We meet Sian Berry, the Green Party candidate for mayor, in a small cafe on Kentish Town Road. Having served as a councillor in Camden and worked as a campaigner for the Campaign For Better Transport since June 2011, it quickly becomes clear that transport is a topic she feels more than comfortable talking about. We begin by asking her what she feels has changed since she last represented her party in the 2008 mayoral race.
“Well since 2008 we’ve had a mayor, Boris Johnson who’s really not, I don’t think, made the effort to plan for the future of the city properly.” She says.
“There’s a lot of things he’s doing that, as far as I can see, mean being buffeted around by lobbyists, not really getting to grips with housing developers or implementing the London Plan. He’s often calling in developments to approve them which local councils have already rejected, which I think is outrageous.”
We ask her what she feels, in transport terms, have emerged as the key transport issues in the same time.
“The air pollution side of things has become much more… well, all the parties are taking notice.” She says.
“So for ages we’ve all been running around saying: ‘The air is illegal. You should have achieved these targets by 2010! Here’s the law. It says you need to do it as soon as possible so why the heck are you expanding roads and airports? Why are you contracting the Congestion Charge?’ All of these decisions, arguably, are illegal decisions that the government are making, that Boris Johnson is making. And now people are starting to accept that this is true and that they have to make a proper plan.”
“I mean, this road?” She continues, pointing out of the window at Kentish Town Road. “Two and a half times the legal limit for air pollution. There’s been a lot of community work going on. Measuring it in local areas, raising awareness in bits and pieces of London, but now it’s becoming a London-wide alarm.”
That the situation has become so serious, she points outs is obviously a negative. That it is something now receiving wider attention is not.
“That’s a good thing, from the point of view of getting something done.” She explains, before giving an example. “I mean, the government have been trying to get London to relax its parking standards. Air pollution is a very good reason to argue against that. As well as things like new roads and the Silvertown Tunnel.”
“You reduce the need to travel by putting things in the right place.”
Addressing the issue of air pollution and congestion, she stresses needs more than simply looking at roads and traffic levels in isolation.
“It’s about integrated transport planning.” She says. “That’s very much something that’s come from my work at the Campaign for Better Transport and from working with CPRE [Campaign to Protect Rural England] and various other big transport NGOs on this. We all agree. Reducing the need to travel… you’ve got to do it in the right hierarchy.”
“You reduce the need to travel by putting things in the right place,” she explains, “concentrating your development around existing transport links. Where you put new developments in place you make sure they’ve got public transport links. You also make sure they’re self-contained communities where you make sure people can walk and cycle to essential services. All of that stuff needs to be brought into London’s new plan and its infrastructure plans. We’ve got a good chance now to do, and change, that.”
“That’s something Boris Johnson’s not been particularly good at.” She continues. “I mean, last week he announced again the road tunnels plan, which he keeps half announcing and reannouncing. They would be an absolute disaster, not only being potentially a ‘predict and provide’ bit of transport planning, but sucking up money that you could be spending improving outer London bus routes, outer London trains, cycling or walking provision all over London.”
The need to look outside of the centre is something that she feels has been missing, both in the last four years and in some of the current proposals from
other candidates.
“In the last four years [the mayor’s] got miles better on cycling investment, and the actual things they’re investing in have got better as well – and that’s largely thanks to the cycling campaigners who ran an amazing campaign in 2012 to make sure we were all signed up to segregated routes and Dutch-style infrastructure.” She says.
“But if you look at where [investment] has been it has only really been in the centre of town.”
“And the Congestion Charge is only in the centre of town.” She continues. “I mean, even Caroline Pidgeon is still only proposing an extra bit of congestion charging around Heathrow, for the Ultra Low Emission Zones to be extended in certain parts of London.”
“And the workplace parking levy,” she laughs, “which is something that the Campaign For Better Transport invented a couple of years ago and Nottingham implemented – we’ve got good links in Nottingham with the Greens there working on that scheme – she’s only proposing that for the very centre of town.”
“But if you actually look at where the work-place parking spaces are, where the ones are that could most easily be retired are, where people are maybe driving to work out of habit or because they have a parking space, not because they’re absolutely car-dependent. Those places where we’d get the most benefit of that are often outside the centre of London. Work-place parking in central London is already going down for obvious land use reasons and the fact that most people using central London don’t have a car. So there’s a gap there, there’s something a bit timid about what Caroline’s proposing and there’s just a massive gap in what Sadiq and Zac are talking about. Because they’re not talking about doing anything to reduce traffic from a demand perspective and their planning policies are deficient as well.”
Again, she emphasises that tackling London’s road problems needs more than just simple thinking. Nor is it just about cars.
“London’s traffic has been going down for twenty years and now it’s gone up. The rise we’ve seen in the most recent year is two per cent! Two per cent! I mean we have to ask, how in one year, do you get two per cent?!
“The only real type of vehicle that’s going up is vans.” She continues. “Yes, and leisure trips, but commuting is not so we can keep on the same track with commuting and keep providing more capacity on the Tube, encourage better work balance and keep commuting down. But the threats are vans, and internet shopping and so we need consolidation. Have you seen the latest figures?”
We admit that we haven’t. Pulling a pen from her pocket, Sian opens up her notepad and begins sketching. The resulting graph gives a rough idea of the growth of van traffic, showing it curving upwards as a proportion of all road usage.
“Now this is very rough, but extrapolate that…” She says, and continues the upwards curve across the page.
“Now that extreme is not going to happen, obviously,” she says, “because we’re talking about a logistics industry here. So, at some point on this curve the industry is going to go ‘oh! consolidation!’ just based on market forces. But we’ve got to worry about where that point comes and we can do much more to incentivise it, and put practical things in place for that.”
More effort, she says, also needs to be made to manage heavier freight traffic, particularly at peak times.
“Again, predict and provide isn’t the answer to this.” She says. “We have to have demand management. But the freight industry are up for solving this. They would love to keep their HGVs out of the city in rush hour, but there’s lots of planning agreements that say they’re not allowed to deliver before 8am and that’s a real issue. You need someone with the political will to go about renegotiating planning agreements.”
“Shops aren’t going to mind.” She continues. “But local people are going to mind that you’re having deliveries early or late at night. I mean, I used to live opposite a Sainsbury’s and the rattling of the cages was awful. But these are all addressable problems. Quieter cages, electric engines, a lot of the reason those engines are running is for refrigeration and that can be dealt with. We can do more with that and we can get the consent to change those planning agreements and then we’ll have real reason to spread that freight traffic out and not have it all in rush hour. And that’s great for safety as well!”
“They’re not against this, the freight industry, but again they’re not yet putting forward their own solutions to it either, and we need to work with them on that.”
“Predict and provide isn’t the answer to this. We have to have demand management.”
Returning to cycling, Sian argues that it’s not just the schemes themselves that need a look, but also the models we use to assess what’s worth doing and what isn’t. There are some positives in London here, she feels, but there is still room for improvement as well.
“TfL are better than the Department for Transport, they look more at the strategic case for cycling, for example.” She explains. “For the Cycle Superhighways, for example, they had a benefit/cost calculation done according to the DfT methods which showed something like a £200m net loss because of driver time increases and bus times and things. But TfL were able to get past that. They were able to say ‘no, the strategic case outweighs this’. That’s a good sign.”
“Another one of the problems we have with the current model, both at the national level and London level is that there’s lots of old, proprietary software. That’s a real issue because it means the data can’t really be opened up. So the cyclists, who were TfL’s allies on the Cycling Superhighway along the Embankment… I mean, we had meetings and we talked to the modelling people and talked about modelling, but they just were not able to open up their model, give us a copy, let us fiddle with some of the assumptions and make the case properly.”
“London could make its own traffic model.” She continues, with enthusiasm. “We’ve got loads of scope to feed in origin and destination data… route data from people’s mobile phones – with their permission of course! – but there’s loads of apps we could use that would give us much more fine-grained data to feed into a new model.”
From a transport perspective, hearing a candidate talk about traffic modelling and investment hierarchies is something most would consider a good sign. We
ask whether she considers the fact that she can do so to be an advantage.
“I think definitely.” She says. “And I’ve had to, in my job and as a councillor, because Camden is working on some excellent schemes to remove road space. Also I think the fact that, as a campaigner, I’ve had to fight through all the detail of this means I have a much deeper understanding of how this all works.”
“And it’s not just the technical understanding.” She stresses. “It’s the understanding of the best way to get the schemes through the planning system as well. I mean, my job as a road campaigner has largely been trying to stop schemes going ahead, trying to argue why a bypass won’t help, but will just increase traffic in the end. But from that I have quite a good understanding of how local communities should be consulted and how they react when they’re not, and how to go through that process.”
It’s that second part, she says, which is just as important as the first, explaining that when communities aren’t engaged with properly it can damage the way they think about schemes.
“A good example of that was Walthamstow [Mini-Holland Scheme].” She says. “There, there was an overemphasis on benefits to cyclists, with whom most drivers don’t identify – although obviously quite a few drivers also cycle. But all drivers and pedestrians will support the amount of the money, up there, that’s being spent on pedestrian improvements and improvements to the town centre, because those are things that’ll improve the local economy. Those are things that are genuinely happening up there and much more should have been made of that.”
Her emphasis on community engagement, we suggest, seems to indicate that she envisions a mayoral administration as much collaborative as dictatorial, at least where cycling and walking are concerned.
“Completely.” She says. “And, I mean, that’s where Andrew Gilligan, the Cycling Commissioner, has been useful. Because the mayor themselves is always going to be very busy. And it’s understandable why TfL’s people might be as well. What we need, like Andrew Gilligan has been, is an expert cycling and walking person at a London level who is a political appointee. Someone who is out there speaking and negotiating with boroughs, and getting
the boroughs on board.”
“The other thing about the next phase of these schemes is that we have to send them to where people want them. And the big Mini-Holland Schemes where it’s all within one area, all done at once, and where every bit of the scheme has to be done? Perhaps that’s not the best way. We might want to start by seeing which communities want them and getting some really good schemes done, so then the other communities come round to the idea.”
Dealing with pollution and London’s road space, however, isn’t the only issue she sees the city as facing. Rail and bus fares, and their impact on people’s daily lives, is something she feels also needs serious attention.
“Five of the poorest boroughs are now London boroughs,” she sighs, “and that’s not how it used to be. The flat fares policy? That’s the motivation for having that. Because it’s just not fair. You’re forced now to live further out and suddenly you’re in zone five or six and your fares have now gone up astronomically. I mean, the difference is huge and the very least we can do is reduce those differences.”
The result of this thinking is a proposal to reduce the number of zones within London to four from next year, and to bring in a flat fare structure by freezing outer London fares whilst inner London fares are left to rise with inflation.
“What I really wanted to do was introduce a flat fare,” she explains, “so I got the data off TfL and we ran a model year by year. And you can get down to every journey that includes zone one being one fare by 2025. I think you still have to include a discount for avoiding zone one, but I’d rather not have a zone though – and obviously we don’t have to make a decision on this stuff now – but what we might want to look at is asking people simply to avoid certain stations. A ‘Monopoly’ option where it’s easy to remember which stations to avoid.”
“Because some zone one stations aren’t interchanges and are not so busy, and it’s a bit arbitrary when the real priority is to be able to say ‘look, if you can avoid Victoria you get a discount’.”
Redefining the zones is certainly one of the more radical proposals put forward by any of the major candidates. This is something that Sian is happy to acknowledge, but says that challenging the traditional thinking is important.
“TfL are very wedded to the zone system, and they have been for years.” She says. “No one’s even thought of changing it. So to have someone come along and say ‘why not change it?’ Well, they could have found that incredibly challenging and just said ‘no’. But I had a really excellent meeting with them where we discussed the possibilities, and they’re just far more open than they used to be to new ideas.”
She also stresses that any major fare policy changes are something that must also be approached pragmatically and over time, not rushed into. This, she feels, is something that separates her own approach from the immediate fare freeze advocated by Sadiq Khan.
“We – and I mean all the other candidates – are right to point out that just doing that is a dangerous policy.”
She says. “I mean, we’re taking a hell of a lot out of the farebox in my plan, in the end. But I know why and I know the principle by which I’m going to raise more money from drivers in order to get it back. And that’s an integrated policy. We need policies to deter drivers and to reduce demand for road traffic, so we can comply with the legal rules and avoid gridlock, and the best way to do that is with some kind of charging system. And a by-product of that, almost, is a load of money that you’ve collected.”
“And the only way that is acceptable to people,” she continues, “and we know this from all the polling, from all the discussions with drivers, is by spending it on transport alternatives. So spending it on cheaper fares for outer London seems the obvious thing to do. I mean, I talk to people in outer London and one of the reasons they want to keep their cars is because of the cost of going into town and because it’s £8, £12 to go in and out again. So it’s about having a long-term plan and making a statement about direction of travel. Saying ‘over eight years your fares are going to go right down’ and then that’s a real incentive to change. And it’s a long lead time as well so they can make plans. It’s not just tomorrow, and that’s really important.”
“And I think that’s really important, having that long term plan.” she says, with a sigh, “Not just bribing people on the eve of an election with promises that you might not be able to keep!”
“TfL might turn round to Sadiq Kahn and say ‘you haven’t got the money for this’,” She laughs, “and I’ve given TfL my spreadsheets and they haven’t said
that to me!”
Alongside flat fares in Sian’s campaign literature sits the concept of the ‘ONE Ticket’, which would represent a switch to fare calculations based on origin and final destination, rather than mode or points of interchange. She admits that this has proved a more challenging part of her fares policy to define.
“TfL have been a bit resistant to this.” She says. “I talked about one-hour tickets in 2008 and at that point there were still paper tickets, and they were just ‘we’re not introducing that. It’s too complicated’. But now it isn’t so they are actually open to ideas.”
We suggest that to a certain extent opportunities to interchange without penalty already exist between many stations that are physically close to each other, thanks to the existence of Out of Station Interchanges (OSI) within the Oyster system.
“They do,” she acknowledges, “but they’re not promoted. Indeed originally I thought ‘let’s just increase the number of those and publicise it’ but the ONE Ticket turns out to not be as expensive as we thought. So we thought ‘let’s just make that about everything you do’ – about origin and destination. Because I think that the new Oyster back-end system, when it comes in, will be able to allow you to change between buses. And intelligence and algorithms used to work out that you’re on the same journey can be used across all modes. So we’ll do origin and destination.”
“As long as you’re not taking a cheeky return!” She adds, laughing. “And that’s why calling it a one-HOUR ticket would be difficult, I think. Because when Tufnell Park is open I could get on at Tufnell Park and pop down to Camden Town, go shopping and pop back again within an hour. And I don’t think that should be free! That’s a return trip! Time and direction is crucial!”
Given the level of changes she envisions, we ask whether there has been a temptation to scale her proposals back, if only to avoid the tired accusation that Green Party proposals can be too ‘blue sky’ and won’t stand up to scrutiny.
“Argh!” She laughs, with a hint of frustration, “But because of that attitude we know we’re going to get that scrutiny! So we think everything through twice as much as the other person!”
“And partly that’s also down to the people we have on the campaign. We’ve got Jenny Jones and Darren Johnson who have long experience on the Assembly and they’ve also got some amazing people on their teams. Tom Chance, who’s our national spokesperson is also a bit of a scientist, like me, so likes to make sure things are right before we put them out.”
“We want to be able to put things out with confidence as well as knowing that we need to. So it’s all good.”
We point out that even with careful thought, radical change, however beneficial, by definition makes future costs. It also makes it difficult to predict exactly where that investment might be needed.
“You’re right.” She says, and once again stresses why this makes effective transport modelling so important.
“There could also be lots of places that people now want to use that we didn’t realise needed the investment sooner.” She adds. “But it’s about being really engaged with TfL I think, and hopefully, as mayor, I’d get on with them quite well!”
“I mean I’ve talked to them a lot – on river crossings, on Cycle Superhighways – and I’ve known a lot of people there as both an adversary and as an ally. And I think I do understand pretty well where they’re coming from. So I don’t think I’d be at all ham-fisted about it the way that Boris has been, for example, with the New Bus for London. I mean, he just imposed that upon them. And the Garden Bridge! I bet they’re DELIGHTED about having £30m of their money promised to that when it has no transport benefit.”
Ultimately, she says, it’s going to be important for the incoming mayor to be flexible anyway. Having a framework of ideas but still recognising the need to adapt them to changing circumstance is a strength, not a weakness.
“It goes back to consulting the public and putting forwards ideas.” She says. “You’ve got to put things forward at the right time, and at the right level. Flat fares policy? I didn’t talk about that until we knew it was possible. We haven’t specified every detail of what the fares would be at every stage, because that would be silly. You’ve got to be a bit flexible to how people react and, indeed, to other forces. I mean, if the government decides to cut back our grant completely then at some point we might not be able to afford to do it quickly. The assumptions in the calculations are around inflation – which is suspected to go up quite a lot, in fact, if you look at the ABR forecasts, which is what we used. But if it isn’t as high as that then obviously the centre won’t catch up as quickly as the outside. It also makes some assumptions about passenger growth.”
This thinking extends to other aspects of transport policy, she says, as well.
“I understand why some journalists can get a bit frustrated at the fact that when they send us a set of questions asking exactly what’s going to happen in 2019 under this exact set of circumstances that we can’t answer that.” She says. “But we deliberately can’t, in a way. We can put it forward in the right level of development for a manifesto, we are putting it forward to see the public’s reaction and we are putting it forward for further development.”
“There are simply aspects where I don’t want to decide now. I’d rather wait and see how it goes, or talk to people first. And I think we’re putting it forward at the right level of detail. And, you know, I can explain in detail why I don’t have the number, or where I do have, but I don’t think it’s a robust number and therefore I don’t want to put it out.”
We finish by turning to the subject of housing. Given her clear belief in the need for integrated transport, we ask, how does housing fit into that mix?
“Obviously we’ve got brownfield sites, and everyone is talking about TfL’s massive estates.” She says. “They’ve just signed a deal with 14 massive property developers. And I think they want everything to be like King’s Cross, but I’m not sure that’s the right approach. I think what we need to be doing a lot more is using the smaller sites better. The big companies also tend to do things much slower, I mean King’s Cross has taken years and years and years. Getting a massive masterplan through, especially when you’re knocking down some things, takes a long, long time. So with smaller sites, if you give them to smaller developers, give them to co-ops, self-builds, councils, people who have people queuing up to live there or people who want to live in the houses, then they’re going to get them done quickly and be less controversial. So I think a small-sites policy has to be a big part of the mix.”
“Long term we also need to have a discussion about the green belt and one thing Zac Goldsmith’s been saying is that it’s okay to go out and use some of the green belt, as long as it’s near railway stations.” She pauses and laughs. “And okay… that’s transit-oriented development, I suppose, so it’s not completely wrong. But it’s too soon to be doing that before we’ve made the most of the brownfield sites. We’ve got to stick to the planning principle that
we know has worked and has reduced traffic over the years – which is town centre first and increasing densities in walkable distances. Not just public transport distances.”
She goes on to suggest that green belt development is also too simplistic because it can fail to take into account the impact such development might have on public transport further down the line. But there are also, she stresses, some other issues that are often overlooked.
“You have to think about what goes into, and is around these developments.” She says “If you’re not near the hospitals, and if you don’t build cultural venues – and Darren Johnson’s done some really good work on music venues – then you get whole areas like Nine Elms being built with no music venues whatsoever, hardly anything for culture, hardly anything for social life and so suddenly you’re generating loads of trips there. I mean, I mentioned the growth in leisure trips before. There’s a real danger that we end up concentrating all our leisure and shopping opportunities in one place and we kill our local high streets as well.”
“And we address that at a local level. But at a London level we’ve got a clear transport planning reason to encourage it as well. It’s not just about diversity, or heritage, or preservation. It’s about what city we build for the future.”
That future city, she stresses will need to be built on integrated transport and community, and borough, engagement, but will also need a mayor who accepts that it is impossible to have all the answers now.
“It’s un-politician-like not to just pledge everything and promise everything isn’t it?” She laughs. “I’m saying ‘we can do that’. You’ve seen the language I use about fair fares – ‘doesn’t everyone want to do this?’ And hopefully that’ll get taken up.”
“I mean, hopefully, this will be everyone’s policy in the next election!”
“What’s really good about us Greens, I think, is that we raise good ideas and we put things forward that other parties can take up.”
We suggest that her openness about her desire to see her proposals appear in future manifestos by other parties is somewhat unique.
“Yes.” She, laughs. “Well, what’s really good about all us Greens, I think, is that we raise good ideas and we put things forward that other parties can take up.”
She says. “We’re not belligerent about it. Now if I was to be all belligerent and say ‘flat fares?! – you’re never going to do that!’ or ‘you’re all going to think that’s awful’ then other people wouldn’t react well to that, and they’re never going to think about doing it. But if you say ‘this is a great idea, don’t you want this?’ then you can keep pushing for it, keep studying it.”
From her final statement, however, it becomes clear that this openness is about promoting real change, not just ideas.
“There are stories,” She says, growing serious once more, “about people who have really low-level jobs, who are cash poor, and can’t afford to get the Tube or a Travelcard. They will get on one bus and stay on it, for the entire journey, even though it goes completely out of their way, because of the reduced cost.”
“And that’s just…” She sighs and looks up, with determination in her eyes. “That’s just unfair. That’s just… We can do away with that with our policies.”
This article first appeared in LR Magazine issue three. If you enjoyed it and would like to read more, or simply want to help us keep writing, then why not buy a subscription?
If you’re looking for partnership opportunities for an event or your company, then speak to us about our print advertising guidelines and options.
Fascinating article, and the page layout at the top looks great.
You mentioned that LR was trying to speak with the leading candidates – did you succeed, and if so will there be coverage before May 5th? (Not that it will swing my vote!)
I really REALLY like that she talks about the big picture, understanding the links between transport, development, local economies etc and how to shape them, but I have a lot of concerns over the flat zoning.
The first is that it’ll only work within the capacity of the tube and train system, and Berry explicitly states that the driver for flat zoning is to remove the barriers that exist for people in the outer zones using the tube and rail to access work opportunities. If it achieves this aim, it’ll create huge problems at stations within zones 2, 3 & 4.
The other, more major issue though, and one I’m surprised she hasn’t grasped, given that she uses it to criticise the other candidate’s proposals, is that flat zoning will turn many outer suburbs into dormitory areas, as zone 1 hoovers up growth in the jobs market.
For example, she’s right to point out that it’s ridiculous that Nine Elms has no cultural or social facilities, and she points out that we should be looking to town centres as the main focus of development, but she doesn’t seem to grasp that flat-zoning will have huge impacts on local and regional economies and shops etc., in areas that will really struggle to evolve into cycling-centred sustainable neighbourhood areas.
And with my architect/urban designer hat on, there simply is not the land to build the houses London needs without using the Green Belt. If you’re going to build the houses that London needs, you will have to push new and improved transport infrastructure into the Green Belt if you want to realise the goal of integrated and sustainable communities.
I also thought that it was interesting that she went after Caroline Pidgeon so much too – the Greens obviously feel that LibDems-sympathisers are a winnable constituency.
The flat fare would lower a Zones 1-6 fare including season tickets, so would this result in a lot of ‘border tourism’ i.e. current Home Counties commuters driving to their nearest TfL station to save money?
Why have I bothered buying LR magazine 3 (to see what it’s like) (not that I’ve got it yet) when you seem to be publishing the content on-line anyway? Sian seems to have a grasp of reality which I like.
An interesting article with some interesting ideas, but as is often the case with the Greens, some seems to be a degree of naivete when it comes to understanding how politics works and the cost of making significant changes.
I used to work in the logistics industry and it operates on wafer thin margins. The owner-operator (ie. one person in their own van) is always cheaper than a larger company with many vehicles because the overheads don’t scale up well. In the late 1990s/early 2000s when IT systems were introduced into Logistics, the larger companies became more efficient through tools such as route optimisation, vehicle and package tracking, load management and vehicle usage optimisation, but these sorts of tools have filtered down to the owner-operators and so any economies of scale have been wiped out. This is why companies like Amazon now use owner-operators rather than big logistics companies, because Amazon provide the optimisation systems but don’t have the overheads of operating vehicles.
So whilst logistics companies will be happy to talk to politicians about consolidation and changes to delivery times – anything to swing trade back to them from the owner-operators – they have no money to invest in improving solutions themselves. Consumers like free delivery, but that results in more white vans on the road. Getting rid of the vans would probably require unpopular legislation – requiring each delivery “company” to have a license for example, or introducing drivers hours on all delivery vehicles.
An interesting article and more “transport” than “politics” which is refreshing for a change. That article shows more clearly than I’ve seen via social and conventional media that Ms Berry does appear to have a clue. I like the fact she recognises the linkages and complexity between a number of key issues like transport, development, land use and then how people react to it.
I completely agree with the remarks about places like Nine Elms being soulless and with little or no local social activity or even the prospect of small scale businesses being able to afford to open there. I see similarities with the Olympic Park – only corporates like Sainsburys can open shops. I assume this is because the rents are way beyond what can be afforded by local business people. She then rightly points out that all you do is create a lot of unnecessary travel demand as people *have* to go elsewhere to do anything. People might condemn places like Old Kent Road, Plumstead High St or Walworth Rd as “dumps” but at least they can still manage to sustain a massive variety of niche and local businesses that serve the local community and allow for more sustainable transport solutions.
I remain sceptical about the fares proposals. It’s all lovely in theory but the avoidance of putting out a fare value is a cop out no matter how she frames it. Yes life changes but you can sit on the fence forever. It doesn’t take a genius to work out what happens if you basically pull outer fares down and price up Zones 1, 2 and 3 until they meet. You end up with a massive disincentive for short hop travel. Perhaps she wants that in order to force people to walk and cycle and to allow public transport provision to be scaled back? It would be nice to know. I find it odd that she states that the current Oyster system can’t do “origin and destination”. Colleagues and I deliberately wrote it into the spec – yes through single fare, multi modal through charging. Perhaps it was taken out of the spec after I left the project?
I also note with interest (and approval) that she doesn’t want her intermodal “one” ticket to be used for return travel for one fare. This, of course, makes administration of the journey and charging data much more involved. This aspect has not been covered by anyone else proposing “one hour” bus tickets but it’s pretty fundamental to how a number of transfer / multi ride ticket schemes work. If people are allowed to make return trips on 1 hour tickets then the income loss will be even higher.
There is a lack of commentary from Ms Berry about larger scale transport investment which I find a little surprising. Perhaps it had to be taken out of the interview to avoid being too long? No mentions of trams (which I think the Greens support) nor buses, tubes, trains or Crossrails. The manifesto acknowledges the need for more investment but isn’t big on committing to large scale spend even though some level of that will remain essential in London for decades to come (if only to keep upgrading / replacing worn out assets). The discussion about vans and freight is useful but I suspect “controlling” those will be a real struggle for whoever is Mayor. There is also the question about who funds changes to vehicles, consolidation etc etc.
@Jim Cobb – another type of consolidation is the delivery lockers that Amazon have introduced in various locations around central London. So a single van can make one delivery out of hours rather than driving around houses doing multiple deliveries during the day only to find that people aren’t in and they need to leave a note through the letterbox…
There may be potential to expand this sort of scheme although I don’t know if it could work for multiple retailers and logistics companies and clearly it won’t work with larger or perishable items.
… also, I’m sure there was a comment on a previous LR article that the convenience supermarket chains basically use their delivery vans like mobile stock rooms because the stores themselves have little space set aside for stock in order to maximise shelf space.
@WW
“I see similarities with the Olympic Park – only corporates like Sainsburys can open shops. I assume this is because the rents are way beyond what can be afforded by local business people.”
Rents are a big factor, but the main ones are that developers like to have the leases sown up before a project completes, for the leases to be of a longish duration, and most importantly, for there to be as small a risk on their side as possible.
This means that unless you have a particularly benevolent developer, or strict planning conditions, chains and franchises are always going to take the majority of the retail spaces underneath new developments.
Jim Cobb mentions “introducing drivers’ hours on all delivery vehicles”. That could be expensive, and goes against the grain of “deregulation”. But it could also have significant safety benefits. Safety is the ostensible reason for all drivers’ hours rules.
There is a lack of commentary from Ms Berry about larger scale transport investment which I find a little surprising. Perhaps it had to be taken out of the interview to avoid being too long?
The approach we took with both this and Caroline’s piece was to encourage a conversation that went into depth about something (or some things) that they felt were important rather than to try and give a full overview of their policies.
Full overviews exist in lots of places elsewhere, and – to put it bluntly – I’ve always felt it is far easier to feign knowledge when you’re spreading your conversation topics thin rather than demonstrate your qualities and knowledge as an individual. It’s much harder to do that when someone says to you: “okay, you think that X is important? Let’s explore that in a bit more depth.”
Why have I bothered buying LR magazine 3 (to see what it’s like) (not that I’ve got it yet) when you seem to be publishing the content on-line anyway?
That’s why the magazine features are ‘timed’ exclusives not just exclusives. As I mentioned in the launch post we wanted a model that allowed people to support the site without it compromising the overall model of creating inclusive content.
What the magazine does do, however, is add content to articles where we are revisiting topics – e.g. the extensive photos for Harrow & Wealdstone and Brunel’s tunnel or additional content on St Johns.
As this is as good a place to mention it as any, we will though be adding at least one print “Special” this year for Subscribers (at no extra cost). Basically it looks like we’ll have the advertising to pay for it, so figured it would be a nice, exclusive “thank you for supporting us.” As it stands, the topic for that is likely to be innovation and disruption in the rail industry.
not that I’ve got it yet
To my knowledge we’re up to date with the post (with the exception of three orders)! I’ll check your email address though and do email the “questions” email address if you didn’t put the same one in as your contact address when you commented.
We’ve had persistent issues with the Post Office, although they were much better this time around, and so it is possible that your copy has gone missing – so if I can match you I’ll put another copy in the post to you as I’m (literally) about to do a Post Office run.
We’re looking at alternate delivery options, but I’d much rather sort out the PO issues definitively if possible – as I don’t want people having to go to obscure depots to pick up issues if they won’t go through someone’s door for some reason. It’s a balancing act though.
Anyway, end of magazine side-bar!
@Reynolds 953 – that is true, but that kind of delivery is a small percentage of the overall market. On the flip side, Amazon are aggressively expanding their same-day evening and 1-hour delivery solutions, which makes any kind of consolidation and optimisation impossible as there is not the time to do it.
I used Amazon as an example, but other retailers have to follow or lose out, particularly as consumers get used to free or very quick delivery. Consumers generally choose not make the connection between their shopping habits and increased congestion/pollution. I wonder if Sian Berry uses Amazon !
@WW
“intermodal “one” ticket to be used for return travel for one fare. This, of course, makes administration of the journey and charging data much more involved. This aspect has not been covered by anyone else proposing “one hour” bus tickets but it’s pretty fundamental to how a number of transfer / multi ride ticket schemes work. ”
I’ve inadvertently benefitted from Oyster not recognising such a trip. Going from station A (in Zone 1) to station B (in Zone 3), out through one barrier line, collected what I had gone for, back through the other barrier line within the (generous) OSI time limit and back to station C which is only a few minutes walk from A. To my surprise I was only charged a Zone 1 fare.
marckee 22 April 2016 at 14:45
I think you’ll find that the majority of East Village E20 retail units remain vacant.
It is difficult for neighbours and visitors to pop in for a meal due to ferocious car parking restrictions until as late as 21:00 on private land and quite long walks from bus stops.
Also vacant are a large minority of the homes west of Celebration Avenue.
@ Marckee – thanks for the info re developers and leases. Looks to me then that we are building in a lot of problems for the future. We are creating a tie in to corporate businesses with probably little immediate competition meaning higher prices and a lack of retail diversity. Most older areas manage to sustain small corner shops, take aways, bakeries, dry cleaners, hairdressers, chemists, post offices, cafes etc. If you’re very lucky you’ll get a butchers or a green grocer. Alternatively there are long established traditional markets offering a mix of lower price goods or fruit and veg. It’s quite clear people like to have these local choices otherwise the businesses would go bust.
I don’t see how a developer led model can ever permit these options for people which means residents in these new areas are going to be deprived of choice or forced to travel. That strikes me as two unsatisfactory outcomes for starters. It is also means London is going to become very sterile and boring as these horrible developments crowd out the old neighbourhoods. Let’s hope the next Mayor starts to flex some muscle and encourages the boroughs to push for more diversity through the imposition of tighter planning rules. I suspect, though, central government wouldn’t like to see any of that because it wants deregulation and less control.
Re JIm Cobb,
Delivery Hours – it would work for deliveries to big retail groups or chain restaurants but many offices are only staffed to accept deliveries between 0800-1800 so in the middle of the night won’t work for many places. So I suspect removing restrictions on bulk deliveries in larger vehicles (including construction materials) is what we really mean rather than <3.5t vans being restricted.
re Reynolds 953
“… also, I’m sure there was a comment on a previous LR article that the convenience supermarket chains basically use their delivery vans like mobile stock rooms because the stores themselves have little space set aside for stock in order to maximise shelf space.
Yes and I suspect that comment will have been from me!
One of the supermarket groups recently opened a new mini format store so it was interesting to watch how they initially stocked up the store and also using vehicles that were too wide for the loading bays outside preventing 2 way flow of traffic on an A road, this ceased after 4 day of opening obviously when the parking fines hit the relevant person desk. A council parking warden earned his commission after spotting the carnage caused by the very first delivery vehicle and camped out there for a few days! (so did the police as the borough office is just up one end of the road the road and the council offices at the other end).
A previous linked comment:
Also worth noting that none of the tradesmen in the city tend to have any spare parts stock on them (minimising working capital) so it takes 3 van runs to get anything fixed. 1x Tradesman 1x part delivery 1x Tradesman.
@Alan Griffiths:
My point was about new developments generally – too many new blocks of flats have a Tesco Metro or Sainsbury’s Local underneath them for my liking, and for the best interests of an area in general.
I actually think that the East Village area of the Olympic Park has been one of the more successful in terms of getting in independent retailers. A look at this map shows that they’re pretty much all let and a surprising amount are not to chains or franchises:
https://yonder.e20.org/t/shops-and-retail-units-in-east-village/232
As was touched upon in the comments under the Caroline Pidgeon article, I think lack of parking will not be having the impact that you suggest – far more likely factors are that the Olympic Park is still a construction site with people and institutions yet to fill the area, plus the close proximity of Westfield.
It’s all very well arguing for night time deliveries to shops, but the main problem is not planning restrictions (though these do exist) but that most small and medium sized businesses cannot afford too staff up their premises out of normal hours just to receive deliveries. Even the big supermarkets, as Reynolds 953 points out, use their delivery vehicles as mobile stock rooms. Big supermarkets will continue to need 5 or 6 carefully timed deliveries throughout the day as they have no prospect of storing all the stock needed for one day’s sales on site. Although the logistics industry would love to move most deliveries to night time, practical business economics mean it’s unlikely to see a very big increase irrespective of any regulatory or exhortatory moves.
@ Timbeau – not so much benefitting from a policy than it simply being far far too involved to try to work out it was two journeys rather than one. There are all sorts of “relaxations” in the TfL fares / PAYG charging structure if you know where to look. Sometimes it makes sense to take a notional “loss” on some origin / destination pairs than tie the technology in knots trying to charge people a higher fare. Of course that becomes more difficult when your income is being eroded through subsidy cuts and fare freezes.
Oh and coming back to Ms Berry’s fares policy she seems unaware that the DfT have got any Mayor and TfL tied into a “no cliff edge fares changes at the boundary” policy. Her proposal to reduce outer London fares considerably would create a massive cliff edge unless she is going to shovel tens, possibly hundreds, of millions of quid from City Hall to the DfT as “revenue compensation” for every TOC that runs into London (regardless of any TfL takeover of inner area services).
I think the emphasis on small-scale and local is right. Currently, the money seems to be going for the big, just like Nine Elms, leaving the local borough to try to fill in all the missing ingredients. If a developer thinks their development will sell regardless, they will have little regard for the future pressures on transport or social cohesion.
The green belt doesn’t have to be sacrificed. The outer boroughs have loads of potential for densification. The transport issues that creates are challenging, but no more so than building around existing transport hubs in the green belt.
I agree that the one zone fares proposal seems to lack any appreciation of reality. We have discussed the 1 hour bus ticket already after the Lib-Dem interview. If the Oyster system can detect that after a change of bus, the journey has reversed, then it should be possible to police that. There would be some journeys where an (unpublicised) relaxation was allowed, just to make the system simpler to operate, and avoid expensive over-complication.
@timbeau
“I’ve inadvertently benefitted from Oyster not recognising such a trip. Going from station A (in Zone 1) to station B (in Zone 3), out through one barrier line, collected what I had gone for, back through the other barrier line within the (generous) OSI time limit and back to station C which is only a few minutes walk from A. To my surprise I was only charged a Zone 1 fare.”
I’ve done something similar. I went to Liverpool Street from Stratford via the Underground system, then picked up a replacement box from Richer Sounds, came back though the TfL-Rail barriers from Liverpool Street and changed to the DLR and touched out at Stratford International.
I got charged for a single Zone 1 to Zone 3 fair, not two as the Liverpool Street-Liverpool Street was taken as a OSI.
@Fandroid If we are to meet the anticipated housing demand for London (and I appreciate that this ‘predict and provide’ methodology is something that is not tolerated in road space provision these days, and we should question whether doing the same for housing might have the same effect), there are essentially two choices:
– push through the demolition and rebuilding of areas in proximity to existing tube/rail stations.
– improve or drive through new public transport infrastructure in areas where new housing can be accommodated.
Either way, unless we accept that very large areas of the capital should be demolished and rebuilt, the numbers we are looking at cannot be achieved without the sacrifice of some Green Belt, however much every Mayoral candidate claims otherwise:
https://england.shelter.org.uk/professional_resources/policy_and_research/policy_library/policy_library_folder/report_when_brownfield_isnt_enough
I note Ms Berry’s discussion at the end of how significant bus fares can be for the very poorest people. What I consider to be a major argument for a “one-hour” or some other free-transfer kind of bus fare is the disproportionate and unpleasant choice between inconvenience and expense that bus disruption causes for the poorest commuters. For example, repeatedly in the last few months I have noticed that in the early evening rush hour there have been thirty minute gaps between eastbound buses on routes towards the City from Tottenham Court Road because all the buses heading west have been turned short at Holborn Circus or Red Lion Square due to heavy traffic.
I don’t want to reignite a general debate about the rights and wrongs of turning short, but giving people chits for further travel only helps people who are on the turned bus. It doesn’t do anything to help people waiting at the terminus who are faced with waiting forever and travelling on a hugely overcrowded bus, or paying a double fare. Which is especially bad since bus commuters tend to be some of the poorest people in London.
I’d love for her to win the mayoral election, but the UK in general is rapidly approaching a US style two party system, and I think it unfortunately find it unlikely that someone from a similarly small party or an independent would win such a mayoral election. 🙁
“Increasing densities in walkable distances” = densification. I wonder how many people in outer London would pop down the polling station to support that?
@ Anonymous 17:15 – don’t forget the elections on 5 May are for mayor, constituency and “party list” and Ms Berry is number 1 on the Green’s list.
Given the Greens had 2 GLA members in the 2012 election I think it is likely they will have at least this number in 2016 so Ms Berry stands a good chance of being a GLA member, if not mayor…
@Anonymous
“I’d love for her to win the mayoral election, but the UK in general is rapidly approaching a US style two party system”
[Historical digression snipped. LBM]
The London Assembly *members* are voted in using the statically amazing d’Hondt system that allocates seats so they match the share of the vote, allowing the smaller parties such as the Greens to have a decent part.
The mayor can be of any party or none, the system (unlike the voting abomination at Westminster) is free and fair, democratically.
[Some bold claims here Briantist, It is possible that someone might want to challenge you on some of the details, or some of the adjectives. But I would ask that they don’t, as we don’t want the thread to veer too far away from the base subject. Malcolm]
Political allegiance aside, while she makes good points, I feel that a key point overlooked by herself is capacity. While plunging the fares is lovely in principle, it is simply not currently feasible with the upgrade work and new lines needed. Raising the congestion charge and reinvesting into the Tube is simply necessary if you’re suddenly opening the network to tons of ex drivers and less affluent bus passengers who can now afford the train.
There’s two umbrella strategies for removing cars from roads: Carrot or stick. Personally I think that a more effective method of removing cars from London’s roads is to create a park and ride style of system for those arriving from the commuter dormitories. Speaking from personal experience when I was younger, if we were going to Central London then we’d come down the M11 and park at Redbridge or Epping and get the tube into town. For me, that would be a much better way forward. Parking at Epping is already full by 6:15am. Expand those facilities and you lower emissions in town, and easing pressure on the North Circular and A12. A similar view could be adjusted for other stations: [Other examples snipped. Malcolm]
This doesn’t really fit into the fare structure of Green policy, however. People in Harlow, for example, use Epping because the NR station is such a price hike. Bringing suburbans under TfL control would go a way to resolving problems like this: Make no mistake, the Underground puts an equivalent NR station out of business, every time, even if it results in a longer journey. I believe there’s campaigns to reopen North Weald to this day.
South London obviously doesn’t have the luxury of a tube station to park in and hitch a ride into town (except perhaps Morden) – so when the Bakerloo extension gets built, it’ll be massively oversubscribed. It makes me wonder if building it to Tube gauge, then, would be worth it. Are there any alternatives? Perhaps the Bakerloo could be split into two lines.
[Digression into wax ballast snipped. LBM]
@bakerludicrous
2Parking at Epping is already full by 6:15am. Expand those facilities and you lower emissions in town, and easing pressure on the North Circular and A12. ”
but massively increasing pressure on commuters further down the line. If the trains are full of car commuters on leaving Epping (or High Barnet, or Chessington (nice big car park at World of Adventures…..) , or Tattenham Corner (all the racecourse car parks lying idle most of the year…..) what use is it to anyone in Zone 4 and 5?
@bakerludicrous
Big attractive car parks for park and ride at the edge of London would also provide a massive incentive for those who currently take the train from places like Chelmsford or Redhill to abandon the train and drive to the park and ride site, thus benefitting from lower fares. The net effect, of course, is wholly counterproductive.
What the proponents of park and ride on the M25 do not seem to understand is that very, very few people commute by car from beyond London into central London. Similarly, a very tiny percentage of those cars in central London (less than 0.1%) come from outside London. Not only has congestion charging cut the overall number of cars in the centre but they are now only a tiny percentage (about 5%) of the traffic in the centre in any case.
Park and ride for London as a whole is not a viable strategy unless the ‘park’ is in the car owner’s front drive.
The trouble with park and ride is it just doesn’t scale.
If you wanted to give a car parking space to everyone on a single tube train, you would need floor area equivalent to about 3 football pitches.
There is no way significant quantities of land will be used just for car parking given its scarcity and value in the south east.
A signifficant number of people in the GLA area own cars, not to take them into London, but because they need to travel outside of it (or round it) for work purposes. What many in Politics seem to forget is that while public transport in London may be good, with all the facilities (shopping and social) in easy reach of said PT, the same cannot be said outside London.
Therefore, has been noted, however noble the suggestions of Ms Berry might be with regarding lowering zonal fares the effects on persons (or indeed infrastructure demands) outside of TfL / GLA land need to be considered – and as the DfT have made quite clear with respect to fare boundaries they will not accept things that cause them issues. This is also true of road maters as the DfT will not sit idialy by if the private sector complains too much that the Mayor is unfairly greeting them.
As others have hinted at talking about reducing the need for van movements is easy, actually getting London voters to give up their ability (which nowhere else in the country has) to have stuff delivered from Amazon within the hour (by one of those said vans) might be a tad harder.
Finally I note that, as to be expected Ms Berry is another ‘no vehicular river crossings in East London type. Such people seem to forget that not everyone works in ‘office’ type jobs – if you are a tradesman or a business that sells big things then it Is unreasonable to expect you or your customers to cart stuff round on public transport – which means that road transport has a big part to play. Yes emissions are an issue but with advances in hybrid or electric technonolgy it can be dealt with. Rather than effectively sticking their fingers in their years when such matters are raised or try and claim its someone else’s problem (i.e. the DfT to sort out via Dartford) they should recognise the truth. Namely that if you want East London to develop in the same way as west London then extra ROAD links across the river (even if you keep them as single carriageway standard) are essential.
Here in middle commuter land SWT are busy putting decks over their car parks, so those extra football pitches of space are created on the same footprint. Park and ride has its place, but only where walking or buses are not practical. Those parking places must be aimed at the local catchment, and not be the mega-parking that attracts people away from local trains and distorts the travel market.
As for river crossings, I suspect that a crossing that gave local connectivity only (a modern Rotherhithe tunnel) would be grudgingly accepted by those who don’t want huge junctions and approach road systems. However, it would still be massively expensive and its limited capacity would never wash its face in any cost benefit analysis. Having said that, there might be some sort of case for a rail bridge which included limited vehicle capability together with foot and cycle traffic.
Could everyone please hold back on East London river crossings, as we have an article in preparation about this issue.
quinlet
What the proponents of park and ride on the M25 do not seem to understand is that very, very few people commute by car from beyond London into central London.
I have occasionally driven N up the M11 at the tail of the AM peak.
I don’t believe you, having seen the crawl-to-standstill on the S-bound lane, backed up to at least the Debden junction.
Ms Berry’s party’s proposal for a flat non-zonal fare is, IMHO, as Upney as Mr Khan’s fares freeze. Oh dear.
Greg: How do you know, without asking them all, how many of the occupants of the cars you saw were on their way to central London?
But that apart, we probably all know that many cars come into London (as a whole) from outside during the morning rush hour. But the idea of large park-and-ride car parks near the M25 seems flawed to me, until and unless there is extra rush hour capacity on a relevant railway line. Anyway, any such capacity which is ever added will probably be immediately taken up by occupants of new build houses within London.
It’s the whole approach to residential development by all candidates that is fundamentally dishonest. It’s local democracy inaction with mass petitions organised against every single housing development proposed whether it’s for a 20 storey block or three terraces on an old garage block.
Theses large scale schemes take a decade to come about because of the current planning system and local democratic input. While smaller schemes are quicker they face just as strong opposition from local residents and I’m not sure how them being built by warm and fuzzy co-operatives are going to make them be built any quicker.
London’s population has increased by over a million in over a decade. How has it actually responded to this so far. Soaring property prices and vastly increased overcrowding with more people sharing accommodation and the growth of new slums with people sleeping in garages and garden sheds. Meanwhile commuting has soared from further afield.
Meanwhile the large schemes have chugged their way through planning and a big recession. Meanwhile everyone is fighting small scale schemes as residential values soar above the rents available from retail, offices and pubs.
The reality is if you want London to cope with another million or two people then large scale greenfield building and redevelopment in the city will be required. Where ever these extra houses are built you will need extra transport facilities to move them from their homes to their jobs. Plus we will need extra employment zones as how are we going to physically fit all those workers in Zone 1.
@ Reynolds 953 1730 – clearly the Greens are pushing extremely hard to at least maintain their number of Assembly members. However I think they are going to have to be very lucky to retain two members. If you look at the polling then it seems to show UKIP in third place which may well give them 2 or 3 of the Londonwide seats. I think the Greens and Lib Dems have a big fight on their hands to hold on to their combined 4 Assembly seats. The make up of the next Assembly could be very interesting in terms of the weight given to any investigations they undertake and if they can ever get a 2/3 majority vote to carry their motions / affect the Mayor’s budget.
@Malcolm and others:
In all fairness, I did mention that [as an example] on the Epping branch near the M11, with car parking improvements, firstly the branch should be taken over by Crossrail 2, which I believe has been previously planned for. High capacity, full size trains on surface platforms that can be lengthened as well. [Presumably most Central line trains would then run all the way around the loop and terminate at Woodford, to improve interchange]
Cockfosters does not have the fortune to be sighted outside of the M25, which makes it less useful in that you can’t avoid it. That’s probably a good thing though, considering that the Picc, with its age, doesn’t need any more pressure on it now.
While in some cases it may be true that improving car parking will suck people away from the mainline and onto the tube, Epping is unique in that Harlow, for the size of town, is not a well used station – it is vastly overpriced considering the proximity of Epping, and rather than TfL bending down to the DfT, for the good of the city the National Rail service should be made more competitive. And despite that Cambridge, for example, has a good enough and competitive enough rail service (and, for good measure, bus and bike facilities) that the majority of users won’t be attracted to use the Tube.
As for the M11 being stuffed to the gills every morning, for me, that indicates most of those workers are destined for London. There can’t be that many people travelling south of it, because that is *a lot* of time for a daily grind. You could well be stuck on either or both of the motorways all morning if events conspire against you!
We cannot assume that nearly all M11 southbound morning peak drivers are London bound. There must be some research somewhere to enlighten us.
I know one of the M11 cloggers and he commutes from north of Harlow to the outskirts of Dartford (an improvement on his old commute to Slough).
Many commuters drive from beyond the suburbs to INNER London, but few will drive all the way in to CENTRAL London – there is a big difference.
One statement of Sian’s that I find difficult to believe is:
“ I mean, I talk to people in outer London and one of the reasons they want to keep their cars is because of the cost of going into town and because it’s £8, £12 to go in and out again.”
Has anyone here met such a person? A car owner in outer London who uses their car to go in to town. And if there are such people, it must surely be for leisure, in evenings or at weekends, and are they really going to change their behaviour if the tube fare goes down from £10 return to £6? Their first drink would eat up the whole saving!
ngh refers to a large chain stocking up a convenience store “using vehicles that were too wide for the loading bays“.
I’m not condoning this at all, but sometimes people don’t realise why the largest vehicles which will fit are usually used to provision stores. If smaller vehicles are used, then there have to be more trips, more expense for the supplier, but also more congestion (a truck carrying 15 tons of groceries takes much less road space than 15 vans with a ton in each).
@ Chris Mitch – to be a tiny bit pedantic people coming down the motorways may well be headed to a variety of locations in Outer London, may be traversing the A406 to reach another radial road out of London or may, as you suggest, be heading to Inner London.
It is quite clear that few people are heading into Zone 1 for a variety of reasons already suggested. You only need look at some old Thames News footage from the 80s to see how much less car based single occupancy traffic there is in Zone 1. Also as noted by Ms Berry and other candidates we have market shifts occurring with freight, logistics and private hire industries which are affecting road capacity in Zone 1 and elsewhere.
A reasonable place for people to at least get a flavour of what is happening are the Travel in London reports produced by TfL. There are also spreadsheets of the supporting data. Unfortunately TfL have somewhat rationalised the data provision in Report 8 compared to Report 7. Report 8 is the most recent and was further updated in Feb 2016. One of the spotlight topics is about changes in car travel in London showing how trip rates have declined and exploring the reasons why as well as which modes have gained.
https://tfl.gov.uk/corporate/publications-and-reports/travel-in-london-reports
@ Malcolm 1917 – I’m not going to claim to have met anyone making those statements because I’m a grumpy old *** 😉 However there was a lot of comment on social media about the pricing up of the one day travelcard and reduction in the number of off peak variants. This has certainly affected outer London travel if you believe what people say on social media (I’m in no way claiming it’s statistically robust or objective).
In Travel in London report 8 one factor identified for declining trip rates is the fall in household incomes in Outer London since the 2008 recession. Clearly if fares are going up and your income is “under pressure” then you have less money to spend on leisure activity anyway and rising transport costs just makes things worse.
Briantist (in Gigabit internet heaven) 22 April 2016 at 16:17
“to Liverpool Street from Stratford via the Underground system, then picked up a replacement box from Richer Sounds, came back though the TfL-Rail barriers from Liverpool Street and changed to the DLR and touched out at Stratford International.”
These days that would be Zone 1 to Zone 2. Perhaps the anomaly you benefitted from was a while ago and has been corrected?
Re Malcolm 1939,
There happen to be several appropriately sized loading bays (for 2x 18tonne rigids) just round the corner which also happen to be outside their main competitors store! The supermarket delivery drivers probably just didn’t want to push the roll cage trolleys past a station entrance, a take away, a minicab office and a restaurant. They now do use the appropriate ones round the corner probably after the parking tickets and TfL having quite word about 20 minute delays on 4 bus routes…
I’ll check the planning consents but I suspect they will have specified the use of the loading bays round the corner instead of in front.
I note one previous incident where the same supermarket group picked up £75k in fines one year for doing the same thing in a different London Borough a few years ago after opening a new store, so they have form!
ngh: Reminds me of a Temple Fortune store which has most of its deliveries in 18 ton rigid trucks which can just get to the rear entrance (subject to no rogue parking in the awkward little lane). One delivery early each morning is (or at least was then) done by a full sized artic at the front door. All agreed and legal, until one driver of the early morning delivery tried to get to the back door. The vehicle had to be removed by a crane.
Malcom
There’s always a different [Snip for politeness. LBM] way of doing that sort of thing
Oh dear – the very last picture is a fake, btw, the others are genuine.
Sorry for my late-night digression – let’s leave automotive mishaps, real or exaggerated, and get back to our muttons.
@Malcolm:
“Could everyone please hold back on East London river crossings, as we have an article in preparation about this issue.”
A “Coming Soon…” list of topics / titles currently in the works might reduce the need for such notices. I’m not sure how easy it would be to add, but it doesn’t need to be anything fancy.
Re. Traffic coming down the M11:
The junction between the M25 and the A13 (close to the Dartford Crossing’s northern approaches) is having some major surgery done on it, which is bound to affect traffic flows temporarily. However, there are a number of other reasons why traffic may be going into London rather than around it…
Firstly, Stansted Airport is often quicker to get to from the south-east of London than Heathrow and Luton. (Gatwick is the other choice, but it can be a bit of a pain if you’re driving as the M23 is shared with traffic from Brighton and the south coast. Also, some carriers and / or destinations are only served by one or the other, not both.) This means a lot of that traffic might not be going into central London, but merely passing through it, via the Blackwall Tunnel. Nobody who lives in Penge or Norwood is going to choose to drive to that airport via the M25.
Stansted Airport also handles freight as well as passengers. Quite a lot of it. Inevitably, some of that will be heading into south London.
The M11 is a very convenient route to/from the Blackwall Tunnel – a popular alternative to the tolls at Dartford. The south-eastern quadrant of the M25 is also clearly optimised for Kent’s port traffic; if you want to get to Bromley or Croydon from the north-east of London, the Blackwall Tunnel is usually quicker than going round the houses via Dartford, even during the peaks.
Since we’re on the subject of locality and changing how zones work, does anybody know if any politicians have suggested using the Danish travel zone system?
And finally…
I’m curious to know whether the Greens have ever suggested requiring new-build apartment buildings to eschew conventional underground parking facilities and maintain a small pool of shared vehicles instead. (Some space for private vehicles would still be needed, such as for on-call doctors who would need their own car, but you wouldn’t need more than a few of these.)
What follows seems to me like something they should have already thought of, but I can find no such suggestion or policy on their website…
[Further elaboration of a scheme which the Green party did not suggest snipped. Malcolm]
This proposal also addresses the criticism of adopting the Über model: the cars are driven by those who need them, so there’s no need to have cars endlessly circling each district like vultures. Once self-driving vehicles become mainstream, you also have a ready-made set of mini-depots where they can be based, charged and, in some cases, even maintained.
Given the densification push, the above seems almost a no-brainer as it would be trivial to implement. Even without the Greens, I can see many developers seeing this as a key differentiator for their projects.
tyteen4a03: Even if there is something about the Danish zone system that would make it particularly applicable to London, I doubt if it would be suggested in that way. There is just too much resistance (among some people and newspapers) against anything “from Europe”.
But if there is some particular way that you think the system in Denmark is better than Oyster zones, do mention it.
One apparent feature of the Danish system is that it applies (if I understand it correctly) to the whole country. That aspect, however attractive, will clearly not fit with the fragmented nature of Britain’s transport providers.
@ Malcolm – I understand the point you make about a nationwide fares system. However there are a couple of initiatives which may start to construct parts of a possible scheme. I understand there is work underway on a Scottish travel smartcard. Transport for the North has aspirations to pull together all the disparate PTE smartcard schemes *and* have some form of zonal smart scheme for all of the rail services within its remit. Finally the major bus groups are also signed up to a “nationwide” smart ticketing scheme. Now I’m not saying that’s going to create a magic panacea but things might merge together especially if there are some attractive products for passengers in the initial areas that indicate a wider offer would be commercially beneficial. I’d rather see a different approach but it’s clear the DfT are never going to put enough oomph behind a “smart clearing house” concept. They couldn’t deliver SEFT so that’s a counter argument to my observations above. London, of course, would remain very difficult to deal with given the different regulatory structure and shift in the technology London uses even though ITSO cards can be read on London ticket gates and bus readers [1].
[1] no comments about ENCTS ITSO card acceptance please. Irrelevant.
[ENCTS is English National Concessionary Travel Scheme. SEFT is South East Flexible Ticketing. Malcolm]
@ Anomnibus – I am deeply sceptical that “driverless” vehicles are ever going to become a viable, safe and attractive form of transport. I appreciate a lot of “tech brain power” is going into developing the concept but I can see the whole thing foundering due to consumer resistance. I also don’t see anything that proposes using potentially competing fleets of “robots on wheels” on London’s roads as being something that eases congestion. I can’t believe anyone would grant a provider monopoly rights so that means competing providers (and technologies). I am not aware anyone is proposing an “industry standard” which means a repeat of competing technologies battling for market supremacy. This immediately gets us to a parallel with the issues Ms Berry raises about white vans, Uber and freight deliveries. These are all unregulated provider models and look at the fall out everyone is suffering even if people do enjoy having parcels delivered to their front door or a “lift home” at the end of an app transaction.
The Travel in London report clearly shows car usage in decline across London for the last decade. It’s too early to say if we’ve had a reversal in the last year or not but the report does warn that we could return to a grow in car usage if certain factors reverse from the past trend (rising incomes being a key issue). I really do not understand why you appear to be so pro car / pro road in the context of arguing about *London’s* transport network. There is massive scope to do something other than provide for cars but a lot of work is needed to persuade people that alternatives are viable and potentially useful for them on an individual level. That’s a key to breaking down resistance – especially in Outer London.
@ Anomnibus – it is quite common at the moment for new developments to have parking spaces allocated to car sharing clubs like Zipcar.
I don’t think there are any specific requirements for this but it forms part of the developers’ undertakings to meet council planning requirements for sustainability
WW
Wandering slightly off-topic, re “Driverless Vehicles” …
AIUI, the Google (etc) trials in the US have all been in “built-up areas” or equivalent, where the street/roadscape is almost-constantly monitored.
It is known that translating this to “open / country / rural” settings is (at least) a whole order of magnitude more difficult.
However, coming back to nearly the actual topic (!) it does look as though truly “driverless” vehicles, especially small(ish) personal transport will soon become a viable prospect for cities. A lot of these would, presumably for hire, rather than owned, which would reduce the number of “cars on the streets”.
Whether it would reduce the number of moving vehicles on the road at any time is another question.
All of which really will change the “game” & is highly relevant to this discussion.
@Anomnibus
Car free housing is now quite common in inner London where residential property owners agree not to own a car, no parking spaces are built as aprt of the development and the residents are not eligible for a resident’s parking permit. This has been discussed on another thread quite recently (I forget which). The fact that these are commonplace is probably why Sian Berry didn’t think to mention it as a proposal.
One reason for some moderator reluctance about discussion of driverless cars is the considerable uncertainty about how soon (and whether) they will become widespread. So comments saying “I think it will be next year/2020/never” are not required. However, useful extra thoughts about what the implications and ramifications of their adoption might be (such as have been recently posted) are OK.
@Malcolm 23.04.16 19.17:
“One statement of Sian’s that I find difficult to believe is:
“I mean, I talk to people in outer London and one of the reasons they want to keep their cars is because of the cost of going into town and because it’s £8, £12 to go in and out again.”
Has anyone here met such a person? A car owner in outer London who uses their car to go in to town.
I think there’s an extra subtlety in there: People who live in outer London, and for whom it is expensive to catch the tube into inner London, want to keep their cars because they need them to drive around those outer London locations to access the services, jobs etc that central London and local neighbourhoods provide for those living within zones 1-3. The low-density areas beyond zone 3 have been shaped by, and are currently still dependent upon, the private vehicle – unfortunately. Housing, jobs, services, public transport provision etc are not of the density to support car-free living in the same way that most areas within zone 3 are.
Marckee: Interesting point. So for example someone living in Bush Hill Park can only conveniently get by public transport to inner London, or places on the way there. So a chiropractor, say, might only be available by public transport by paying the £8 or £12. Whereas with a car they can drive to Oakleigh Park, say (an easy drive but awkward on buses). Is that the sort of thing you suppose Sian meant? (Other alternative health practicioners are available, and sorry to any chiropractors practising in Bush Hill Park).
I have never understood how driverless cars are supposed to reduce traffic congestion. I can see that they may reduce parking congestion if a shared-use model becomes prevalent (which in itself is a big assumption), but just replacing a driven car with a driverless car does not reduce the number of cars on the street. I saw a comment recently that the current generation of driverless vehicles actually cause *more* congestion as they drive more conservatively than driven vehicles – the Google test vehicles often have a queue of traffic behind them.
I wouldn’t dream of using my car to go into or through central London on a weekday, but on Sundays, with only 2tph (when there are any at all), we use the car if we have occasion to go into London or visit our relatives on The Other Side. Even a relatively simple trip into Victoria last Saturday proved to be problematic.
@Jim Cobb
“but just replacing a driven car with a driverless car does not reduce the number of cars on the street”
It depends on how it’s used. Many car journeys are for transporting people who cannot drive themselves, due to infirmity or youth. Unless the driver of that car has nothing better to do than hang around until their passenger is ready to go back, there will be some “dead” mileage involved. (e.g Mumsy going home after taking little Timmy to school, and going back later to fetch him) If you trust the car to do the school/hospital/station run on its own, it can park itself somewhere nearby out of the way until it’s time to go back to fetch the passenger.
However, if the existence of driverless cars is used as a reason not to provide parking at places which currently do so, such as workplaces, hospitals, stations, theme parks: the roads will be full of driverless cars being sent home after taking their owners to work, etc, and coming back later to fetch them.
A pool or hire system is unlikely to be able to match peaks in demand – look at the distribution problems the Boris Bike scheme has – and unlike bicycles, you cannot easily put a couple of dozen cars on the back of a Transit to redistribute them.
There will be a stampede at the school gates for any driverless hire car turning up at the school gates at 3:45:05
@Malcolm
Yes, kind of. Compare and contrast the density of the outer London suburbs with that of the inner London suburbs. The former is much lower than the latter, and that doesn’t just refer to housing or population, it also includes:
Public transport provision (especially for orbital routes)
Public services
Shops and amenities
Jobs and employment (places of work and potential markets for work)
The reasons for this are partly market-driven (lower value land = lower densities) and also historic. The inter-war and post-war suburbs might have been centred around the new tube stations, but they were designed and built around the automobile. If you can offer people space and a garden and assume that everyone is going to drive everywhere you create road-oriented, low-density neighbourhoods.
What this means is that if the cost of travelling into central London is prohibitively expensive for those in outer London, those in outer London need access to a comparatively larger catchment area in order to connect to the equivalent facilities, and it becomes much more convenient/attractive/necessary to own a car.
@marckee – 25 April 2016 at 11:12
The inter-war and post-war suburbs might have been centred around the new tube stations, but they were designed and built around the automobile
Bit sweeping? I can think of many where the roads are too narrow to allow on-street parking on one or both sides. Some of these have been resolved by concreting over front gardens to provide hard-standing.
I think there was a tacit assumption until the end of the 1950s that the cost of motoring was beyond the reach of the ordinary man.
The streets may be too narrow for on street parking, but many inter-war estates included provision for lock-up garages, often accessed from the rear of the house.
Example here
https://goo.gl/maps/DLpex7tpWSz
Unfortunately both the garages and the access lanes were designed when a typical car was an Austin 7. Furthermore, in the 1920s cars needed more cossetting – they were not the hermetically sealed boxes we have now, so the seats would be cold and damp in the morning – and unlikely to start – if they had been out in the cold all night. They were also much easier to break into or steal. Moreover, people are lazy, and prefer to park outside their front door rather than negotiate a (possibly muddy) garden path.
Indeed. I think current advice for avoidance of the dreaded tin-worm is not to put a wet car in a garage (even if you have one). And sufficiently thorough drying (including underneath) is vanishingly rare.
Jim Cobb says “the Google test vehicles often have a queue of traffic behind them.”
I am tempted, with my grumpy hat on, to observe that any car driving at the legal speed limit and no faster will often have a queue of traffic behind it.
Re Malcolm,
The google cars are limited to 25mph (or lower if the local speed limit is lower) not the legal limit in most places hence the queues. One “test driver” even got pulled over by the police for driving too slowly before he explained the 25mph vehicle limit to them…
[California is mix of 25 or 30mph urban speed limits]
My personal bugbear is Technophiles panglossian enthusiasts. Just a few years ago it was Personal Rapid Transit that was going to sweep away all that old fashioned bus and train systems. Then Google car was debuted and they went wild. If it can work, then, great the carnage from road accidents will be eliminated and it will be worth it from that perspective.
Anything else is being oversold and other effects will all depend on the costs.
The idea that it will kill cities with people commuting ever further from low density exurbs, ignores the cost of fuel and the fact they will go no faster than current vehicles.
This also kills the idea that people will sending their autocars on lots of empty journeys to avoid parking charges does not realise they are talking about doubling someones weekly fuel bill. Driving a car is not free.
Others who talk about just parking in some side street a mile out from a town centre to avoid parking charges, seems to forget the competition for existing roadside parking in many residential areas. As some one who had their tyres slashed for parking in front of someones house you realise how many people feel they own the kerb outside their own home, even though it’s a public street. If such parking took off, all that would happen is resident permit parking would spread to every street to stop it. So no reduction in parking requirements unless there is a massive rise in Auto taxis, but that brings it’s own problems.
The key factor will be how much a taxi drivers wages are as percentage of running costs of a taxi. In rural and the small towns of the Home counties where bus travel is very expensive and used mainly by pensioners and students, then if costs fall far enough then auto cabs could take over public transport duties, but in larger towns and cities then I expect buses to continue to prevail ( at least till 8pm or so).
In London we can already see the effect of cheaper cabs are having on traffic, if the price was to halve again or more, the streets would be jammed all day and night. I just don’t see it happening, Or to be exact such an eventuality would lead to policy responses such as city wide road pricing. It is not as if the bus industry would not respond, with either a completely automated bus and no humans on board or a ticket seller/inspector on board who is paid much less than a bus driver.
@ ngh – I think that proves Malcolm’s point! As part of a campaign by residents on my London street I’ve been measuring vehicle speeds and over 80% of vehicles exceed the 20mph limit, many by a considerable margin.
If cars stay below the limit, at peak times the cars just bunch up and crawl along.
@ Marckee – I think I’m going to go on a rampage if I keep seeing people referring to poor orbital transport services in Outer London. If we look at buses then almost all services in places like Romford, Bromley, Croydon, Kingston etc are either very local or “orbital” in the sense of linking to neighbouring centres like Barking, Dagenham, Downham, Orpington, Streatham, Elmers End, Crystal Palace, Twickenham, Richmond, New Malden etc etc. Croydon is blessed with Tramlink that goes nowhere near Central London but does lots to provide E-W transport links. The two main N-S buses (109 and 468) only get to Zone 2 (OK the boundary with Z1 for the 468). Rail provides the “heavy haul” into the centre from all of the above main centres.
We really must get away from this farcical notion that there are “seemingly” no orbital transport links in Outer London. This is a political construct which is just utterly false and Tories, Labour and Lib Dems in London government have all indulged this myth. We even have these nonsenses being churned over again for this Mayoral Election. The answer is not necessarily to recreate the X26 bus all over Outer London but there may be a small selection of links where an express bus could build the market and take people out of their cars. [The bus “crayons” are locked away]
There is a different debate to be had about how you add to and improve capacity and links in Outer London. I would agree with you that some areas are very car orientated and not able to economically sustain high frequency bus services without a lot of subsidy but there may be more nuanced ways of getting some level of increased public transport usage. I’m thinking here of places like Orpington, parts of Bexley etc. There are also the “unspoken” (by the politicians) issues about what you do with public transport when the destinations are “over the border” outside of Greater London and under a different funding and regulatory regime. Ms Berry has made a brief remark about recognising the need for more buses in Outer London but hasn’t said what that means. She has also cleverly (from her policy viewpoint) suggested more smaller scale cycling interventions rather than the current larger scale Mini (!) Hollands. Of course what she wants to do is generate positive reactions to these small schemes so that pressure builds for more schemes and, crucially, to then link them together. I remember Darren Johnson, Green Party AM, saying that he would simply have a policy of extending Croydon Tramlink until it covered all of London! It’s the same idea for cycling – start small and local and then push and push in increments until you achieve a larger (unstated) goal of London wide provision.
On the issue of driving speeds, it is worth mentioning that “driving too slowly” is not an offence. It can be a cause of failing a driving test, it may also be a cause of police investigation (was the person driving slowly because they were drunk, have poor eyesight, having an argument, not concentrating etc, each of which could be an offence). But if google car, or any other driver for good reason chooses to drive at 25 mph, that is legal. (And things like “there seem to be a lot of children about” could be a good reason).
Except, of course, where the speed limit is 20 mph, which is increasingly the case in many parts of London and elsewhere.
My original grumpy remark was thinking of the person who is fully aware of the speed limit, and drives about 5 mph faster (because their time is “precious”). My anecdotal observations see enormous numbers of these.
@WW: I specifically didn’t use the word ‘poor’ and wasn’t only talking about bus service provision.
I was talking about all public transport (so including rail and tube); I was talking about density of routes; and I was also comparing the areas outside of zone 3 with those areas within it.
I don’t think anyone could argue that the density of public transport provision in zones 4, 5, 6+ is anywhere near that of zones 1, 2 & 3.
WW: I think the “poor orbital bus service” meme might arise from completely unrealistic expectations. If a journey is short, then the difference in time taken between bus and private car is often only a few minutes, and when you’ve added in parking time, bus can be quicker. But for longer journeys (say, over 8 km) this is less likely to be the case, and if you want to compete with car speeds (outside central London), you probably need a train. Orbital train services are quite rare in outer London. So the “poor service” perception arises just because typical outer London orbital journeys required (like my chiropractor) are generally over 8 km, sometimes quite a lot over.
I find it difficult to believe that cars travelling at or just below the speed limit make any real difference to road capacity in peak times. Real modelling of traffic would give us the answers, otherwise we are just airing guesswork and prejudice ( exactly as I have done at the start of this post).
Similarly, the impact of ‘auto-cabs’ (nice name). Their potential impact is so wrapped up in human behaviour, social pressures and regulatory restrictions, that we can all speculate magnificently and that’s about all.
Road capacity for transport is the core issue. Everything that has been discussed already illustrates how most debates cover specific little corners of the problem without ever trying to answer the question ‘how can we make overall best use of our road space?’
@Reynolds 953 – If cars stay below the limit, at peak times the cars just bunch up and crawl along.
Precisely what happens locally since the council narrowed the lanes, installed speed humps etc. The jams are incredible now. Previously you only got jams when there was a crash on the A406 and the traffic diverted. It’s such a backwards step given the road functioned perfectly well before. I’m struggling to identify what “gain” there has been and to whom it has accrued [1].
[1] possibly the highway contractors and aggregate and paving suppliers. 🙂
@Walthamstow Writer – I’ve had discussions with council traffic planners about traffic problems on my street and they told me, in general, reducing vehicle speeds does not reduce traffic volume. So if a street is seen as an attractive rat run for speeding vehicles because it avoids a junction with traffic lights, slowing the traffic down does not reduce the volume. Drivers prefer to keep moving, even slowly, compared to a route with traffic lights.
So traffic calming measures may just substitute one type of traffic issue with another – instead of speeding vehicles the traffic is backed up nose to tail.
Again in my particular case, I know there is work being done looking at options for traffic filtering in the area as part of a cycle superhighway and quietway scheme. This would block through traffic so streets in the area would be for access only.
I’m anticipating the usual pitched battles when plans come out… 😉
WW
Precisely, And, of course said local council are vigorously denying that anything of the sort is going on at that all of their traffic measures have 99% or even 100% public support.
R 953
“Rat Runs” – yes, that excuse was used here. Unfortunately about 95% of the “rat-runners” were local people diverting off the main roads, when they got into the “large block” in which they lived. As above, the council deny everything.
It is very clear that the main capacity constraint on any road in built up London is at junctions, especially signal controlled junctions. Even if the stop line has twice as many lanes as the road leading into it, the capacity across the stop line is less because the light is going to be green for less than half the time. Once that point is accepted, almost anything that happens on the links, whether it is slow moving vehicles, pedestrian crossings or bus stops, will have no overall effect on capacity. Instead of drivers going faster and waiting longer at the stop lines, they will move slower but wait less at the stop lines. For the same reason bus lanes, provided that the reservoir behind the stop line just clears in one green phase, have no impact on overall capacity.
Hence, WW and GT, you may well be appreciating more congestion at the moment, but whatever the cause, it’s not going to be traffic calming.
Traffic is a complex phenomena and not all causes are obvious. Some years ago I had a debate with the stall holders in a street market. The council had put in traffic calming and other measures to reduce and slow traffic going through the street market (there had previously been too many unfortunate accidents involving both traders and shoppers). But, shortly afterwards trade declined alarmingly. The traders all blamed the traffic calming measures but completely ignored the fact that a large local estate had just been emptied for redevelopment. Two things involving traffic can change at about the same time but it doesn’t necessarily mean that the one has caused the other.
A month or so back I was at a number of presentations / discussions about city growth and how to deal with it. One of those was by a rep from Camden Council on how they have gone into the consolidation business. All of their supplies are now delivered to warehousing out in the Lea valley, and then brought into the centre of London by their own vehicles, with the latter full to the brim. He also mentioned that they were inviting staff to use the facility for their online shopping (instead of getting it delivered to the Town Hall, for example)
This idea of a flat fare, to me, seems a non-started in that whilst it would create a flat fare *for journeys to (the present) zone 1* it would make all other trips overpriced.
@Rational Plan:
“Just a few years ago it was Personal Rapid Transit that was going to sweep away all that old fashioned bus and train systems. Then Google car was debuted and they went wild. “
PRT and self-driving cars are fundamentally the same thing. Hence the going “wild” over Google, et al. We know PRT works, because Heathrow’s T5 has its ULTra system, and it’s been working very reliably. There are other examples around the world, but these are all basically self-driving vehicles that use cheap tricks to help them steer: they’re essentially running on rails, like a tram, except the rails are partly virtual rather than purely physical.
Segregated PRT might work well as an alternative to guided buses, but the preferred goal is to create a PRT technology that doesn’t require its own dedicated infrastructure, which can often be difficult to insert into the existing public realm.
[Started ok but drifted off topic. Also please keep in mind that this is not the place to debate PRT or guided busway systems in detail. LBM]
@quinlet – some good points, hence why I prefer to trust in data rather than anecdotes when it comes to traffic.
There has been a road closure to stop rat running near to me in West London and quite a lot of local controversy between the residents on the street concerned who are pro closure and other locals who are anti.
Following the closure, the antis blame every traffic jam on the closure but TfL bus route timings on the roads affected have shown no change whatsoever from before the closure.
People have forgotten that before the closure, traffic was pretty heavy with occasional jams and after the closure, traffic is still pretty heavy with occasional jams.
Additionally, there can be other effects such as Braess’s paradox where road closures can actually reduce congestion. Where there is a choice of route, drivers make their own selfish decision but that doesn’t mean their decision leads to optimal performance of the network as a whole.’
@ Marckee – I wasn’t arguing Outer London public transport was the same density as closer to Z1. Clearly it isn’t. I am sure you could legitimately argue that it doesn’t need to be at that level either. Nonetheless I find political remarks (and to be clear, not yours!) that “diss” the level of transport provision when set against an undefined “panacea” as unhelpful to say the least. Taking a more pointed stance it’s a deceit and sets out to create an incorrect perception on the part of the public which then partly justifies woolly and ill defined “promises” and policy initiatives (Outer London Commission anyone?).
@ Quinlet – I fully appreciate the point you make but I am struggling to find an alternative but simultaneous issue that could explain what I’ve noticed with our local traffic recently. I’m also referring to a main road not side roads. I won’t drag the debate into detail as that will help no one here but the main local development has been an awful lot of highway engineering and road works with slow speeds the result. We have many more months of this on a local basis and the works nearest to me haven’t yet been made public. There is quite a lot of new housing going up in Waltham Forest but much of it is not yet finished so the traffic / public transport impact is yet to be seen. I’m not hopeful given how overcrowded existing peak time services are and it’s not yet clear whether TfL have reviewed the local network (a lot of local bus routes are out to tender over the next year).
@WW Pretty much all roadworks produce congestion when they are being undertaken, and any ‘road diet’ produces short-term congestion, but just about every study also shows that in the long-term, if coupled with increasing cycle and pedestrian provision, congestion is reduced and safety improved. It wouldn’t work if driving didn’t become harder to the point that unnecessary journeys are unattractive.
In Waltham Forest (I’m a resident of the borough too), much of this congestion can be laid at the feet of the works being ongoing along some of the major routes through the borough, and one would expect this to ease off around six months after they are complete, when people (and satnavs) learn where the modal filters are located.
In my side of the borough I’m more concerned that the blockade of the GOBLIN coincides with 20 weeks of gas main-induced closures on Blake Hall Road – it’s inevitable that Mini-Holland will be blamed for the resulting queues that will build up in the south east of the borough despite minimal work being undertaken there during this period.
@Malcolm 14:53
Driving too slowly can be deemed to be “careless driving” and fines implemented. An actual example is reported on the BBC news website (search for “Pensioner fined for slow driving”) while the AA website lists “inappropriate speed” as potential grounds for an offence.
Man of Kent: Yes. I should have said that there is no official offence of driving too slowly. But presumably people who have been penalised in this way were not able to provide the “good reason” to which I referred. (Children, ice, car fault, visibility, whatever).
And this one
http://news.bbc.co.uk/1/hi/england/bristol/7171154.stm
Vehicles incapable of achieving 30mph are not permitted on motorways. But there have been several incidents of mobility scooters on the A27 expressway – even the eight lane section!
http://news.bbc.co.uk/1/hi/england/hampshire/7544509.stm
Perfectly legal, but extremely foolhardy.
On a slightly different topic, I see that almost all the major party candidates (Berry, Goldsmith, Khan, Pidgeon) have now signed up for the London Cycling Campaign’s pledges.
The same thing happened in 2012 so it is interesting this has become tablestakes for a serious candidate rather than the preserve of the Greens.
R 953
That reminds me … concerning a n other form of transport/movement we all participate in … walking
Are any candidates actually supporting continuing with the “Garden Bridge” ?
Greg: unfortunately this is not a hustings meeting at which such questions can be put. Others: please no garden-bridge related follow-ups unless you are saying something genuinely new about this unpopular idea.
@Jim Cobb
“I have never understood how driverless cars are supposed to reduce traffic congestion. “
Density. For a given safety risk tolerance, driverless cars are expected to be able to travel much closer together and/or at much higher speeds than manually driven cars. There should be a small improvement as driverless cars start to replace manually driven ones, but the really substantial increase would come on driverless-only roads, because they will be able to reduce distances accordingly. As well as allowing more cars to use the same lanes (and even more passenger kms, if speeds increase), more lanes may become available, too (two lanes may become three, for eg). Alternatively, the extra space could be used for cycle paths or wider pavements. But a boon one way or another, irrespective of how it’s used.
@Walthamstow Writer
I am deeply sceptical that “driverless” vehicles are ever going to become a viable, safe and attractive form of transport. I appreciate a lot of “tech brain power” is going into developing the concept but I can see the whole thing foundering due to consumer resistance.
When minicab firms no longer need their drivers because the cost of driverless tech has fallen sufficiently (given the billions being invested now, quite likely fairly soon), and it therefore become substantially cheaper to run a car driverlessly, do you think consumers will resist the much cheaper fares on offer from driverless minicabs? If not, and some are prepared to do the novel, modern thing in exchange for a much cheaper fare, how long until resistant slow-adopters start feeling silly for paying more, perhaps double (if drivers’ salaries represent half the running costs of a car used 24/7 by a minicab firm – this is pure speculation, I have no idea how much it actually is) what others are paying?
Aren’t most minicabs operating in an owner driver environment, as with black cabs? Why would the companies they use (effectively booking agents) ever want to take on the capital costs?
Re Kit,
“Aren’t most minicabs operating in an owner driver environment, as with black cabs? Why would the companies they use (effectively booking agents) ever want to take on the capital costs?”
They are but they would be replaced by new companies such as Google or Apple (who both happen to have large cash piles!) or operator (e.g. uber) and leasing company combinations (there are large number of vehicle leasing companies)
@Harry Crayola
I really don’t believe there will be much increase in traffic density in stop and go traffic that is typical of most of London’s roads, save early early mornings or late evenings. I also can’t see two lanes becoming three being practical at all given that there is a need for space for the swing of larger vehicles such as buses and lorries going around corners. Plus the need to have extra space on the curb side to open a door to pick up or let off a passenger, without smashing the door into parked cars or street furniture.
@KitGreen
Yes – that’s correct. But not especially relevant, for the reasons ngh explains. The ownership/provision structure is much less important than the costs underlying them. Whether individual drivers upgrade to a driverless car and spend their freed-up time doing something else, or whether they’re replaced by directly-owned or leased cars as per ngh’s suggestion (much more likely of the two, I’d guess), is not a substantial matter in contrast to the fact that you’re stripping out what is, I’d guess, the single biggest cost and perhaps even over half of it.
@Long Branch Mike
Good point re stop-start. But I suspect there are efficiencies even here. I’d be very surprised if driverless cars couldn’t wait more tightly-packed at traffic lights, and if more couldn’t flow through them during the alloted time for a green light. So I suspect there are non-trivial capacity improvements to be made here, too, although less substantial than those available on smoother-flowing routes.
“the need to have extra space on the curb side to open a door to pick up or let off a passenger, without smashing the door into parked cars or street furniture”
I’m afraid I don’t understand how this is relevant, however. Could you possibly flesh out a situation of how the more tightly packed driverless cars (or perhaps the narrower lanes) I described might relate to what you said?
@Harry Crayola
In making two lanes into three, the full width of the two lanes would be used and filled with the width of three vehicles, and that would mean right to the curb/kerb. So where there is street parking, lampposts, signposts, bus shelters &c there is no wiggle room for opening the door on the curb side. Plus sloppy (human) driving could not be accommodated in such 2 to 3 lane conversions. It appears to me that there would be no margin of error in the width of the road, unless the original lanes themselves are wider than typical and can be reduced.
Re Harry Crayola,
Unfortunately I can only see the bigger benefits coming from some kind of central control system or the cars having access to huge databases with the latest set of traffic light phasing timings for each set of lights and even then there is unpredictability for example with a left or straight ahead lane at lights is the (human driver car) car 2 in front turning left or going straight ahead as this would make a difference to what the driverless car would do, what if the car 2 in front indicates late etc.
Would the different makes of driverless cars need to communicate with each other to maximise benefits?
you would only need the two lanes converted to three in the approach to traffic signals. For the rest it would not add anything. It’s also possible that the autonomous taxis could be programmed as to where and where not to stop, avoiding hazardous places.
@WW
Driverless cabs could also provide greater coverage (because of the lack of need to provide a driver) and thus increase availability. For passengers, therefore, there might be a choice of a driverless cab or no cab at all. I know which one I would opt for.
Efficiencies through tighter packed cars will be cancelled out if driverless cars mean more cars moving around with just one (or no…) person in them.
Efficiency in terms of people per square metre of road space used by a driverless car still won’t come close to buses or bikes.
Re R953,
Indeed in the example far above the driverless car doesn’t just need to pick up Timmy from school but Tom, Dick and Harry as well…
Re Quinlet,
“You would only need the two lanes converted to three in the approach to traffic signals.”
Err – exactly where the pre junction bicycle lanes start???
The performance of many junctions could be improved with a large quantity of red line paint in the vicinity, cheap but some of the shopkeepers etc. wouldn’t be happy!
Man of Kent,
For the sake of accuracy, driving too slowly would much more likely to be classified as “inconsiderate driving”. This is typically invoked for things like not avoiding puddles and consequently splashing pedestrians or driving slowly with a heavy load and failing to pull in when opportunities arise at lay-bys to let the following traffic pass. The difference between inconsiderate driving and careless driving is that in the the former there is not an element of creating a potentially dangerous situation. If both offences appear to have been committed then the charge will almost certainly be careless driving in preference to inconsiderate driving though the person could potentially be charged with both. Obviously specific traffic offences (e.g. blocking a box junction) are more preferable still when it comes to charging as it is hard for the person not to be found guilty if the facts are proved.
Harry Crayola, Walthamstow Writer,
For the reasons stated driverless cars are unlikely to reduce traffic congestion by much if they simply replicate the route the human driver would have taken. However it is the start of an opportunity to re-route vehicles automatically. So if a particular junction or section of the road is congested it is possible for the autonomous vehicle to be rerouted and the decision making process could potentially take into account the benefits to all users (or at least all road users). One could also prevent the situation, often seen, where one selfish or incompetent driver manages to block a set of lights for a whole sequence in a generally pointless desire to get a bit further forward.
Reynolds 953. Good point earlier about bus timings being a source of fairly useful data for deciding whether road layout changes are really causing delays (or not). Your later point supports what I wrote earlier. Measuring effective road use in terms of vehicles moving around is entirely missing the point. Those vehicles are a means to an end; either shifting people or shifting goods. Just counting vehicles is absurd. Roads are barriers to non-motorised traffic. If phalanxes of self-driving vehicles are whizzing along in close formation then no mere human is going to be able to get across any road.
When I took my driving test oh-so-many-years ago I was warned by my tutor that the examiner I was due to have took a firm view of ‘proper speeds’ when driving, and would fail people on the grounds of “failure to make normal progress”.
So far as the benefit of “tightly packed driverless cars” at traffic lights just watch what happens presently: the first car moves off, the second has to wait a few moments before moving off because that driver needs to be certain (a) where the first car is going, and (b) that they won’t slam on the brakes. The third car repeats this delay ad infinitum. With inter-communication autonomous automobiles they can all start moving *simultaneously* because there are no drivers involved.
Two relevant questions about driverless cars , in particular, seem not to have ready answers:
– to what extent is the software risk averse. If the programming is designed to stop immediately any sort of risk appears (particularly risks from other vehicles) then that will consume more road capacity than manual driving where, rightly or wrongly, many drivers take risks at junctions, roundabouts and traffic lights.
– how the software delivers the vehicle to the precise destination in the absence of any manual override – and I don’t just mean delivery to GU8 5JB, as it were, nor indeed to Waitrose’s car park but to a precise place such as my front drive (with the added choice of my garage also), or if that’s full, to the layby opposite. How do we avoid autocars being stuck looking for the specified parking space, taking up road space awhile?
– Oh and to pick up Fandroid’s point – it’s unclear how long “trains” of selfdriving lorries can be prevented from entering junctions from which there is no way forward, so blocking the roads behind them. (Single manual drivers are already quite good at this sort of blocking manoeuvre, give them a longer vehicle train…. )
@Harry Crayola:
Some of the benefits of computer-driven vehicles (note: NOT just “cars”) are possible only when meat-driven vehicles are no longer sharing the same roads. During the transitional phase, the primary benefit should be safety and more efficient routing, but they’ll still have to interface with human drivers, cyclists, and pedestrians.
There will be some road space benefits, but they’ll be based around more efficient use of the vehicles (i.e. no need for ’empty’ trips back home; the same vehicle can just move onto its next customer), and no need for a driver’s cab in HGVs, buses, and the like. The humble white van becomes little more than a self-driving box. Removing the redundant cab from these vehicles should lead to some reduction in congestion, but it’s unlikely the benefits will be noticeable given the high latent demand.
Electrically powered vehicles will also allow for some size reductions as these don’t require a big lump of hot metal to be kept under a bonnet. But that’s about it.
I’m as much a cheerleader of technology as anyone else, but I’m not blind to its limitations. Trains will still be more efficient at moving large numbers of people.
@ngh
“Would the different makes of driverless cars need to communicate with each other to maximise benefits?”
yes, but those benefits won’t begin to be achieved until all cars can communicate with each other.
“even then there is unpredictability for example …. is the (human driver car) turning left or going straight ahead as this would make a difference to what the driverless car would do,”
This illustrates the more immediate problem – mixing driven and driverless cars. (Even on the Tube, driving a train on a normally driverless line causes some interface issues)
Arthur C Clarke once predicted that there will come a time when it will be an offence to drive a car on a public highway.
(ACC claimed that he refused to travel in any vehicle in which he could not read a book)
@Graham H
“it’s unclear how long “trains” of selfdriving lorries can be prevented from entering junctions from which there is no way forward,”
As long as they are not physically coupled, each lorry will still have to make the decision about whether to enter the junction. Given a long enough amber phase, one lorry in the train can “drop the tow”. The junction controls (which may be lights but need not be) will be communicating with the vehicles and know to change in time for the last vehicle it allows into the junction to clear it before the “train” of which it is a part comes to a stop.
Graham H: As I understand the concepts, driverless cars are not intended to be individually owned, but used on an ‘as required’ basis. This is their USP – that you don’t need to provide parking space for one (or pay for it when it isn’t being used). You would actually end up removing the majority of the need to parking meters / car parks too.
Other jet packs and flying cars are available.
@Long Branch Mike
I agree that when/if we narrow lanes to take advantage of the technology, you wouldn’t be able to open the passenger door kerb-side if there was a parked car or streetlamp or something there. But don’t people normally drive on the extra few feet to avoid these things currently anyway? I know I do. I suppose in such circumstances where people cannot or do not want to drive the extra distance to pass the streetlamp/whatever, then it would represent a loss to the alighting/boarding passenger. In my opinion, however, that seems very trivial.
I also agree that we wouldn’t be able to narrow lanes unless the road was driverless-only and was part of what I meant when I orginally said “the really substantial increase would come on driverless-only roads”.
@ngh
I don’t agree that there would have to be such centralised direction to take advantage of the efficiencies. Making traffic data and junction signaling data available should be enough for each car to effectively plan its route accordingly. I suspect the implication of your question about car-to-car communication is correct: ie, that they would need to do so to be maximally efficient, but I don’t read enough on the subject to say with any certainty.
Reynolds 953
Efficiencies through tighter packed cars will be cancelled out if driverless cars mean more cars moving around with just one (or no…) person in them.
OK. If this is true (and it will be an empirical question), then it will demonstrate that the efficiency benefits of driverless cars have accrued wholly to additional journeys (more cars moving around) or car availability (owners empty summoning cars from home or sending them home to conserve precious/expensive city centre space). This would all still be a good thing, although I do not share the opinion that this is the likely distribution of benefits. FWIW, my own belief is that driverless cars will accelerate the move away from car ownership towards membership of car clubs and/or increased use of minicab firms. The difference in the product offering between Uber and Streetcar, for example, will narrow markedly once both have driverless cars. As car ownership declines and people hire driverless minicabs more instead, I suspect that opposition to vehicle tracking devices will weaken, which will enable much more sophisticated road pricing systems, which will itself further advantage hire rather than ownership (why pay for a space to hold your vehicle, or pay for fuel and road charges to send it home to park in your garage when you can only pay for the journeys themselves?)
“Efficiency in terms of people per square metre of road space used by a driverless car still won’t come close to buses or bikes.”
It depends on your definition of close, I suppose, and how valuable that difference is, as to whether or not it’s relevant. Buses will likely go driverless, too, of course (probably well before tubes!) So they will become more efficient, as well. But without knowing the facts, my guess is that their efficiency potential is much less than that for minicabs because I suspect that driver employment costs are a much smaller percentage of total costs for buses than they are for minicabs. So, for that reason and for the road capacity reason there is likely to be a shift in demand from bus (and rail/tube) travel towards driverless minicab travel as a result.
@Graham H
All interesting questions. Another similar one is whether the car will be programmed to save the life of its passenger at the risk of multiple pedestrians or vice versa, in cicumstances where that is the only choice.
The big elephant in the room with driverless cars is responsibility when the car is involved in an accident – each decision has to be coded by the software designer, including deciding between saving the driver and saving a pedestrian. How will the law accept those sorts of decisions ? Whilst the development of driverless cars is happening at technology speed, changes to the law far, far longer. I wonder if we will see the metaphorical may with a red flag walking in front of them to clear the way ?
I see various suggestions on how driverless cars will improve capacity. These are based on a shared-use model, which is all very egalitarian, but will it really happen that way ? Taxi’s maybe, but will people really give up their personal vehicles ? On top of that, what happens during the transition. It is not as if all vehicles will become driverless overnight – there will be many years where there will be a mix of vehicles, so you cannot start changing lane sizes whilst you still have cars on the roads. If people believe the promise and they become popular quickly, they could easily cause more congestion until the road infrastructure is adjusted to make best use of them.
Personally, my belief it will happen, but the change not be quick or simple, and there will be many unintended side effects, because there always is with disruptive technology. We are struggling to cope with implications of Uber, and that is relatively simple change in comparison. You will know driverless cars have arrived when you see the campaign to ban them !!
@AlisonW -yes,I’d assumed that the “GU8 5JB” problem would disappear* if you had *only* communal/commercially pooled vehicles although whether that is likely/acceptable remains to be seen (ask any taxi driver how often he has to clean his vehicle on a Saturday night…). Jim Cobb is right to raise the cultural issues – and if they exist then all the sorts of problems mentioned above arise.
Harry Crayola is surely right about product liability; there is a great deal of evidence from US class suits to show the corporate damage that can occur in this area. The surprising thing is that Google seem to carry on so blithely. That suggests to me that their software is incredibly risk averse, with allthe implications of that for traffic management. (I think it was Alan Griffiths who remarked on how little lawyers change the world?)
*and how often have you tapped in a postcode to your satnav only to be conducted to a field in the middle of the countryside? “Your personal car is now only half a parish away…enjoy”
“will people really give up their personal vehicles?” – as mentioned in the interview, there are already housing developments where that is a *requirement*. If it wasn’t for the fact that the nearest car club spaces are too distant from me I’d be seriously considering it (and I have friends who have already made that jump.)
So far as postcode destinations are concerned, every UK property (actually every *entrance*) is fully coded by a number of organisations so a far better degree. Postcodes have been being misused since they were first developed.
@Alison W -even “every entrance” isn’t good enough. It has to be every possible parking space otherwise you will be left somewhere that isn’t what you need – eg your drive rather than your garage/the space by the front entrance rather than the space round the back accessed from the front entrance. Furthermore, the auto system needs to know that that space is actually free. Short of satellite and CCTV surveillance of every possible space,how is that to be achieved?
@Fandroid, Graham H
Platooning of HGVs has been touted regularly as an early win for autonomous vehicles because:
(a) the technology already exists; and
(b) the cost of installation and operation is more than self funding through fuel savings
However, the operational problems, such as joining and leaving are likely to present much bigger problems in practice than have been anticipated. It is noteworthy that the UK trial of this approach is scheduled for the M6 in Cumbria which has both one of the longest stretches between motorway junctions and has a motorway service station for the platoons to form up in. This somewhat reduces two operational problems but no-one has yet solved the problems of what happens when a platoon comes alongside an entry ramp with other vehicles wanting to join, unless the platoon is in the middle lane – yet the platoon will still be amongst the slowest vehicles on the motorway as they are limited to 50mph so this will just encourage undertaking. Much work remains to be done before any actual trials will take place, I think.
@Jim Cobb
I suspect the liability problem is overstated as the principle as it applies today is equally valid for driverless the vehicles. That is, the driver is liable unless it can be shown that someone else is liable, either the keeper , for not undertaking foreseeable preventative maintenance, or the constructor if the design is at fault. The company responsible for the software will have a place in that hierarchy.
quinlet: I cannot quite understand the problem you describe of a platoon alongside an entry ramp. It is already possible (in these pre-platoon days) for a car on an entry slip to be confronted with a stream of vehicles in the left hand lane without a sufficient gap to slip into, and in such a case you obviously have to stop. It doesn’t happen all that often, but every driver must already be able to deal with it when it does.
Graham: I accept there is a need for the person in charge of the journey to convey to the self-driving software exactly where it is meant to go, drive or garage or front door, and for suitable negotiations to occur if the stipulated place turns out to be occupied when you get there.
But solving this problem strikes me as child’s play compared to the (apparently fully-solved, or almost so) problem of driving through heavy traffic without carnage.
@quinlet – not sure about “the driver” when there patently isn’t one. How about “directing mind”? And if so, when the directing mind has no control other than to start and stop a vehicle on a journey which may be full of incident, in what sense does the directing mind actually have control?
Allowing the public to use driverless minicabs at market-driven cost will cause such a huge switch from trains and buses to autonomous vehicles that the entire road network will grind to a halt (except in the small hours), and the railway network will become seriously underused (except in the peak). Road tax on all vehicles will have to be massively increased, or the autonomous cabs will have to be drastically to bring their cost to the user up to the current cost.
Jim Cobb ” wonders if we will see the metaphorical man with a red flag..”. We are already seeing him, in the form of the human driver who is currently sitting in the cars with a big red emergency stop button (perhaps metaphorical). Presumably the need for this person will be legislated away in due course, just as happened to the red flag man.
Forget reallocation of road space or tightly packed cars, these vehicles will have to interact with non motorised traffic and for a long time other manually driven vehicles.
Lots or people keep things in their cars either permanently or during long days of multiple stop trips. Plus with auto cabs their will be a lag time between ordering a cab and getting into one. In less dense areas or at peak times you could easily wait a long time compared with your own car.
Remember an autocab is still a capital heavy piece of public transport and the owner of such a vehicle needs to achieve a high utilisation rate to offer competitive fares and maximise profit they can’t afford much slack in the system. With government set fares that means a shortage of taxis at peak times and hence a longer wait time or as in Uber higher prices at peak times, which encourages more peak capacity.
Either even if all of Londons main roads were 6 lane boulevards there would not be enough auto taxis to carry people to work at an affordable price.
Mass transit is here to stay in big cities and as an inter urban form of transit. The hardest area will be bus transport in rural and semi urban areas where traffic congestion is not a serious issue and the most obvious one when pensioners can keep driving and no longer need to rely on a bus, the survival a universal bus pass will be in doubt. Much more likely will be a mobility subsidy either as an extra payment or as a separate card that can be used for auto taxis or buses and will most likely be means tested. What the bus industry does in such an environment is hard to know but a retrenchment to major cities and or the launch of driverless bus services to cut their operating costs.
In a far enough future taxi firms and bus companies could merge and sufficiently powerful computers could aggregate consumers smart phone journey requests into bundles of similar journeys either taking several people in a minibus from the same suburb/business park/shopping centre to the a several nearby destinations.
In bigger cities taxis or minbuses could ferry people to the nearest transit hub where long distance or mass transit trunk lines whisk people to other transit centres near their destination. Your personal digital assistant chirping away guiding you when to get off at a stop and helping you identify the correct vehicle to get into for the next journey stage.
Re Harry Crayola & Timbeau,
The Chinese work on driverless vehicles has shown that the driverless vehicles need to communicate with each other to achieve significant benefits (especially in mixed environments with humans in control of other vehicles or themselves, the benefit isn’t just for segregated driverless vehicles) and they have been working on developing a suitable communication protocol as a key part of their work.
Re Alison W,
Exactly UPRN and the various utility company systems have already done this for long time. RoI as a late adopter of postcodes is rolling out a 1 letterbox = 1 unique postcode system that has very accurate letterbox coordinate data.
And just yesterday Google and Uber partnered to lobby to aim to prevent California from requiring controls and a qualified driver to still be in the vehicle.
Uber’s long term aim is fairly clear then… (and they have obviously looked at the density issue with the attempts at ride sharing).
Everything I’ve seen so far suggests that the google car is actually very conservatively programmed for safety reasons (and now re-programmed to expect bus drivers never to be helpful!)
@Rational Plan
“Forget reallocation of road space or tightly packed cars, these vehicles will have to interact with non motorised traffic and for a long time other manually driven vehicles.”
The really subtantial benefits will have to wait until we have driverless-only roads. But there will be significant benefits before then, too. A driverless car can react much faster to the car in front than a human driver, meaning they don’t need as much stopping distance. This benefit is immediate, but only in proportion to the number of cars which are driverless and the aggregate benefit will therefore increase gradually as driverless cars replace mannually driven ones.
“a lag time between ordering a cab and getting into one”
As there is now. I’ve never had a problem with Uber for this. Surge pricing keeps availability high. The contention isn’t about whether minicabs/car clubs will *entirely* replace owned cars. Instead, it’s that driverless cars will hasten this already movement, by reducing the costs from their current level above cars down closer or perhaps below the journey cost for an owned car.
“Remember an autocab is still a capital heavy piece of public transport and the owner of such a vehicle needs to achieve a high utilisation rate to offer competitive fares and maximise profit they can’t afford much slack in the system.”
I’m sure Uber/Zipcar/etc will be thinking about this.
“With government set fares that means a shortage of taxis at peak times and hence a longer wait time or as in Uber higher prices at peak times, which encourages more peak capacity.”
I think you make a very powerful case against governments setting fares and letting the price mechanism work its magic. I’m convinced!
“there would not be enough auto taxis to carry people to work at an affordable price. Mass transit is here to stay in big cities and as an inter urban form of transit”
I agree that mass transit is here to stay. My contention is that it will allow for more private road transport. The balance will shift. A little in central and inner London (although I expect that its effect in inner London will be outweighed by continuing densification which continue to encourage a shift from private transport to mass transit). But in rural and low-density suburban areas I see very little role for mass transit or even buses. If taxis cost about half their current amount, how many people will continue to get buses or infrequent trains?
The same goes for intercity rail, especially as motorways are likely to be the first roads to be converted to driverless-only use, allowing much higher speeds and capacities than at present. I can’t see it surviving the competition.
“In a far enough future… sufficiently powerful computers could aggregate consumers smart phone journey requests into bundles of similar journeys either taking several people in a minibus from the same suburb/business park/shopping centre to the a several nearby destinations”
Isn’t this basically UberPool?
@PoP: if a particular junction or section of the road is congested it is possible for the autonomous vehicle to be rerouted and the decision making process could potentially take into account the benefits to all users (or at least all road users).
But what incentive is there to any individual car’s occupant to take a longer route in order to reduce congestion overall? Traffic congestion is often a classic ‘tragedy of the commons’ situation where individuals pursuing their own interests create a situation which is worse for everybody. Are we willing as a society to outsource power over the routes we take to the likes of Google, on the basis of a promise that the algorithm will enable the greater good of the greater number?
And will we then start to see pressure on governments to regulate the routing of vehicles, so as to avoid residential areas, areas of high pollution, etc?
@AlisonW: With inter-communication autonomous automobiles they can all start moving *simultaneously* because there are no drivers involved.
Although as their speed increases, you need an increasing gap between each vehicle, just as in a railway moving block signalling system – so they can’t all set off at the same speed – the vehicles behind need to accelerate more slowly than those in front. Even in the highly controlled environment of a metro system the pure ‘platoon’ concept (ie of vehicles at speed separated by less than their braking distance) has never been accepted, because of the catastrophic consequences if the vehicle in front should collide with something.
@Harry Crayola: If taxis cost about half their current amount, how many people will continue to get buses or infrequent trains
Half their current amount is still more than buses or trains, so I’m not sure the economics really work there, especially assuming that full cost recovery through road pricing is introduced, as you seem to favour (which would make most rural and suburban roads very expensive to travel on).
The same goes for intercity rail, especially as motorways are likely to be the first roads to be converted to driverless-only use, allowing much higher speeds and capacities than at present. I can’t see it surviving the competition
Higher maximum speeds mean higher energy and maintenance costs and lower capacity, of course. And motorways rarely reach the centre of cities either.
Techno-utopians have been predicting the imminent demise of the rail network for decades. As a general rule, I would suggest that any argument based on a comparison of a technology that actually exists and is in use, versus a technology that does not exist, tends to be stacked in favour of the non-existent technology because the proponents of the newer technology haven’t yet realised the real-world issues that will prevent it from reaching its theoretical maximum potential.
sufficiently powerful computers could aggregate consumers smart phone journey requests into bundles of similar journeys
Better still, you could predict in advance, based on past demand, where the journeys are likely to be required. Then encourage people to congregate at optimised kerbside locations to be picked up by their vehicle, which would be optimised for efficient shared usage. You could even let people know weeks or months in advance when pickups would occur, to allow them to plan ahead. You could give this highly optimised service a friendly name that emphasised its convenience for everybody. Maybe put it into another language? Does anyone know what the Latin for ‘for everybody’ is?
One group who rely heavily on taxis are those who are physically incapable of driving. For many of this group, a driverless taxi would be useless because their infirmity prevents them getting in and out of the taxi unaided, or of loading and unloading luggage. Others would be unable to see the control panel to set the destination.
How would these people be provided for?
Regarding platooning of HGVs…….based on my last week day trip on the M1/M6, it has already started. Long lines of HGVs spaced too close together were a feature of both roads. Perhaps platooning might stop the interminable process of an HGV travelling at 55.5mph being overtaken by an HGV travelling at 56mph.
Regarding autonomous vehicles in London, it will be interesting to see if the white van autonomous vehicles will be allowed to block redways, park where they liks and so on, also, if this is all connected, perhaps the system might say to a vehicle…….we’re full, for now, come back in an hour.
As a pensioner, I am not seriously expecting to see autonomous vehicles in routine use except where they have their routeway fo themselves (Heathrow T5 to business car park)
Perhaps we will know that auto-cabs have truly come of age when one pitches up at the door and tells us ‘traffic’s dreadful. You would be better off walking’.
However wonderful these beasts become, well-used buses (driverless or not) will always be better users of scarce road space. Perhaps the (sort of) ideal will be for the road space to be equally divided between bus-lanes and the rest, as auto-cabs will apparently be more efficient than their human operated equivalents.
Graham H
That suggests to me that their software is incredibly risk averse, with all the implications of that for traffic management.
Entirely correct – & remember, at present they are limited to more-or-less built-up areas & 25 mph.
P.S. All this talk of narrowing lanes … errr … cyclists? Will still, presumably be around?
EE
Your scenario is self-limiting, though, isn’t it?
Why do you think people take “trains” into & around London, anyway?
In many places the traffic is already close to or at the point you describe.
HC
The same goes for intercity rail, especially as motorways are likely to be the first roads to be converted to driverless-only use, allowing much higher speeds and capacities than at present. I can’t see it surviving the competition.
Wrong, completely wrong.
This is the same prediction that was made in the early 1960’s & pushed by [Snip!] Marples & his cronies, & is still being peddled by the IEA … & it’s still wrong.
Unless, you are positing significantly higher road-speeds for the vehicles. (?)
Outside Kent, Sussex & Surrey, the train is going to start later & get there first, easily, even without HS lines.
Or were you trolling a bait, for some of us, like my self to rise to?
@Malcolm 2217, 100andthirty
The difference between what you see now and automated platooning is that, in the latter, the HGVs will be much closer together and that the spacing will not vary. Even on the M1/M6 it is almost always the case that if you want to leave at an exit (and suitably indicate) at least one HGV driver will slow down sufficiently to let you through. With automated platooning, as currently planned, that cannot happen.
@Graham H
‘driver’= person in charge of the vehicle. Even if all they have done is to press a start button they will still be ‘in charge’ even if there is no emergency stop button available, as they will have made the conscious decision to press the start button with, presumably, the knowledge that they have made all the other checks they might reasonably be expected to make. This will not, of course, include checks on whether the doftware is working correctly in exactly the same way as today a driver is not expected to check that the software in today’s cars is working correctly.
And … just for laughs:
Ian J – it’s “Omnibus” isn’t it?
The potential, as I’ve hinted at, for mischievous people, like bored teenagers, to utterly screw with autonomous vehicles is immense.
All you need is an old tin can & a length of string to swing the can into the sensor-area of said vehicles. [ And plenty of other tricks, too … ]
Oh dear.
@Ian J – I think you are right to be sceptical about driverless car utopia. One point about the Google car trials in California is they are being done on streets that Google themselves extensively and continually map to much greater detail than publicly available maps. It isn’t the case that you could plonk a Google car anywhere and it would be able to navigate purely based on its sensors and a bog standard GPS map.
Of course, Google aren’t going to make this mapping information available for free and will be selective about the areas they map as there is both the initial cost of the mapping and continual updates. I don’t know how a Google car deals with workmen plonking down a “diversion” sign and a few cones, for example, but I’m guessing they will want to communicate that a lane is out of service to the maps in other Google cars (but not their rivals…)
Maybe the sensors and software will become so good that the cars won’t need very detailed maps but as you say, this is an example of non-existent technology.
Regarding product liability, one way or the other, the cost of this will just get passed on to the consumer. Todays manufacturers have product recalls, class action law suits, defective components from suppliers and one way or the other, the consumer pays…
The story about jaywalking laws in the US being introduced at the behest of the car manufacturers is well known but I wonder if driverless cars will also mean the manufacturers try to legislate everyone else’s behaviour to suit their creations?
I can see kids discovering that “bounce the ball in front of the driverless car” is a hilarious game…
Practical reasons to be against platooning on the Motorway. How will manual drivers break through a convoy to get in the correct lane for a junction. Remember the idea of safe stopping distances.
It’s all very well saying they stop within milliseconds, but that does not take into account catostrophic failure, say with a blown tyre or sensor failure. Instead of frequent smaller tragedies, we’d get massive multiple vehicle crashes with mass loss of life.
The motorways are already full of near stationary traffic near major cities, there won’t be any big increases in road capacity without serious road building. It’s the same techno wishful thinking as the capacity upgrades expected from the latest signalling technology.
@Ian J, well yes I expect bus routes to survive in most larger towns and cities.
@Harry Crayola. I imagine lag time for calling Uber in Zone 1 is low, in Zones 4 and outwards though? Having your own car brings advantages of instant availability at the busiest times, your cost is owning a fast depreciating asset that you don’t use most of the time.
In a truly digital transit environment your digital assistant will have all your favourite destinations saved and each time you press ‘Mum’ or ‘Office’ a list of options could roll down your screen giving estimated timings and how much it will cost.
Those wanting a quicker journey may spring for a personal cab to the nearest transit centre or even for the entire journey depending on the time of day and traffic conditions. Your own assistant will know your own tolerance to walking or cycling and could throw those options into the mix.
A more sophisticated version will have no fixed bus routes just mini buses aggregating peak flows with people either willing to wait and pay for direct trips or accepting multiple journeys with your bus seat vibrating to remind you to get off to change buses or you have reached your destination. Fixed routes could survive where they can offer high frequency such as every 5 minutes or less.
A networked transit system probably won’t be much needed in Inner London, but I can see the appeal lots of automated mini buses and some cabs replacing the need for cars for a substantial portion of the population.
Each bus could add their destination on the front screen with affix ‘Only’ added and only stop at the stop where those passengers are waiting or divert to pick them up if their willing to pay a little more, their phones would remind them this is the bus they wanted and then once it had picked up enough passengers for it’s destination it could speed on to it’s destination with no stops.
Virtual transit hubs could be based at all major destinations from, Uxbridge town centre, Brunel University, the Hospitals and all the railway stations etc
Main roads could end up with express services and local services depending on the characteristics of the road and whether passengers are clumped of dispersed along the route.
[Far too long and rambling comment that has had several paragraphs abridged. LBM]
I think it is also fanciful that a driverless car would replace a lot of main family cars although “two car” households may become “one car” households.
Thinking about a typical suburban family car, it may have child seat strapped in the back, children’s toys scattered all over the floor, smelly blanket and dog basket in the rear… I don’t think people will cart all that stuff in and out of a car every time they make a journey…
Anomnibus 26 April 17.54 wrote of the white van becoming little more than a self-driving box….
Interested to know whether it would be programmed to park half on the pavement and half on the road with indicators flashing and, more seriously, who is going to unload it or run up two flights of stairs to get a signature! ?
@0775 Indeed I can see automatic trucks not needing drivers transporting goods between large warehouses where there are people to load and unload at each destination but local distribution? Where the driver also unloads the van and gets the customer to sign! At most it might save a premium on HGV licenced driver but urban deliveries will require a drivers intervention to judge on what non standard parking he can get away with. I can imagine many areas where an official loading bay is either full or a long way away from it’s customer. You might find customer resistance to the idea they have to trudge down to a waiting van and enter a code to open the door and then pick the right package and no others. A major appeal is a the safe delivery to their office/shop on a little trolley. Who has the spare staff to leave a shop to get a delivery?
Ratty et al
Just a thought: automated goods reception / shop lockboxes. If you can solve vehicle automation, then this should be a doddle.
@Ian J
“Half their current amount is still more than buses or trains, so I’m not sure the economics really work there”
Well, current cost structures support current useage patterns. So what effect would radically shifting the relative cost of driving compared to taking a train/bus have? Any effect from such a move would be reinforced by any changes to service patterns resulting from the first level effects, too.
“road pricing… would make most rural and suburban roads very expensive to travel on”
The theory I’ve read suggests this might be true of rural roads but not suburban ones. The expensive ones would be those where the land value plus maintenance costs per journey are high. Suburban roads have sufficient usage to benefit from a large denominator while not being too expensive in terms of the land value numerator. It’s basically congested (ie, urban) roads and unused (ie, most rural) roads which would not be able to sustain current expenditure/usage patterns. In other words, deadweight costs would be reduced.
“Techno-utopians have been predicting the imminent demise of the rail network for decades”
They were right in the mid 20th century. But point taken. Perhaps there are things that all the investors pumping billions of their money into this haven’t seen yet that, even with another 10, 20 or 30 years of technological progress, can’t be overcome. It is, after all, a bit of a gamble that they’re taking. I’d be nervous about betting against them, though, especially given how close they already are.
“motorways rarely reach the centre of cities either”
No, but
1) not everyone’s destination is the city centre,
2) they could reach a metro station on the outskirts of the centre, for those unwilling to pay the road-pricing/congestion charging premium for the last section of a journey. And
3) trains rarely reach the front door of your destination, unlike cars, for those who are willing to pay any necessary congestion charge/road pricing.
With the knowledge that Google’s experimental cars rely on intensive local mapping, it might well be that a bus system could pioneer automated vehicle control first. Only the bus routes would need to be mapped intensively, and that could be done by equipping just one of two of their ‘normal’ vehicles with the survey sensors, and ensuring that they are rostered to sweep the entire route network in order. Stops for loading/unloading would be pre-programmed, so the Graham H garage conundrum does not need to be solved. Which city system do we think will be the pioneer?
Re. “White Vans”:
Surely, if you’re expecting a delivery, you’ll have staff on-hand to collect it from the computer-driven* van?
If there’s a problem with parking, then the same computer-driven technology can also be applied to, say, small motorised pallet-sized trolleys (similar to those seen today in supermarkets) that could exit the vehicle and make their own way to the loading bay or service entrance.
(It might even be possible to use drones for smaller deliveries. Amazon are already trialling this. There’s no reason such drones couldn’t be based in a computer-driven “drone-carrier” vehicle that brings them within flying distance of their destinations.)
I think this is the part many people really struggle with: just because the vehicles we use today look the way they do now, it doesn’t follow that they will continue to look and work that way in future. Once you have the core computer-control technology, entirely new designs and concepts become possible that would have made no sense if human drivers were required.
—
* Arguably a more accurate term than “self-driving”: Just as the railways began with effectively independent trains separated by fixed time intervals, and eventually moved towards trains that are told what to do by computers based miles away in a dedicated control centre, so essentially the same process of centralised command and control is now being applied to road-going vehicles.
While each vehicle will, of necessity, be able to operate autonomously at a ‘tactical’ level, ‘strategic’ control would be handled by a central Traffic Control Centre of some kind.
Road users on foot will not interact very efficiently with driverless cars. What if a pedestrian starts to cross the road in a space where a driverless car has just decided to pull in? Will we need to install KED’s (kerb edge doors) alongside all roads with pavements?
As to what will happen when the more assertive members of the cycling population and aggressive pedestrians discover that a driverless car will ALWAYS give way to them…….
Perhaps the following might help:
1. A driverless car may not injure a human being or, through inaction, allow a human being to come to harm.
2. A driverless car must obey orders given it by human beings except where such orders would conflict with the First Law.
3. A driverless car must protect its own existence as long as such protection does not conflict with the First or Second Law.
(apologies to the estate of I Asimov)
Re Nameless,
“What if a pedestrian starts to cross the road in a space where a driverless car has just decided to pull in? ”
The google cars will probably slam on the brakes – they already understand joggers tend to cross in different places to walkers so get very nervous if the jogger is near the edge of the pavement as that is sign they might be thinking about crossing.
@Nameless:
“Road users on foot will not interact very efficiently with driverless cars.”
If road accident statistics are to be believed, they don’t interact very efficiently with non-driverless cars either.
Anomnibus above…
I’m in no way disbelieving of the ideas about white vans and self-driving cars being presented above but it seems to me that it is not a simple as making the technology work. It will no doubt be capable of doing all the things that you say – but in order for such a future to develop it requires the good old human being to be accepting of the changes and regard this form of the future to be a “good thing”.
Perhaps the best and most useful future development would be for a new experimental greenfield community to be built incorporating all the ideal advances needed and in a setting where everything would be run using “smart” technology from transport through to domestic central heating systems through to shopping and waste disposal. Then those who wish to be involved in developing such systems can test at their hearts content on willing Guinea pigs.
Self-driving cars arriving to take you to the station already knowing that your bus/train is delayed and forewarning you that that they will be giving you an extra 15 minutes at home so no waiting on cold platform etc etc would be quite a boon..
It surely must be cheaper to start from the ground up rather than convert Victorian thoroughfares to make space. [Last bit about repurposing London snipped as is unrealistic, and we do not want a flood of urbanista comments. LBM]
@ Walthamstow Writer re orbital transport links..
The problem isn’t so much whether the links exist, but at what frequency – especially if you need to use more than one. This is particularly true of rail – frequencies on outer-suburban lines tend to be limited 1) by bottlenecks further in to town, and 2) by TOCs who, unlike TfL, don’t seem to recognise that weekend travel patterns have changed massively since the 70s/80s, when everything was shut & so everyone either stayed home, or drove to the countryside/seaside.
This in turn means that if you’re unlucky with connection times, especially on Sundays, you can be waiting for ages.
Case in point – West Norwood has supposedly easy one-change connections to Wandsworth (via Clapham Junction), Wimbledon (via Tulse Hill), Bromley (via Beckenham Junction). All 5 miles, give or take. But actually try and do it on a weekend & unless you live literally above the station, you’re almost certainly much better off driving – even if it involves the South Circular.
Realistically the capacity issues closer to town are going to be difficult and expensive to fix, if Thameslink is anything to go by. So I’d like to see them look seriously at linking segments of under-capacity outer-London rail together to run orbital trains in addition to the existing radial routes – allowing a higher overall frequency, less need to change trains, and shorter waits when you do have to. No specific examples lest I be accused of crayonism.. but it does appear that both TfL & the TOCs are far more focused on the rush-hour capacity crunch on the radial routes than in seriously trying to make orbital rail competitive with other modes.
(As an aside, 5 miles is a very cycleable distance, competitive with rail door-to-door unless train frequencies are high, and far, far quicker than the bus, but I suspect it’ll be a very long time before we see links between outer London town centres suitable for everyone to cycle).
Harry Crayola et seq: widespread, mobile, pervasive technologies have a different set of security problems from desk (or server room) based systems. “Smart cities” are pretty hackable; see the original “Italian Job” for details. I worry about the robustness of these new technologies. There is already a market for wallets lined with metal mesh so that your ‘contactless’ cards are protected by a Faraday cage from “swipe by” theft.
@Nameless Re Asimov’s 3 laws
I would worry about the prospect of driverless cars crashing into other vehicles to prevent said vehicles from knocking over cyclists, in accordance with the 1st law!
@ Binary Funt
It’s surely better that property is damaged than people are hurt or killed. Though such a scenario would rather strengthen the argument for cyclists (and indeed pedestrians) to be required to have 3rd party liability insurance.
Ian J: “you need an increasing gap between each vehicle, just as in a railway moving block signalling system” – Actually, no. Wrong comparison. In the case of road vehicles think more of them being like an Aussie road train – fixed together – rather than independent vehicles. The tow bars just happen to be wireless not metal rods.
Personally, I’d always assumed that ‘the system’ would require appropriate transit types, ie trains for long distance, via trams and busses and down to pods for the final home connection. Like the calculation between whether to get an APEX train fare or drive oneself (and have to find parking) it will all come down to money in the end.
0775john: “who is going to unload it or run up two flights of stairs” – I already have this problem as I live on the fourth floor with no lift. Whilst food deliveries are no problem (they want COD!) nobody else will deliver to my door.
“driverless cars crashing into other vehicles to prevent said vehicles from knocking over cyclists” is a good thing — if it is based on a prediction about the specific vehicle in question apparently about to knock over a specific cyclist.
However, if the self-driving logic goes “Moving cars sometimes kill cyclists. I see a moving car, therefore I must stop it” then I would argue that it has misunderstood the laws. (As quite often happens in Asimov’s stories, as I recall).
AlisonW: whether you need an increasing gap as software-coupled vehicles set off simultaneously actually depends on the separation strategy. If the strategy is fixed-distance, then it is as you say, the software is just virtualising a tow-bar. But that strategy risks catastrophic consequences if the first vehicle collides with something – they will all collide and a 2-vehicle accident becomes a 22-vehicle accident with probably unacceptable consequences for the thing-collided-with and other stuff in the area.
More likely is a separation strategy where the separation distance depends on speed – not necessarily a full braking-distance one, but a compromise something like the distance that human drivers typically leave in front of them. Any such strategy would indeed require a different acceleration rate for each vehicle in the platoon.
@ 0775John – and when all those clever cars all converge on the station at the same time because everyone has decided to have 15 mins more at home? Result – huge traffic jams on the approach roads, a shortage of stopping positions and people miss their trains. Yep that sounds like a solution.
@ Various – I’m sorry but all this blind belief in an unproven technology just leaves me shaking my head. Perhaps I’m just a luddite but I just cannot see how you make driverless cars viable, safe and acceptable to the vast numbers of people who have a massive range of transport demands. I am also left wondering just where these “driverless vehicle only” roads are going to come from? Are these existing roads that are converted and thus become the monopoly of private sector provided transport services? How funds their maintenance and repair? Are the emergency services and utility companies banned from accessing them for essential purposes? Are we really going to turn over existing roads to huge concerns who will want a monopoly of use *and* control? Are we really going to prevent human beings from crossing them if they are in urban areas? This concept simply cannot work in established urban areas unless you flatten them as was threatened with urban motorways in the 1970s and we know what became of them.
There also seems to be a proposition being put forward that all of this just glides into place with no resistance from people, no legislative issues etc. Although you’d never know it from the palpable lack of “joined up” policy across government transport and mobility are essential enablers of a massive range of other societal functions as well as individual wellbeing and health. I simply don’t see all the respective interests in those areas surrendering themselves and their priorities into a Google/Uber/Amazon electronic vehicular dreamland / nightmare. It simply isn’t going to happen. Oh and retaining my luddite qualifications I actually quite like using buses and trains and finding my own way about rather than being encased in a robot like bubble which I have no control or influence over. I doubt I am alone.
@Rational Plan: A more sophisticated version will have no fixed bus routes just mini buses aggregating peak flows with people either willing to wait and pay for direct trips or accepting multiple journeys
Such things exist throughout the Third World – dolmus minibuses in Turkey, for example. Minibuses ply particular routes, waiting until full and then setting off. You can connect between routes at nodal points. There are two problems with this system – firstly, it takes up a lot of kerb space at these nodal points with buses waiting for passengers. Secondly, journey times are unpredictable because the first passengers on board have to be prepared to wait an undetermined time until the bus is full. (And in some countries, the operator’s idea of how many people should get on before the bus is ‘full’ is, umm, interesting).
Hence the disruptive transport technology adopted by innovator George Shillibeer in 1829 adopted the radical approach of leaving at set times whether full or not. (Like all good entrepreneurs he got the idea from someone else, in this case the French.)
@Harry Crayola: So what effect would radically shifting the relative cost of driving compared to taking a train/bus have
I can see that self-driving cars would shift the relative cost of taxis and minicabs compared to taking a train or bus. But self-driving technology alone doesn’t reduce the cost of driving for people who drive themselves – in fact it increases it, since you have to pay for the self-drive technology, whereas most people value the labour cost of time they spend driving their own car at zero. Only the very wealthiest would pay someone to drive them from London to Birmingham, say, so eliminating the driver alone doesn’t make that journey any cheaper.
For most people the purchase price of a car is a sunk cost. So it is hard to see even shared use cars being significantly cheaper on an incremental per-journey basis for longer trips than driving your own car, in which case the effect on intercity travel is unlikely to be significant.
Suburban roads have sufficient usage to benefit from a large denominator while not being too expensive in terms of the land value numerator
A lot of suburban roads are designed specifically to only be useful to the people who actually live on that road – the pattern of cul de sacs and loops that makes up much postwar suburbia. A more economically rational approach would be to shut them down Beeching-style as hopelessly unviable, and encourage the inhabitants to buy helicopters.
Perhaps there are things that all the investors pumping billions of their money into this haven’t seen yet that, even with another 10, 20 or 30 years of technological progress, can’t be overcome. It is, after all, a bit of a gamble that they’re taking. I’d be nervous about betting against them, though, especially given how close they already are.
Silicon Valley has a long and proven track record of pumping billions of other peoples’ money into things that are not actually viable and have no business case. In any case, it would not be necessary to transform the nature of transport to justify the sums currently being invested – just take a big chunk of the global taxi/minicab/sharecar industry (Uber) or the car industry (Apple, Tesla) and you could make plenty of money. Incidentally the latter two seem to be betting their billions on an ownership model, not a shared vehicle model, being the future.
@Hilltopper: I’d like to see them look seriously at linking segments of under-capacity outer-London rail together to run orbital trains in addition to the existing radial routes
AKA the “R25”, which doesn’t seem to have been mentioned in the current election campaign at all (or by anyone at all since Boris’ transport plan was published).
@AlisonW: As Malcolm suggests, a fixed separation distance is only safe if the vehicles are physically braced against each other, so that deceleration forces can be safely transmitted from one vehicle to another, as through the couplings in a train.
Moreover, the braking distance of a vehicle with such a high combined mass is much higher, which is why road trains are only permitted in remoter areas of Australia where there not too much other traffic, and there is a shared understanding amongst road users that the onus is on you to get out of their way.
a separation strategy where the separation distance depends on speed – not necessarily a full braking-distance one, but a compromise something like the distance that human drivers typically leave in front of them
A problem here is that the Highway Code strongly encourages full-braking-distance separation. This is not observed by most drivers, but given that the Highway Code can be used in establishing liability, it would be a brave technology company that programmed anything less into their system.
@Hilltopper: the argument for cyclists (and indeed pedestrians) to be required to have 3rd party liability insurance
“I’m sorry Mrs McGarrigle, little Johnny can’t come to school today because he can’t leave the house. His pedestrian insurance has expired and I can’t afford to renew it. The insurance company have had to double the premiums for children because of the billions they are paying out to Google”.
I’m sure Sian Berry would have something to say about that.
[Far too long a comment, please break up responses to other commentators as individual comments so that it is easier for them to find. LBM]
@Malcolm – One could envisage multiple cars attempting to save the same cyclist. Actually there was an article in Nature about this.
Clarification: the article discusses the dilemma of who to save in the context of Asimov’s 1st law, not the particular scenario I suggested
@Walthamstow Writer
I really do think you’re missing the picture with this, I’m afraid. There is big, big money being poured into this and the technology has already advanced. It’s working already in small controlled environments. So, in all likelihood, it is going to happen. Bear in mind that, similar to “driverless trains”, there’s a scale of automation and it’s unlikely to be fully automated cars on the road next year, not least as public acceptability, consumer demand and legal liability issues take time to resolve/evolve. But it’s not slow-adopters and sceptics they need to convince any time soon. It’s technophiles and neophiliacs.
Interestingly, there has been a bit of news today about this, about Volvo testing cars on London streets early next year, with Patrick McLoughlin talking of the UK leading the way and George Osborne recently talking about reforming legislation to allow driverless cars on motorways. Meanwhile, China has been busy, too.
I have reordered/regrouped some of your questions for my convenience, I hope you don’t mind:
“I just cannot see how you make driverless cars viable, safe and acceptable to the vast numbers of people who have a massive range of transport demands”
I don’t think this matters. Firstly, because it doesn’t matter if driverless cars don’t meet everyone’s demands. They only need to meet those who will use them. In zones 1 and 2, I don’t think they’ll change that much in terms of the modal share in commuting. But the further out you go, and the further from peak times you get, the bigger the impact I suggest they’ll have.
“just where these “driverless vehicle only” roads are going to come from? Are these existing roads that are converted and thus become the monopoly of private sector provided transport services?…Are we really going to turn over existing roads to huge concerns who will want a monopoly of use *and* control? How funds their maintenance and repair?”
They won’t happen soon. They’ll have to wait until the demand is sufficient to justify disallowing other traffic. My guess is we won’t have the first one until 2030 at least (assuming driverless vehicles become mass-marketed within 5 years, and then fleets take another 5 or 10 years until over 50% of cars are driverless – 2.5 million a year sold, 37 million total on the road). I expect the first road to be converted will be a motorway, because that’s where the gains are most obvious and they’re already highly restricted and controlled environments. I don’t understand why you fret about “monopoly” or “private sector provided” vehicles.
Nothing needs to change on these driverless-only roads except that manually-driven cars will no longer be permitted and, I’d guess, lanes narrowed to add capacity. I don’t see why driverless public sector coaches/buses couldn’t use them too, if desired. I also don’t understand the worry about “control”. Motorways might be privatised by then anyway, but that’s a completely separate question. The existing Highways Agency could simply add a symbol on the road signs indicating driverless only (together with some enforcement penalty) and that could work fine.
“Are the emergency services and utility companies banned from accessing them for essential purposes?”
Of course not – why would we want to do that? Firstly, I expect most/all of their vehicles would be driverless, too. Second, of course a manually driven emergency vehicle would be able to ignore the road sign and drive anyway – the driverless cars would detect that it was a manual vehicle and react accordingly.
“Are we really going to prevent human beings from crossing them if they are in urban areas? This concept simply cannot work in established urban areas unless you flatten them as was threatened with urban motorways in the 1970s and we know what became of them.”
Driveless-only roads in urban areas I’d guess would be much further away (2040? 2050?) But I expect they’d operate much as they do now. Eg, no pedestrians allowed on the Westway etc, but fine in most other places. I certainly wouldn’t want to see any restriction here. Car companies in America lobbied for jay-walking laws in the US. This kind of hi-jacking of the law by corporate interests is my only substantial worry about driverless cars. But as long as we don’t put up with it, it won’t happen!
“retaining my luddite qualifications I actually quite like using buses and trains and finding my own way about rather than being encased in a robot like bubble which I have no control or influence over. I doubt I am alone.”
You’re not. I agree. Sometimes. If I’m short of time I’d rather be taken as fast and conveniently as possible. But often little detours are interesting, I fully agree. But three points here:
1. I don’t think anybody suggests you can’t instruct your driverless car to amend its route if you fancy it.
2. You’ll be able to drive yourself manually for a long time to come.
3. Nothing to stop you taking trains, buses (subject to commercial viability for your route, as now) and walks if you would rather do that.
@Ian J
“self-driving technology alone doesn’t reduce the cost of driving for people who drive themselves…most people value the labour cost of time they spend driving their own car at zero”
I don’t think this is backed up by the pricing evidence of homes with shorter commutes, transport modes with faster journey times, shop rentals in locations that are more spatially convenient for customers etc, etc.
“Only the very wealthiest would pay someone to drive them from London to Birmingham, say, so eliminating the driver alone doesn’t make that journey any cheaper.”
And what happens when you halve the cost of being driven to Birmingham? That’s the whole point I’m making. For people who don’t want to drive, but can’t afford a taxi now, ie train passengers, driverless cars are going to have a very, very substantial impact on the relative utility/price ratio for trains compared to the next best alternatives. If a train fare costs £30 and petrol costs £25 but you don’t want to drive, what happens when you introduce a £25 or £30 taxi option to the relative attractiveness of taking the train?
“Only the very wealthiest would pay someone to drive them from London to Birmingham, say, so eliminating the driver alone doesn’t make that journey any cheaper.”
My contention isn’t that driverless cars will eliminate individually owned cars. I don’t think that will happen. Instead, I think it will accelerate the move from owned to hired/shared cars. But whether or not I’m right about that acceleration aisn’t that important. Even if ownership patterns remain unchanged, you’ll still have people able to summon their owned driverless car to take them on an intercity trip, or hire one via a driverless car from a rental company/car club/minicab firm (the distinction between the three will surely vanish once minicabs no longer have drivers). I just can’t see many people opting for a train any more for these journeys.
“Moreover, the braking distance of a vehicle with such a high combined mass is much higher, ”
Why should that be so? Surely the mass is proportional to the number of vehicles in the train/platoon. But the braking capability depends on the number of brakes fitted, and assuming all-wheel braking (which is normal on most road (and rail) vehicles) then the braking capability is also proportional to the number of vehicles. A fully-laden lorry may take longer to stop than a car because the weight on each wheel is bigger. The brakes are bigger too, but friction is proportional to surface area (of the brake discs) area and mass is proportional to volume, plus a car has more empty space on board. But an eight wheeler weighting 30 tonnes should be able to stop as quickly as a four wheeler weighting fifteen.
timbeau
IF the mass of a vehicle gets large enough, yes, then the braking distance will increase – the energy dissipated has to go somewhere …
But, for private-car-sized vehicles, the differences are minimal.
[We’ve heard enough about your vehicle, so snipped. LBM]
But your 4-wheeler massing 15 tonnes will not be able to stop as quickly as my 4-wheeler massing 2, will it?.
This article, from yesterday’s WSJ, explores how Google/Alphabet are pushing to construct a prototype ‘smart city’. Essentially, asking hard-up American cities to let it redevelop large areas as testing grounds:
http://www.wsj.com/article_email/alphabets-next-big-thing-building-a-smart-city-1461688156-lMyQjAxMTE2MTI5NjcyMTYyWj
I think the issue of liability in driverless cars is being glossed over. The person using the vehicle cannot be held responsible for any issues with the vehicle – if you get into a taxi and the taxi has an accident, it isn’t your fault. The driver in a driverless car is the software, not the passenger. Any lawyer would easily counter any kind of suggestion that the passenger is in control and so is liable. So if not the passenger, who is liable ? It has to be either manufacturer or the software designer. As the law is based on precedent, it is going to take a several high profile accidents for the law to be adjusted sufficiently.
There are large numbers of people who are not going to accept this change, either because they don’t trust it (quite rightly in my view – is Google the right sort of company to be writing safety-critical software?), but also because they will lose their jobs to it. The Unions are currently planning to strike over changing the role of guards on Southern trains – what will be their reaction be when they are told “we are getting rid of all bus and taxi drivers” ? Driverless cars are nice and shiny at the moment, but when people start seeing the downsides, there is going to be a huge backlash.
The introduction of driverless cars is going to be very, very messy, and any one who plays down the issues needs to stop and take a cold hard look at it.
@Jim Cobb – I can see that liability will shift from the individual to the manufacturer so another loser in this case will be insurance companies who sell people car insurance policies.
I can also see the manufacturers trying to limit their risk so they don’t get sucked into interminable court cases with everyone and their dog jumping on board to claim damages. So they may look for legislation that caps liability.
I don’t think there is any software developer on the planet who will warranty their product to be defect-free so manufacturers will be limited what they can do passing the buck on to sub-contractors as long as the sub-contractor can show they have rigorously followed the defined quality control process. And of course, the more rigorous the process, the higher the price…
@ Harry Crayola
As at today, how does a driverless car obey the instructions, verbal, by illuminated sign or by hand signals, of a police officer or other authorised person.?
For example, a road is closed following an incident such as an RTC or major fire. The police are instructing vehicles to stop or divert. Or there is a census or other checkpoint.
How about recognising cyclists’ or horse riders’ hand signals?
How do they react to a vehicle with “blues and twos”?
Surely until ALL of the potential hazards and situations have been catered for they cannot be allowed to mix with general traffic.
The other problem I see with liability is people will be less forgiving of an automated vehicle. If a child runs out in front of a car with a driver, the reaction would be “it was just an accident” (assuming the driver withing the limits). If the same happens with a driverless car, particularly early in their adoption, the reaction will be “these automated cars are supposed to be safer, so how can this have happened”. There will then follow a media storm and numerous lawyers trying to pin the blame and get a huge payout.
A couple more issues to consider –
– Defining what “safety” actually means and testing to ensure the vehicle complies
– Customising the solution to support regional signs, laws etc., particularly when driving between regions
– Securely updating the software and ensuring vehicles are using the latest software, particularly if the changes were to support changes of laws
– Deciding where to store these vehicles when they are not in use (eg. overnight)
This has been an interesting discussion so far. A few strategic points emerge:
– Many technical problems remain to be resolved, and perhaps can be resolved but at a very high cost with the consequent risk that marginal residual users of manual vehicles (eg motorbikes, agricultural users, people living on the remoter parts of the road network) will be excluded as not worth “converting”
– there is a high risk of attempts to supply closed source kit
– the deals on liability are likely to disbenefit end users and 3rd parties
– culturally, many may find the idea difficult to accept
– it’s not likely to generate much extra capacity in already congested, slow moving streets.
@Jim Cobb – there are different approaches to liability even without driverless cars. For example, the Netherlands has strict liability for collisions involving children so a driver will ALWAYS be liable for 50% to 100% of the assessed damages (an exception being if there is intent to cause the collision)
Essentially the Dutch approach is the driver is responsible for controlling lethal machinery and liability for collisions with children is one of the responsibilities they need to accept even if “it was just an accident”.
@Reynolds 953 – Interesting.
So will that mean that the Dutch Driverless car will be configured to avoid children more than the UK version of the same car ? If the manufacturers have to pick up the liability, they will have to consider how regional customisation affects their liability.
@Jim Cobb
“if not the passenger, who is liable ? It has to be either manufacturer or the software designer. As the law is based on precedent, it is going to take a several high profile accidents for the law to be adjusted sufficiently.”
I software supplier liability is the most common working assumption I’ve seen. It’s been my assumption. Vehicle owners would need to subscribe to a software service and comply with its updates policy and reasonable hardware maintenance to ensure that liability was transferred to the vehicle manufacturer or software provider. I suspec this would happen gradually, as cars become gradually more automated before eventually not requiring a driver at all. And yes, the beauty of common law is that it adjusts rapidly and reasonably predictably to these sorts of social changes. But as you say, we’ll have to see how courts apply the “reasonable” principle to the details.
“is Google the right sort of company to be writing safety-critical software?”
When the legal liability is established by the courts, investors and insurance markets are best placed to answer this question, balancing their cost efficiency with risks of legal liability in their investment decisions and insurance policies.
“Unions are currently planning to strike over changing the role of guards on Southern trains – what will be their reaction be when they are told ‘we are getting rid of all bus and taxi drivers'”
I’m sure they’ll stamp their feet and demand others give them cash payments or progress-blocking regulation so they can carry on financially benefiting from other people’s (previously inherent, now due to technology artifical) inconvenience. But we needn’t worry too much about that. Similarly to the risk of corporate interests hi-jacking the law like Big Auto did with jay walking laws, it won’t happen if we don’t put up with it!
@Jim Cobb – I think that the approach to liability in the UK is just one of the many things that would need to change with driverless cars.
So perhaps strict liability for some circumstances and presumed liability for others and a sliding scale depending on the actions of the injured party, such as recklessness.
Driverless cars should have a whole set of instrumentation to show what happened in a collision anyway.
Those types of liability rules can equally apply to drivers – they do in many countries already.
As has been mentioned, the worry would be if the manufacturers lobby for laws to keep those pesky humans away from their creations to limit their liability payouts. This isn’t a problem on motorways but I don’t see the UK being ready to accept jaywalking laws on urban high streets but who knows…?
Or maybe the vehicles become so good that strict and presumed liability applies to humans..
I reckon that driverless car manufacturers will strive for one basic approach to the European market. Minor variations will be permissable, since crossing national borders is usually) easily detectable by the cars. Expect an attempt to harmonise driving or liability laws where these are sufficiently different so as to require fundamental changes to the cars’ logic.
This approach suggests that UK/Ireland will have driverless cars later than European countries that drive on the right.
@Graham H
Thanks. Good summarising points. But a few questions:
“Many technical problems remain to be resolved, and perhaps can be resolved but at a very high cost with the consequent risk that marginal residual users of manual vehicles (eg motorbikes, agricultural users, people living on the remoter parts of the road network) will be excluded as not worth ‘converting'”
A couple of queries with this:
1. I don’t know how you define “very high” on costs to resolve technical problems. We can assume that driverless car tech investors don’t share this assessment, certainly not in proportion to the monetisable benefits that the investments will be able to benefit from.
2. I don’t understand the point about people on remoter parts of the road network. Do you mean that they are excluded from congestion relief benefits because their roads aren’t congested and there is therefore no congestion to relieve? Or that some will keep their old manual cars for some time to come and won’t bother upgrading to a driverless new one? Does this count as being “excluded”?
“it’s not likely to generate much extra capacity in already congested,slowmoving streets.”
I think it’s in stop-start traffic which is constrained by junctions where the benefits are lesser, rather than congested roads. What capacity benefit might there be in these streets? I guess current congestion is largely a function of how many vehicles can pass through junctions in a given time. For instance, at traffic lights. So if faster reaction times of driverless cars mean that, say, 10% or 20% more cars can pass through a set of lights in the “green” time, we therefore have an improvement of only 10% or 20% in capacity. I suspect the improvement on uninterrupted main roads and motorways could be significantly better, especially when we can paint in more lanes from the same space.
Assuming it works what are we looking at? More and more cars that do more for the driver, you can already buy cars that parallel park, maintain safe distances, maintain lane discipline and automatically stop in time for obstructions in the roads (if below 35 mph).
[Generalities which repeat previously made points snipped. LBM]
Computer-driven vehicles will likely have their own equivalent to a “black box”, as well as video cameras, precisely for insurance and liability issues. Insurance companies are already encouraging fitting such technologies in cars today, so this isn’t new.
There are any number of videos on YouTube showing why dashboard cameras are already very popular in Russia and other parts of Eastern Europe. (If nothing else, most of the accidents seen in those videos would be very difficult, if not impossible, if computer-driven vehicles are the norm. Computers don’t tend to get drunk or high.)
If an accident occurs, there would therefore be no shortage of hard evidence to prove culpability in a court of law.
The hard part will be dealing with people who don’t believe in taking responsibility for their own actions. (Including their parenting skills, if applicable.)
Refining and updating the software to handle unexpected edge cases is likely to be rather easier by comparison.
@answer=42:
The side of the road we drive on should, if anything, be the least of our problems. The main thing it affects are which side of the vehicle gets the steering wheel, pedals, and dashboard. If the vehicle doesn’t have a human driver, none of that applies.
Any computer-driven vehicle’s software should be able to cope with roads regardless of which side they’re driving on. Failure to do so would effectively rule out sales to any European logistics companies that send goods between the UK and continental Europe.
Japan and India also drive on the left. India’s population numbers over a billion people, while Japanese car companies are also doing R&D in this field, so they’d certainly have an interest in selling into their home market.
@Harry Crayola – my fault for compressing my argumentation… By “very high” costs, I was assuming that that assessment would be a judgement made by the system supplier and if asked to provide a “universal” system extending to the whole of the road network, they would argue that Mr Pareto made 80% of his journeys on 20% of the network (as is indeed the case) and that the cost of adding in the lightly used parts/vehicles shouldn’t be an obligation – the usual debate with regulators, in fact.
I agree that the principal beneficiaries are likely to be people in the quieter rural areas where roads are not congested and public transport is hardly visible. Whether urban capacity is released on any significant scale depends essentially on two things: the degree of risk averseness in the software and the ability to use parking space as efficiently as now.
I’m not sure Harry Crayola’s London-Birmingham example really works.
When I drove it took me an hour to reach the junction of the M25/M1 from my home in inner London. Following the slog on M1 my best total time was 2hrs 35 mins to the centre of Birmingham, ‘sometimes’ at illegal speeds.
That will never compete with HS2 and the 49 mins journey. I also have a Senior card nowadays, but I’m not sure if that will be valid on HS2.
@Greg
“But your 4-wheeler massing 15 tonnes will not be able to stop as quickly as my 4-wheeler massing 2, will it?.”
Of course not, but what we were comparing was a single truck vs a platoon, convoy or train of trucks. There is no reason why a 16 wheeler massing 60 tonnes should take any longer to stop than a 4 wheeler massing 15 tonnes, provided each wheel has the same braking capability. (It would indeed take longer to stop if it were to hit something – where all the braking force has to be dissipated through the front end into whatever it hit, as can be seen by comparing the effect of a train braking (where the number of vehicles is not a factor) and the effect of hitting something (where it most definitely is)
But the only difference between the brakes in the convoy all being under the control of one driver , or of individual drivers in each truck, is that in the reaction time depends on how quickly the control signals are passed down the convoy – in the human case that is essentially the reaction time between the brake lights coming on and the driver responding, in the one driver in charge case it depends whether control is relayed from one vehicle to the next, or whether the control system controls each vehicle separately. Some allowance needs to be made for differences in weight/braking capability, either by adapting to the worst performer, or varying the separation to take account of this.
Thje interesting debate on liability presumes that driverless cars are on their way. But they are already on the roads! Not many yet, but cars capable of driving on real world roads without drivers are here. You can buy one if you want. Their numbers will increase. And while it may not currently be legal to use them in driverless mode, I have no doubts that at least some of the ‘drivers’ will be tempted to use them in driverless mode. Most of these occurences will not involve accidents and will, hence, be undetectable by a casual observer. Numbers will only increase in a way which will be unstoppable. So perhaps the time has come to stop thinking of an ideally planned world where we sort everything out before driverless cars appear on the roads and think, instead, of how we sort out the problems when they do occur.
Anonymous compares his 155 minute drive to Birmingham with 49 minutes on HS2. That is only a fair comparison if you consider only those people who want to go from somewhere close to Euston station to somewhere close to Curzon Street station, and only then for people indifferent to whether they have a car at their disposal when they arrive or not. For most people, a car will either be a benefit (giving more choice of return times, opportunity to drop in on Auntie Sarah on the return journey, etc), or a nuisance (requiring expensive parking).
There must already be insurance precedent for self parking when it goes wrong. Anyone know?
http://www.independent.co.uk/life-style/gadgets-and-tech/news/self-parking-volvo-plows-into-journalists-after-owner-neglects-to-pay-for-extra-feature-that-stops-10277203.html
@quinlet
I think you ought to have said:
“cars *just about*capable of driving on *extremely simple relatively empty* real world roads without drivers are here”
I’d be happier knowing for sure that they would pull over if an emergency ambulance was fast approaching from behind.
There is no way that they should be allowed on UK public roads mostly occupied by other road users until after a huge amount of further development and the implementation of a workable satisfactory legal framework.
I reckon we’ll see a driverless Piccadilly line working before then.
There is a lot of misplaced emphasis on the private car in this discussion, despite their making up only 5% or so of commuter traffic in London. Computer-driven cars will have very little effect on that percentage: it’s a question of infrastructure capacity, not driving technology. If nobody wants to upgrade the road network, no amount of automation in the vehicles will help. Us humans aren’t getting any smaller.
(There may be more single-seat vehicles, but I suspect the desire for room to carry shopping, and minimum safety requirements for road vehicles, will mean they won’t be small enough to make a useful difference to capacity.)
What will make a difference is applying the same technology to other vehicle types. An awful lot of dots can finally start being joined together.
The emphasis on the private car in the discussion may be related to public reports, which are mostly about so-called google cars. Although the technology for other vehicle types may well be very similar (and arguably, as Anomnibus says, the benefits greater, perhaps especially for London), the private car market is certainly the largest in terms of numbers of vehicles, so perhaps the most fertile ground for spreading the considerable development costs over a large number of units.
And when it comes to public acceptance, many people who find driverless cars a bit scary will probably feel the same only more so about buses or trucks, so starting at the least scary end of the market (at least in public) may well be a wise strategy for the developers.
@Nameless
I was driven in one on Westway during a weekday afternoon. Of course it is only category 4(driver present and able to take over at no notice), rather than category 5 (no driver at any time), and using it in driverless form is probably unlawful at present. But, as I said, provided there are no accidents, whose to know?
Walthamstow Writer: Clearly, your “driverless vehicle only” roads will actually be iron rods above the lanes of current roads, supported at regular intervals by some sort of scaffolding with the cars held on by magnets. You could call it a “personal driverless monorail”.
/runs away and hides from the rapidly approaching mod 😛
[harrumph. Malcolm]
quinlet: Are the categories to which you refer (1) descriptions of what the car is equipped and capable of doing, or (2) descriptions of what the car actually was doing? And did the human driver, as far as you know: (a) Watch but not intervene, or (b) watch and sometimes intervene, or (c) do most/all of the driving?
@Harry Crayola:
“most people value the labour cost of time they spend driving their own car at zero”
I don’t think this is backed up by the pricing evidence of homes with shorter commutes
I said labour cost, not travel time cost. Certainly people want to reduce time spent travelling. But they seem relatively indifferent to the labour inherent in the act of driving itself: many don’t perceive it as labour at all. Some even see driving as a leisure activity!
If a train fare costs £30 and petrol costs £25 but you don’t want to drive, what happens when you introduce a £25 or £30 taxi option to the relative attractiveness of taking the train
If petrol costs £25 and there is a £25 taxi option, you have just discovered a very generous taxi company that is willing to ignore the capital cost of the taxi, maintenance costs, road charges of whatever kind, and make a hefty loss on every journey. Even a nominal £30 is unlikely to be enough to cover these charges (most of which a self-driving car owner will just ignore in making a decision about which mode to use). And if the taxi drives at 125 mph to try to match the speed of the train, then the fuel cost will be substantially more than the £25, since energy consumption at such speeds is mainly due to wind resistance, which increases exponentially with speed.
And what happens when you halve the cost of being driven to Birmingham?
Given that it would cost several hundred pounds at a minimum to be driven to Birmingham, not a lot. Double or quadruple the number of people being driven and you have still made no difference to the overall transport market, especially as some would have transferred from self-driving anyway.
you’ll still have people able to summon their owned driverless car to take them on an intercity trip, or hire one via a driverless car from a rental company/car club/minicab firm (the distinction between the three will surely vanish once minicabs no longer have drivers). I just can’t see many people opting for a train any more for these journeys
If the train is faster (which is inevitable unless you believe we are going to see 300mph cars any time soon) and cheaper (because energy consumption per person for intercity rail is much lower, especially if we do see 300mph cars), why wouldn’t people catch the train? The implicit model of journey choice you seem to have in your head doesn’t seem to be able to account for the millions of people currently travelling by train – is it really just that they don’t like driving? Or is your model too simplistic?
@Harry Crayola investors and insurance markets are best placed to answer this question, balancing their cost efficiency with risks of legal liability
AKA the Ford Pinto approach. What amount of money will Google’s engineers attribute to a human life to get the cost efficiency balance right?
There are interesting parallels here to the history of railway safety in the nineteenth century, where the prevailing laissez fare ideology of the times broke down in the face of public outrage over the railway companies’ willingness to tolerate a high level of casualties, and the government eventually stepped in to impose regulation (mandatory block signalling etc). With accompanying complaints from the railway companies about progress-blocking, onerous government interference etc etc.
Ian J: I agree with most of your points, but just have a couple of niggles.
Wind resistance does not increase exponentially with speed, it increases broadly according to a power law – but what you probably mean is that it increases faster than linearly (which it does).
Most people are not interested in speed as such, but in the total journey duration. So to compete with a 300 mph train (which in any case we will not see any time soon, HS2 maximum is about 250 mph), a car does not need to go at 300 mph, since it is (often) door to door. Certainly the train is “faster” (in the sense of shorter journey time) for many journeys, but for others (Oxford to Hillingdon, say) it takes longer, and for yet others (Lynton to Barnstaple, for instance) it is not a feasible choice at all. I appreciate that you, and Harry, were talking about journeys where train or road are both feasible, and for such journeys I think your argument (that having or not having to drive makes little impact – the choice depends on time and cost) does indeed win out.
@KitGreen: Quite a revealing comment from a Volvo spokesman on that incident:
The Volvo XC60 comes with City Safety as a standard feature however this does not include the Pedestrian detection functionality
Pedestrian detection is available but it costs extra. The City Safety function that is designed to protect the car itself comes as standard.
@Malcolm: starting at the least scary end of the market
I think that the first driverless buses will be very small and slow and used in niche shuttle-bus type operations. In fact they already exist. I could see them catching on first in places like airports and theme parks, those petri dishes of transport-technologies-of-the-future (including the unmentionable one!).
Interesting discussion but I don’t recall any mention of what will power these vehicles (forgive me for not re-reading all the contributions – and so I may be wrong) but I assume they will be electric.
It seems to me unlikely that adequate infrastructure will be available in, say, 2025 in anything but large conurbations. And who will pay for this? A market for current forecourt fuel suppliers presumably?
Was the car Quinlet was in electric powered? Again, presumably.
Would the Courts currently have seen that driverless vehicle as being “under control” of the “driver” if it had had an accident?
@ Harry Crayola 0213 – We will not agree so I’m not going to respond in detail. One point I will make is “why might I be concerned about control”? I think that you take a more benevolent view of private enterprise and see it as being essentially “good”. I am afraid I am not quite so trusting and am concerned that the size of the potential prize (as you see it) will lead to that same private enterprise wanting monopoly market control, wanting legislative protection and limits to liability and wanting unacceptable (IMO) restrictions on common day to day activities that would complicate the operation of “their” technology.
I also suspect that we may see some extremely dubious “business transactions” whereby these companies quietly buy up and close down what would be their “old technology” competitors. Shades of what the oil, rubber and motor trade did to urban and inter urban transit lines in the USA. Call me sentimental but I view that as an appalling tragedy and a betrayal of past, current and future generations who’ve been lumbered with the material, environmental and social costs of those pernicious business deals. And I therefore suggest we leave it there.
@Ian J
Your critique of my waft-of-the-hand numbers is very good. I should have been more robust. But I’m afraid that I’m not much convinced by the underlying arguments.
“I said labour cost, not travel time cost. Certainly people want to reduce time spent travelling. But they seem relatively indifferent to the labour inherent in the act of driving itself”
You’re right to pick me up on the labour value versus time cost. But people don’t take the train because it’s cheaper than driving. They do so, broadly, because it’s more convenient, especially when travelling to a central area. Often because they are able to work on the train or to relax so that they can work when they arrive at their destination. Partly it’s because of the expense and inconvenience of parking. These are all advantages that driverless cars will eliminate.
“Even a nominal £30 is unlikely to be enough to cover these charges”
Yes. I should have used car hire plus petrol as comparator. So, I looked up off peak trains today (Friday) from London to Birmingham on National Rail. The cheapest day return ticket is £28 on London Midland taking 2h:04 from Euston to New Street. Alternatively, a Virgin service costs £52.50 and takes 1h:22.
By contrast, driving from Euston to New Street takes 2h:10 according to Google Maps. Hiring a ZipCar costs £5 per hour plus 29p per mile after the first 60 miles. At 126 miles each way, the journey would cost £80.68 (assuming two 2.5h hires at £12.50 each plus 29p times the 192 miles over the daily allowance. If we assume two sets of 60 mile allowances, the price would be £63.28.) Zipcar website says it is testing one way reservations but the page for more details isn’t available.
Anyway, this should give us a better idea for how the prices and journey times will compare. For the train, this is £28 per person at 2h:04 each way, plus costs and time to travel to and from the stations, or £52.20 per person at 1h:22 each way, plus cost and time of transit to the stations. For a car, £80.68 (or £63.28) at 2h:10 each way total cost (on average, assuming that additional and subtractory distance times/expenses net out) for up to four passengers. No changes of vehicle are necessary in the car journey, of course.
So, in summary, a driverless car with four seats will likely cost roughly same as one ticket on a Virgin train, and take about 40 minutes longer door-to-door than the Virgin train will take station-to-station. Or it will take roughly the same time door-to-door (again, with four seats) as a Midland Mainline train will take station-to-station, while costing about the same as two tickets.
Malcolm has addressed the fundamentals of the speed comparison well (although I obviously disagree with his conclusion), so I won’t repeat that. But it is also worth noting that a further saving is expected to arise from more efficient driving.
“why wouldn’t people catch the train? The implicit model of journey choice you seem to have in your head doesn’t seem to be able to account for the millions of people currently travelling by train – is it really just that they don’t like driving? Or is your model too simplistic?”
Yes, my explanation was not very strong. I hope the one above is a little clearer. I really don’t think there’s a bright future for rail journeys like London-Birmingham (although I don’t think things are quite so bleak for substantially longer journeys, where speed may well retain its relative advantage).
Please forgive my “Midland Mainline” mistake. It is late. I meant “London Midland”. Sorry.
@Walthamstow Writer
That is, I think, a very good overview of the difference which underlies a lot of what separates our takes on this. My only quibble is that I would add that I fully share your mistrust of corporate interests and their desire to hi-jack the law and gain monopolies etc. I am nervous about them trying it on again with jay-walking, for example. I am sure there will be other ruses. I would argue that I am happier to focus on stopping specific problems as and when they arise (and, as importantly, clearly signalling in advance that you won’t put up with any nonsense so they don’t try it on in the first place) rather than taking a more sceptical view of the whole development lest it lead to problems. Although that seems unduly flattering to me so I can only think I must have failed to describe it fairly!)
@Harry Crayola:
Your ZipCar comparison makes a fundamental error: the cost of fuel. You still have to pay your own petrol or diesel. Assuming your journey does not conveniently begin and end at the very edges of the two cities, this is likely to add up about £20-30 on top, so at least £100 for the trip.
While electric cars are very probably coming, there’s nothing inherent in vehicle automation that requires it, so it’s not a given that the two technologies must be in lockstep.
Another point you omit to mention is the quality of the journey. An automated vehicle designed primarily for low speeds in urban streets isn’t going to be very pleasant when belting along at 200+ miles per hour. It wouldn’t make sense to design a one-size-fits-all vehicle when the vast majority of journeys will be much shorter. (I’d also like to point out that it’s rather harder to fit toilets and a bar into vehicles that small.)
Much will also depend on how the technology is implemented, not only in terms of engineering, but at the business and political levels. Do we go for railway-style concessions or franchising, or allow a bus-style free-for-all with multiple competing companies running their own fleets of vehicles through the same areas?
This is why I see the emphasis on driverless cars as rather missing the point. That specific use-case is just one of many, and isn’t even all that important when it comes to addressing London’s future transport needs. There more interesting possibilities happen when you look at applying vehicle automation technology to other modes, and even to entire industry sectors. (E.g. postal services, and other forms of delivery.)
To paraphrase a certain industry writer: we must first work out what driverless vehicles are for. How can they help London? What possibilities do they open up that would be impossible without them? Only after these (and other) basic questions have been answered can the hard decisions on business models and regulations be made.
How we transition to new transport technologies is mostly a political issue, not an engineering one.
@Malcolm: 300 mph train
Sorry, I meant 300 km/h. And yes I was using ‘exponentially’ loosely and shouldn’t have done. I would note though that at a given speed, one long vehicle (a train) will have less wind resistance and so use less energy per unit of length than multiple small vehicles (cars). And that is before you get onto the greater rolling resistance of rubber tyres on concrete vs steel wheel on steel rail.
@Harry Crayola: They do so, broadly, because it’s more convenient, especially when travelling to a central area. Often because they are able to work on the train or to relax so that they can work when they arrive at their destination. Partly it’s because of the expense and inconvenience of parking
I do think that you are missing out on one very important factor – they do it because it is quicker.
So, in summary, a driverless car with four seats will likely cost roughly same as one ticket on a Virgin train
I think I’m missing an important part of your reasoning. You seem to be saying that a non-self-driving sharecar currently costs about 40% more than a Virgin train for one person, and so obviously less than the train per person for multiple people. But then you make a jump to saying that a self-driving sharecar will somehow be cheaper, sufficiently cheaper to empty the WCML, just because it is self-driving? Why?
And if it is currently cheaper and, by your reckoning, more convenient for any group of two or more people to drive from London to Birmingham, why do so many people catch the train now? By your reasoning the WCML should already be hardly used. Are Virgin’s trains full of people who pay through the nose for peak tickets just because they don’t like driving?
“Often because they are able to work on the train or to relax so that they can work when they arrive at their destination.”
This is only really possible on inter-city services, or on the relatively small number of regional commuter services where everyone gets a seat and access to a table. For most commuters, rail travel during the peaks involves standing for much of the journey. “Relaxing” isn’t quite the term I’d use to describe it.
Another point worth noting: not everyone can work on the move. Many, like myself, suffer from travel sickness if we do anything more complicated than staring out the window. I can’t even read a book while on the move, let alone muck about on a laptop or tablet, and I know I’m far from alone in this.
Driverless vehicles are also inherently less efficient than mass transit systems when it comes to shifting lots of stuff – including people – quickly from A to B. Not even the biggest driverless long-distance coach can shift as many people as quickly and efficiently as a modern train can.
The only rail-based transport mode that is likely to suffer is the tram, but even that’s not a given.
@Malcolm 2323
There are standard categories for the degree of automation for vehicles ranging from 1 to 5. 1 is basically some automated assistance, such as cruise control, but which requires the driver to be actively engaged at all times. 2 and 3 provide higher levels of automation which mean that the driver does not have to be engaged actively at all times but must still be fully in control and able to take over at all times. Park assist is an example and ‘creep control’ (which automatically drives a car in a slow moving line of traffic, and which will be on the market in the next year or so) is another. Category 4 is where the vehicle will drive itself at all times but does require the driver to be present, focussed an able to take over immediately if needed. Category 5 does not need the driver at all. These are all what the vehicle is capable of and not what is actually happening at the time, but clearly as you get to higher categories the option of resorting to simple manual driving starts to diminish. The equivalent on rail might be DLR trains, which can be driven manually but are not suitable for routine manual driving.
While I was in it the driver watched but did not intervene during the demonstration and yes, 0775John, it was electric.
I agree with IanJ that places like airports and theme parks are likely to be amongst the early adopters of the technology, if for no other reason than that they use private roads and not public roads which are under their complete control. A much easier place to start this type of activity.
Nameless ( & everybody else)
At present, Google are doing very well – they are IIRC saying that “we” (Google) are liable & we are doing our best to make sure nothing nasty happens … so far, apart from one bus-driver maybe doing something legal but stupid, every single “accident” to a G-car has been the humans’ fault, not G’s.
However & contrariwise, this is at 25-mph or under speeds in built-up, intensively-surveyed areas.
Enlarging this to cover whole cities & higher speeds is actually a step-change, not just “more of the same” – so far at any rate.
WW
Indeed
“Who killed Roger Rabbit?”
Anomnibus
we must first work out what driverless vehicles are for. How can they help London?
Oh dear, I do wish you hadn’t asked that, because there’s an obvious answer.
They are for those under 12, over 80, pregnant, have back pain or other “mobile” problems, the disabled generally & those with defective eyesight.
In short all of those who cannot cycle, even if they wanted to.
quinlet – I understood there was no option to override the driverless function at cat 5?
If there was it would torpedo any benefits as a small minority would certainly use it.
@Anomnibus
From what I can see on the website Zipcar costs include fuel. All cars come with a corporate fueling card in case you run out. Most car conventional car hire companies though charge a hefty one-way fee – I wonder what Zipcar’s will be.
@Anomnibus
“Your ZipCar comparison makes a fundamental error: the cost of fuel. You still have to pay your own petrol or diesel. Assuming your journey does not conveniently begin and end at the very edges of the two cities, this is likely to add up about £20-30 on top, so at least £100 for the trip.”
You’re clearly not a ZipCar user! There is no fuel to buy. You transact fuel purchases with a fuel card they give you at petrol stations. That’s what the 29p per mile charge is there to cover. You’re double-counting.
“we must first work out what driverless vehicles are for. How can they help London? What possibilities do they open up that would be impossible without them? Only after these (and other) basic questions have been answered can the hard decisions on business models and regulations be made. “
We don’t have to work this out at all. This is something that investors and managers touting for their investment need to work out. And they have. Basically: driverless cars will combine the convenience and accessibility of taxis (at least for the able bodied, whether drivers or not) with the flexibility and low costs of car rental and/or ownership. Indeed, they’ll go even further than that by reducing costs even more through improved fuel efficiency and lower insurance costs and they’ll make journeys faster. They’ll do this for two reasons:
1. Their reaction times mean will be able to travel much faster on normal stretches of road at the same risk level.
2. Junction communication systems will mean they may be able to safely pass through junctions much more quickly than we can now (substantially reducing congestion). See this video explaining how an automous intersection control protocol could work.
All we need to do is make sure we’re not getting in the way with laws designed for another age (and, indeed, to ensure that we don’t let that argument be used to introduce new regulations under its guise such as jay-walking laws or whatever the driverless cars equivalent may turn out to be!)
“if it is currently cheaper and, by your reckoning, more convenient for any group of two or more people to drive from London to Birmingham, why do so many people catch the train now? By your reasoning the WCML should already be hardly used.”
Driving often isn’t more convenient now, though. That’s my point. Now, trains are often more convenient. What if I can’t drive? What if I left my car at home? Sure, some hire car companies will deliver a hire car to your location for a premium fee, but that’s just swapping the hassle cost for financial cost. What if I don’t own a car? What if I need to use the time to read something or work? A taxi would be wonderful and answer all these questions, but they’re extremely inefficient due to the need for a driver.
“Driverless vehicles are also inherently less efficient than mass transit systems when it comes to shifting lots of stuff – including people – quickly from A to B.”
Firstly, nobody actually wants to travel from the A to the B that railway companies can offer. Who just happens to live opposite Euston and happens to have a meeting opposite New Street?
Second, how are they more efficient? Not environmentally. A small car with two passengers emits more co2 than the train (see here). Not financially, either (see my comment at 02:03 for the personal finance aspect and compare rail subsidies versus fuel duty receipts less road maintenance costs for the fiscal aspect). Nor, with widespread adoption of driverless cars, will trains any longer retain much efficiency advantage on land use (rural land for motorways/railways is very low value to start with and motorways are likely to become far, far more space efficient when driverless-only).
There is a future for both public transport and rail, but it’s not what we have had now since the spread of mass motoring. I am confident that driverless cars will represent another such fundamental restructuring of consumer demand in transport. I find it astonishing that so few people (outside tech circles) seem to have woken up to just how fundamental its impact will be. It reminds me of this 1995 Newsweek article on why the internet will just be a passing fad.
Rail will continue to be needed for dense urban areas and perhaps journeys of over 200 miles or so, depending on how fast driverless cars will be able to efficiently travel. Otherwise, I expect some (driverless) bus services will remain for those willing to share and travel to and wait at stops, and also minibus UberPool style oeprations as a half-way house between taxis and buses.
@Ian J
I’m sorry (to both you and Anomnibus), I ineptly quoted from you in my reply to him. I have copied it below for ease of reference. Apologies to the moderators for my clutter, too!
“if it is currently cheaper and, by your reckoning, more convenient for any group of two or more people to drive from London to Birmingham, why do so many people catch the train now? By your reasoning the WCML should already be hardly used.”
Driving often isn’t more convenient now, though. That’s my point. Now, trains are often more convenient. What if I can’t drive? What if I left my car at home? Sure, some hire car companies will deliver a hire car to your location for a premium fee, but that’s just swapping the hassle cost for financial cost. What if I don’t own a car? What if I need to use the time to read something or work? A taxi would be wonderful and answer all these questions, but they’re extremely inefficient due to the need for a driver.
@Greg
“[driverless vehicles are for] those under 12, over 80, pregnant, have back pain or other “mobile” problems, the disabled generally & those with defective eyesight.”
Most of those categories need assistance at both ends of the journey, if not to get in and out of the vehicle, but to get from it to the premises they are actually travelling to.
Would you seriously let an under-12 travel unsupervised in a driverless car to a remote destination?
As for pregnancy precluding cycling, my mother kept cycling through most of my own gestation, and those of my four siblings.
@GT
What about accidents in which the Cat 5 car is not actually involved but still caused by its actions? eg cyclist colliding with pedestrian after being forced to swerve by Cat 5.
@Harry Crayola
Just a few very simple questions about the present state of development:
How precisely do Cat 5 cars obey the instructions of a lollipop person outside a school?
Can they recognise Stop/Go boards at a temporary road works?
Or road signs like “give way to traffic approaching in opposite direction”?
Can they cope with traffic at a double mini roundabout?
How do they react to birds flying in front – can they distinguish between different animal hazards?
If involved in a collision, how do they provide details of ownership and insurance?
Do they know that they must leave at least a 1.8m gap when passing cyclists?
If the answer to any of the above is “no”, when will such capability become available?
I’m more than a little sceptical about the timescales under discussion.
@Nameless
How precisely do Cat 5 cars obey the instructions of a lollipop person outside a school?
I don’t know, I’m no expert. Perhaps you could try a supplier?
Can they recognise Stop/Go boards at a temporary road works?
Or road signs like “give way to traffic approaching in opposite direction”?
Don’t know, but don’t see how this would be a problem with suitable standards.
Can they cope with traffic at a double mini roundabout?
Don’t know, but I can’t think of any realistic reason why they wouldn’t be able to.
“How do they react to birds flying in front – can they distinguish between different animal hazards?”
Not sure. But something makes me suspect that you haven’t found an obstacle that nobody in the industry has thought of.
If involved in a collision, how do they provide details of ownership and insurance?
Don’t know. But such a trivial matter would be quite simple. I’m sure communication protocols wouldn’t find this insurmountable. Plus I don’t see why they wouldn’t have registration plates in the event of communication failures.
“Do they know that they must leave at least a 1.8m gap when passing cyclists?”
That depends on whether the software has included that information.
If the answer to any of the above is “no”, when will such capability become available?
When it is required.
I’m more than a little sceptical about the timescales under discussion.
Why? What is the basis for the timescales that you think are likely?
Could all commentators please limit comments to a few paragraphs at a time, and split responses to individual commentators in separate comments?
This will ensure that your comments are read, and that the commentator replied to can find your response.
Thank you
LBM & the Mods
@quinlet
Are the categories you mention from the US National Highway Traffic Safety Administration (NHTSA), which defines vehicle automation according to five levels?:
No-Automation (Level 0): The driver is in complete and sole control of the primary vehicle controls – brake, steering, throttle, and motive power – at all times.
Function-specific Automation (Level 1): Automation at this level involves one or more specific control functions. Examples include electronic stability control or pre-charged brakes, where the vehicle automatically assists with braking to enable the driver to regain control of the vehicle or stop faster than possible by acting alone.
Combined Function Automation (Level 2): This level involves automation of at least two primary control functions designed to work in unison to relieve the driver of control of those functions. An example of combined functions enabling a Level 2 system is adaptive cruise control in combination with lane centering.
Limited Self-Driving Automation (Level 3): Vehicles at this level of automation enable the driver to cede full control of all safety-critical functions under certain traffic or environmental conditions and in those conditions to rely heavily on the vehicle to monitor for changes in those conditions requiring transition back to driver control. The driver is expected to be available for occasional control, but with sufficiently comfortable transition time. The Google car is an example of limited self-driving automation.
Full Self-Driving Automation (Level 4): The vehicle is designed to perform all safety-critical driving functions and monitor roadway conditions for an entire trip. Such a design anticipates that the driver will provide destination or navigation input, but is not expected to be available for control at any time during the trip. This includes both occupied and unoccupied vehicles.
@HC
Fair enough, you don’t know if solutions to any of the above are currently available.
You assume that the technology will be ready “When it is required”. My point is that all of these matters and a vast number of others applicable to the safe and fair interface between self driving vehicles and the rest of the road using population *must* be addressed *before* the new vehicles can be allowed to operate completely without drivers on the shared road network.
Incidentally, the points regarding lollipop persons and double mini roundabouts are to demonstrate example situations where drivers’ conduct is reliant upon eye to eye communication and understanding with others.
Complete safe road sharing must be decades away.
LBM & the Mods?
Do you have any old photos of the band wearing parkas or on your Lambrettas?
@Nameless: I think you might be wrong, I would say less than 10…. Given how technology has evolved over the last 10 years and how far we already are down the track, I would say we’re closer than you think.
Crystal ball going back under wraps now, I think I saw an unmentionable extension starting to appear in one corner.
I suspect that self-driving cars will change they way we use road vehicles to such an extent that it’s extremely difficult to make even half reasonable predictions about their impact. It strikes me that parking might be transformed. If only pool cars were allowed in a car park, then they could just fill the entire space leaving no open roadways. If a call was made for one, the last one in would respond and go on its journey. Underground car-parks are not very popular here, unlike in Germany where they are common, freeing up a lot of surface space. If few humans were required to enter them, then they could be built smaller but retain the same capacity, so might be seen as a much sensible use of expensive land than either surface car parks or multi-storey ones.
The inherent problem with the door to door idea in inner cities is that there would be a lot of empty cars (having just dropped their loads) on the roads, causing as much anxt as cruising mini-cabs and black cabs do now. However clever they are at using junctions, in a dense city they will always be obstructing some one else’s journey or road space.
@harry Crayola
“If involved in a collision, how do they provide details of ownership and insurance?
I’m sure communication protocols wouldn’t find this insurmountable. Plus I don’t see why they wouldn’t have registration plates in the event of communication failures.”
And if a collision between two unoccupied vehicles knocks the comms out in both vehicles?
@Long Branch Mike – is there a Fully Self-Aware (Level 6)?
Terminator fans will get the reference…
I saw a cartoon recently that went along the lines that in the future, intelligent driverless cars will imprison the human race in metal cages but it will take humans some time to realise this because they will think it is just really bad traffic…
@LBM
I have come across them as levels 1-5 but apart from that I have abbreviated.
Reynolds 953
The SF author C Stross has already done this.
I recommend: “Rule 34” by said author.
Nameless says “My point is that all of these matters and […] *must* be addressed *before* the new vehicles can be allowed to operate completely without drivers on the shared road network.”
This does sound a little like not allowing anyone in the water until they can swim.
Clearly, while some of these matters are still not entirely worked out (which is probably the case today, though I cannot be certain) extra precautions have to be taken. These may well include the on-board qualified and attentive driver. They may also include extra-conservative program settings of various kinds, special markings for the cars, more restrictive speed limits, etc etc. But unless we all set our collective faces against driverless cars (which would seem Luddite), then carefully planned risks must be taken.
Suggesting that such advances can be entirely risk-free is a mistake. But suggesting that nothing except provenly risk-free experiments should be allowed would also be a mistake in the other direction. The risk must of course be properly evaluated, managed and minimised.
Nameless refers to eye-contact between drivers and others.
I agree that this is important, and a suitable substitute should be sought. Could a suitable dummy visage able to nod and wink be added to the front of the vehicle somewhere? Car-y McCarFace?
One surprising omission, particularly given Reynolds 953’s interests, is the reference to interface with cyclists who will, presumably, remain Man v1.o for the foreseeable future.
quinlet/nameless
I think you will find that technology is further advanceed than you imagine. My current car, delivered January 2016, is fitted with ‘creep control’ and very good and stress reducing it is too. It also is fitted with cameras and interpretation software to recognise roadside and overhead gantry signs (speed limit, no overtaking, give way, etc) and again it is accurate and consistent in recognition capability.
Person v1.0 ? Although my wife was pleased with her upgrade to v1.1 on a recent bike-hire (comprising electric pedalling assistance).
Providing details of ownership and insurance after a collision is still legally required, I believe. (And certainly driving away without speaking to the others involved is, rightly, an offence). But since these details, these days, can be fairly readily determined from central databases using only the registration number (at least by authorised persons), one wonders how much longer this will remain the case.
timbeau says “Most of those categories need assistance at both ends of the journey”.
I submit that this is frequently not the case. Many completely blind people, and almost all partially-sighted people, normally go about much of their business without a dedicated companion. And yes, I would let a child, old enough to walk or cycle to school independently (which includes most 11-year-olds), use a self-driving car instead. (Naturally it would depend on the child and on the neighbourhood and other circumstances, but what I mean is that I would certainly not rule it out just because they were 11 years old).
Will the authorties be able to take command of a driverless car remotely?
How do you stop someone filling half a dozen with explosives and telling them to drive themselves to targets?
@Malcolm
I did say “most” of these categories.
timbeau: the explosives question sounds tricky, but doing the same thing with a human-driven taxi is already feasible. I doubt if the villain will care much one way or the other about one more or one fewer (taxi-driver) death. The techniques (if any) which taxi-drivers already use to detect and avoid this risk could also be adopted by self-driving cars. The same applies to off-vehicle methods (such as police surveillance).
Malcolm 18.16
You refer to the taking of “carefully planned risks” in the development and introduction of driverless cars.
I am asking in all seriousness since I don’t know, but do we approve of this mind-set in the air industry and in the development of rail signalling etc. “As low as practicable” over a century + of experience seems reasonable but maybe not in new technology such as this.
The nuclear industry may have something to tell about how early attitudes to risk were not always up to the required standards and, whilst the scale of risk in driverless cars is minuscule in comparison, I think we should be careful not to be cavalier about it.
0775john: I am not sure what you are asking. Certainly hindsight tells us that risks which we now consider unacceptable were taken early on in the nuclear industry. But they were not judged unacceptable at the time. We must use the best available information, but without access to a time-machine, sometimes this will be wrong.
I quite agree that we should not be cavalier about risks. But demanding that everything be completely risk-free is quite unrealistic. We all take a risk every time we leave the house; but we do so because we want the benefits of leaving the house, and we judge that those benefits outweigh the risk.
@Malcolm
You don’t need to make any threats, or promises, to get a robot to do your bidding. And one fanatic could control several at once, and survive the experience.
If the police have an over-ride switch it’s a different story, (that might also solve the “how does a robot car cope with temporary diversions, emergency service vehicles, etc” questions) but if they have that capability we then get into interesting civil liberties territory. (Imagine if the police had had the capability to control the vehicles at the infamous “Battle of the Beanfield”, or pickets heading for Orgreave.)
@ Malcolm – while I understand the basic point you are making I think you’re being a tad simplistic. I am very doubtful that anyone really makes any real assessment of the risks of leaving the house. (I suspect there is only a tiny element of the population that has ever had any risk training or had to do a risk assessment or been held responsible in their jobs for identified risks. I have.) We might think about putting a coat on if it is cold or taking a brolly if the forecast is for rain but no one rationally considers all the possibilities of trips, slips, falls, heart attacks, strokes, road accidents, lightning strikes, being pecked to death by an errant psycho bird etc etc. All people do is expect to repeat a previous experience of doing the same thing safely and taking simple precautions like crossing the road safely or driving their car while paying what they consider is sufficient attention. Almost all of the time the above suffices but, of course, sometimes life takes an unexpected turn….
It would be entirely and utterly inappropriate for an industrial supplier of driverless vehicles and the supporting control infrastructure to adopt anything like the same approach as the average person. They are *required* to properly identify risks and take the appropriate steps to remove them or else get them down to extremely low levels. Given the lamentable safety performance of private road transport then there will be a great expectation on the part of the public that that current state of affairs is improved upon. Clearly part of the problem is that human drivers make errors or misjudge situations and their skill levels to respond to them. Sometimes entirely unexpected things happen to cause accidents. Now the new technology *might* well improve on human fallibilities but a vast amount of work, testing, evaluation and assurance will need to be provided. Not unlike what a new or modified railway has to do in order to demonstrate it can safely transport its passengers. It will be heartening to see technology suppliers and vehicle manufacturers start to face some of the same rigour that modern railway networks (and their suppliers) face. It will also be interesting to see how public opinion shapes what the suppliers are required to do and if the demands are more or less rigorous than apply to the railways.
WW. I would submit, m’lud, that the people you describe in your 2nd paragraph (professionals used to risk assessment and risk reduction) act as ordinary people (1st paragraph) when not at work, and do all the mad things ordinary people do (cross the road on the red man, ski, climb mountains etc.).
timbeau
If the police have an over-ride switch it’s a different story
And, if someone else also has an over-ride switch?
It’s the usual security “backdoor” problem. If it exists, it’s hackable.
This was the exact problem Mr Stross addressed in his novel that I mentioned.
WW: I agree with most of what you say. Including the word “simplistic”, if it is intended to mean that there are many ramifications of the “risk” issue into which I did not go; for various reasons, including staying somewhere vaguely related to the original topic, also I am no expert and have little experience of formal risk assessment.
I think your point of bringing into road transport much of the formal stuff that railways are already obliged to do is a good one. (Though it is possible, but by no means certain, that car developers are already up to speed on such matters, at least in some countries – if they were then it might not necessarily be made public).
Your mention of human driver errors is a reminder that the possible shortcomings of computers as drivers (doubtless numerous) may well (in due course) be wholly or partly offset by certain definite advantages – less easily distracted, no fatigue, no wish (even in young male computers) to show off to their peers, etc etc.
It will be extremely important that your car will not be hackable. The potential for chaos from rogue states persuing aysmetric warfare is to great. Both Russia and China pay vast number of patriotic hackers to infest the global net and already perform industrial scale corporate and government hacking. Many companies now insist that travelling executives use disposable phones and laptops on visiting certain countries.
Russian has already used cyberwarfare against countries who offend it politically, I think it was Estonia that ground to a halt for a couple of weeks with the closure of most banks and telecoms when the offended Russia by pulling down a war heroes statue.
I’m sure the police would love to lock you in your car or have you delivered direct to the Nick but unless that can comply with safety rules I can’t see it happening, besides they won’t be any traffic offences in the future so there are no safety requirements to stop cars, Besides I’m sure they’d be happy just knowing where your car is second of the day.
timbeau and Greg: I agree that the whole matter of backdoor access to driving software is fraught with difficulties. (Not just by the police and by villains, but also by garages and motor manufacturers and highway authorities).
The phrase “the authorities” is a sure warning of difficulties for anyone who has moved on from the reassuring world of “goodies” and “baddies”. The colour of the hat is no longer any guide.
Rational plan: “there won’t be any traffic offences in the future so there are no safety requirements to stop cars”
Cars are often stopped for reasons other than suspected traffic offences. (Including having the wrong colour skin for the make of car, but also plenty of reasons which we might find rather more acceptable).
Just stepping back from the detailed debates about the how and the what if of driverless cars for a moment, what strikes me is that it is assumed that driverless cars are a Good Thing (or perhaps an inevitable thing, which is not quite the same,of course). Reading the detailed comments suggests, however, that the benefits are uncertain, the risks are unclear, and the costs may be quite high. The main beneficiaries would seem to be the manufacturers, applauded by a claque of technophiliacs, but whether it is actually in the ordinary public’s interest is not stated….
Interesting video about automation, including driverless cars – https://www.youtube.com/watch?v=7Pq-S557XQU
The point this video makes is that driverless cars don’t need to be completely safe, only safer than human drivers. I am not sure that this is correct as new technlogies are generally required to be a lot safer than their predecessors. If manufacturers try to sell driverless cards on the safety benefits (which they will), then they will need to be a *lot* safer than human drivers.
As an IT Engineer, I also have concerns about the security of these types of vehicles, not only from external sources, but also from the owner. How do you ensure someone doesn’t open up the “bonnet” and modify the code to make the vehicle go faster or be more aggressive in traffic ? Google aren’t (yet) terribly good at managing this sort of thing.
Security is likely to mean that a large swathe of users are unlikely to adopt diverless cards – eg. Police, security services, diplomats and so on. Whilst you could argue that human drivers can also be “hacked”, at least these users already have controls in place to manage that.
Graham: We live in a capitalist society (other forms of organisation are available) where it is not considered necessary to choose between benefits for the promoter of any given innovation, and benefits for society as a whole. These two outcomes are presumed to be mutually compatible.
Of course, if no-one except the manufacturers gains, then there is no problem, there will be no take-up. If some of the public benefit, but not all, and the benefits/disbenefits are considered to be “unfairly” distributed, that is another matter, but inequality of this type is, sadly, so widespread as to be entirely unremarkable.
@Malcolm – oh I agree entirely*; my point is simply that the advantages seem unclear for any party (possibly not even the manufacturers of vehicles and software, unless they are assuming that a wave of the legislative hand will wipe away all the more obvious institutional pitfalls).
__________________________________________________________________
* Indeed, I expect closely controlled societies to embrace the technology much more fervently than capitalistic ones
Graham: I would like to deconstruct your “the advantages seem unclear”.
The possible advantages seem to me to be perfectly clear. The main ones would be safer easier travel for many people, with perhaps also resource saving, less congestion and cheaper travel.
What is unclear, though, is how soon, if ever, these advantages can be achieved, and at what price. That’s where the conjectures, risk-taking, entrepreneurship and other ways of dealing with uncertainty start to come in.
@Graham H – indeed. Certainly, I can’t see many advantages for existing car manufacturers. They will be looking to cram new widgets into their cars to persuade people to buy the latest model but that isn’t the same as trying to to produce a driverless car.
The situation is different for new entrants who will be looking at a a new class of product to grab market share at the expense of existing manufacturers .
Perhaps analogous to what smartphones from Apple and Google did to manufacturers like Nokia and Motorola.
A rather extreme (?) view against driverless cars:
It will be a tag-team of the government and the car companies who control (and thereby, effectively own) “your” car.
http://www.zerohedge.com/news/2016-04-29/why-hard-sell-self-driving-car
@Malcolm – yes to “possible” -a matter of faith (and/or commercial hype).
The safety benefits alone are a considerable gain. Then there is mobility gain for older people who don’t have to rely on buses. if we can ever sort out drink driving rules (maybe you have to press a lock out option in the car and say you have consumed alcohol to be allowed to drink and drive) then pubs restaurants could see a increase in business.
More young people could own/lease a car due too much lower insurance premiums and their would be far fewer uninsured drivers on the road.
All important benefits but not essentially anything that reorders society.
Bigger changes require driverless vehicles to be much cheaper to run than existing manned vehicles.
If an auto cab is 30% cheaper, well that’s nice but it’s not going to change the world just keep wages down on cab driving.
If on the other hand its 70% cheaper then we will see much bigger effects. Autocabs will have to be card only, unless you want to add the complexity and cost of ATM pay in machine into the vehicle. Smart card technology will need to be expanded to low credit score part of society and those who are still unbanked.
As to vehicle stopping the police can still wave someone down and the car will still need an emergency stop button and a method to interrupt a trip to pull over or cange destination quickly.
@rational Plan – yes, some possible benefits and some possible disbenefits (slower traffic speeds, unclear or limited liability, etc) ; my point is that there is no clearcut balance of advantage either way (and quite possibly no easy way of assessing that balance), but nevertheless, we are told these things will be coming our way.
I think the problem here is that some people have the view that within a few years, all cars will be replaced by cheap, automated, shared-use pods, run by some benevolent government organisation, solving all the congestion and safety problems.
What is more likely to happen is that there will be gradual progression of technology in cars, which will lead to driverless functionality in limited circumstances (eg. motorways) and eventually under all conditions. These features will appear in high-end cars initially and after economies of scale kick in, will filter down to all vehicles.
As these sorts of the vehicles appear, they will gradually be used to replace existing taxi’s in cities, probably in a similar fashion to Uber but without the drivers. As the cost of buying and running these vehicles reduce, they will eventually become cheaper than vehicles with drivers, and will start taking over, but it will be many years before they become significantly cheaper than todays taxi’s.
How quickly this all happens is unknowable, but it will be a lot more than the couple of years some people assume. Automated parking systems first appeared commercially in 1999 in a Lexus and today are still considered a relatively high-end feature. The majority of cars on the road do not have it, because cars are a long term purchase. For driverless systems, we are in the mid-1990’s, so it could be 20 years before it becomes widely available.
@Harry Crayola:
“You’re clearly not a ZipCar user! There is no fuel to buy. You transact fuel purchases with a fuel card they give you at petrol stations. That’s what the 29p per mile charge is there to cover. You’re double-counting.”
Okay, I wasn’t aware the fees included fuel. That’s good to know. There’s a similar service in Italy, but it’s nowhere near as widespread, and has a rather more complicated pricing system.
However…
ZipCar’s website says they also charge you for the time you spend at your destination, doing whatever it was you went there for. By the hour. Trains don’t do that, and neither do taxis unless you explicitly ask them to stick around. Every hour you spend there is costing you money. (On top of whatever parking fees you have to pay as well, though that’s arguably not relevant to a discussion on driverless cars.)
If you’re only in Birmingham for a quick one-hour meeting, it’s not a deal-breaker, but if you’re there for the day, it’s going to add up quickly, and will soon overtake the rail option. This is, however, an argument in favour of shared-ownership driverless vehicles (which I shall henceforth refer to as “SODVs” as I’m not yet used to typing on my new keyboard and that’s a bit of a mouthful).
While some restricted-user driverless vehicles (“RUDVs”) may be needed for certain edge cases, such as on-call doctors and emergency services, I predict the SODV model will be the most prevalent as this means, once you get out of the car in Birmingham, it doesn’t have to stick around and wait for you to come back. The down-side of this is that both London and Birmingham would need to be using interoperable systems and vehicles, which might not necessarily be a given during the early phases of introduction.
“See this video explaining how an automous intersection control protocol could work.”
I already have, thanks. It has a few issues, not the least of which is that it cannot possibly work with pedestrians and cyclists.
There are very few places in the whole of Greater London where that approach could possibly work at all. The only one that springs to mind is the Kidbrooke Interchange on the A2, which is the only one on that road inside the M25 that’s a flat junction with traffic lights, rather than grade-separated.
There are a number of other issues, but I suspect this isn’t the forum for it. (E.g. Several million vehicles, all trying to communicate with both each other and a control center of some kind, is probably the hardest part of all this. That’s a hell of a lot of bandwidth you’ll need to allocate.)
@Jim Cobb, et al:
Now consider the Shared Ownership Driverless Vehicle (SODV) model:
Instead of spending most of their lives switched off, each vehicle is in use almost all day long. You can “sweat” these assets the same way bus and train operators do, by keeping them working and earning their owners money for as much as possible. Granted, peak hours means there’ll inevitably be more cars available during off-peak hours than are needed, but train and bus operators deal with this too. You schedule your maintenance, cleaning, refuelling / charging (for electric vehicles), etc. for those quieter hours.
[Superfluous paragraphs not germaine to London transport snipped. LBM]
Re. Command & Control…
If the Shared Ownership model is adopted, as I suspect it (mostly) will be … [sorry, too far into the realms of speculation and whimsy, and we would also like others to get a look in. Malcolm]
@ Anomnibus – another factor the manufacturers must be scared about is the trend in many developed countries for a reduction in the level of car ownership amongst younger generations and an increase in the age that people actually get a driving license.
In London, every age group except the over 60s has a lower rate of car ownership than the same age group ten years ago. The increase in ownership with the over 60s isn’t enough to offset the decline in younger age groups as the ownership rate amongst over 60s is less than younger age groups anyway.
There is no sign of this trend reversing, so as each age cohort continues to own fewer cars than older cohorts, the overall market size will decline in certain markets.
@Anomnibus et al
Note that concerns about car manufacturers’ market share is far from the scope of London Reconnections, and we may well snip such musings henceforth.
@Harry Crayola, 29 April 2016 at 11:18
“Junction communication systems will mean they may be able to safely pass through junctions much more quickly than we can now (substantially reducing congestion). See this video explaining how an automous intersection control protocol could work.”
Looks an awful lot like an signalling interlocking with ARS ‘setting routes’ for individual vehicles to me, and an enormously complex one at that! If railway safety standards were to apply, all those processing and communications systems would have to be engineered to deliver a safety integrity level of SIL4 (https://en.wikipedia.org/wiki/Safety_integrity_level), and to be acceptably reliable would have to deliver extraordinarily high levels of dependability for every subsystem and component.
I much preferred the next video that YouTube suggested, featuring something called a continuous flow interchange as developed by the Utah Department of Transportation. That concept works fine for non-automated traffic as well, a vital consideration for migration to any notional totally automated future.
CFI Tutorial – Utah Department of Transportation
https://www.youtube.com/watch?v=_dB25GPPdeU
Continuous Flow Intersections (CFI) are currently in use at several locations in Salt Lake County and are planned for other areas throughout the state. This tutorial was developed to help drivers learn how to use these intersections and their various configurations.
@Jim Cobb
The other likely early adopter (apart from taxis) will be fleets and goods. In fact anywhere where the driver is currently paid but is used only for driving the vehicle. The additional cost of driverless vehicles in these cases will be offset by a reduction in salary costs. In each case the fleet owner has an incentive to keep the vehicle operating for as much of the day as possible.
CFI is unlikely to be much use in London (in my view) because of the extra road space which would be required for the solid barriers. It is also a digression from a digression. (But I cannot simultaneously comment on it and request others not to, so I am not so requesting. Please use your own discretion).
Quinlet: Unless the vehicle is empty, the driver is also responsible for the load (or the passengers). Granted physical loading and unloading is sometimes done by others, but the driver must still be responsible for seeing that the right goods go to the right place, or, if collected, are documented and safely stowed. For passengers, the driver is responsible for their well-being.
Other arrangements will doubtless be possible for self-driving vehicles in at least some situations, but their introduction will be an extra complication.
@Anomnibus
This is, however, an argument in favour of shared-ownership driverless vehicles
Yes. I was counting on that being obvious given my previous comments and the facts you describe and so, given my already long comment, I decided not to mention it and hope it would be assumed.
There are very few places in the whole of Greater London where that approach could possibly work at all.
Agree. Good point. But driverless cars’ mere faster reaction times and safer bunching alone should make for a significant improvement in junction capacity.
Several million vehicles, all trying to communicate with both each other and a control center of some kind, is probably the hardest part of all this. That’s a hell of a lot of bandwidth you’ll need to allocate.
We’ll be on 5G tech at least by then. But I’m not sure why you need them all trying to communicate with each other, or indeed any control centre. Why do you need a control centre? They’re being designed to be autonomous. They only need to be able to communicate with cars near enough to them that they may come into conflict.
@KitGreen
A rather extreme (?) view against driverless cars
I have to admit that I share some of ZH’s cynicism towards the intentions of the firms and how governments will try to hi-jack the technology to compel us to obey their own obsessions. For our own good, of course. But, only some. If we don’t put up with that stuff, it won’t happen.
@Mark Townend
I much preferred the next video that YouTube suggested, featuring something called a continuous flow interchange as developed by the Utah Department of Transportation.
Thank you. This is interesting.
@Jim Cobb
What is more likely to happen is that there will be gradual progression of technology in cars, which will lead to driverless functionality in limited circumstances (eg. motorways) and eventually under all conditions. These features will appear in high-end cars initially and after economies of scale kick in, will filter down to all vehicles.
Yes! Although note how much faster adoption of new technologies is becoming as time moves on. I suspect that once driverless cars are accepted as driving people as comfortably and quickly as manual cars, their mass market place will come very quickly. Why drive yourself when you can hire a chauffeur in your own car or take a taxi at almost no cost?
As the cost of buying and running these vehicles reduce, they will eventually become cheaper than vehicles with drivers, and will start taking over, but it will be many years before they become significantly cheaper than today’s taxis.
I’m not sure I agree with this. I think taxis are the place where they’ll make the first substantial impact. The cost of the tech for driverlessness merely needs to fall below that of the driver’s employment costs in the expected lifespan of the vehicle. And this might well be multiple individual drivers’ employment costs, too, not just one. After all, how many drivers would you need to operate a vehicle 24/7?
‘Continuous Flow Interchanges’.
Wasn’t that the original purpose of roundabouts! (Now succumbing to light control in many cases).
Rational Plan suggested that auto-cabs would allow us oldies to rely less on buses. There is a simple reason why we do so (outside London with its Freedom Pass), they are free to use for most of the day! Will auto-cabs be free too? If so, bring ’em on!
@Harry Crayola – New technology gets adopted as quickly as asset replacement schedules allow. PC’s took about 10 to 15 years to reach saturation level, whilst smartphones took about 7 years. That is not a reflection of technology adoption speeding up, but because PC’s were a lot more expensive.
Driverless cars are going to be a lot more expensive than a driven car initially because they have a lot more sensors and other equipment, as well as taking on a proportion of the development costs. It will take a number of years for that additional cost to reduce to a reasonable level, but even then, cars are generally only replaced every 3 to 5 years, but have a large secondhand value, so the average life of a car is 10 to 15 years. On that basis, it is going to take 10 to 15 years after the price settles for driverless cars to reach saturation level, so possibly 20 years+. The driverless car may be a desirable item, but that doesn’t mean people can immediately afford to go out and buy one. Besides, many people like driving so won’t initially be attracted to them (particularly outside of cities).
The same applies to Taxi’s. You may be able to save the drivers wages, but that has to be offset against the cost of buying the vehicle in the first place. That means 3 to 5 years once these vehicles are proven and reasonably priced, but that means a reasonable adoption in the general market first, so probably 10 years.
Also, you cannot use driverless cars 24/7 because there isn’t the demand. If you have a fleet of 5 cars in London, you will be using all 5 during the rush hours, 3 or 4 during the rest of the day and 1 overnight. This also then gives you time for maintenance and recharging, but your utilisation becomes around 60%, which isn’t far off what many taxis get today.
As for delivery vehicles, the driver isn’t just the driver, he is also the loader/unloader and often customer services rep. You are going to have to automate more than just the driving – how would it cope with doing Amazon deliveries to house ? Would it just beep its horn until you came out of your house and opened up the relevant delivery box on the back of the vehicle ? It is going take a long time for that solution to be cheaper than a person with their 20 year old van.
@Jim Cobb
……..many people like driving so won’t initially be attracted to them (particularly outside of cities).
The reverse argument is also applicable. Due to the lack of local public transport in rural areas many people will be eagerly awaiting the arrival of driverless cars to liberate their social lives and enable them to have a drink when visiting friends, at dinners, visiting the local pub, etc. A look around the age and make of cars in villages in this neck of the woods confirms a straw poll that indicated their arrival is eagerly anticipated (plenty of relatively wealthy and still active retirees).
@Jim Cobb
I think we’re talking at cross purposes. I agree with most of what you say, and indeed much of that is what I’ve already said, albeit with slightly different phrasing.
Driverless cars are going to be a lot more expensive than a driven car initially because they have a lot more sensors and other equipment, as well as taking on a proportion of the development costs. It will take a number of years for that additional cost to reduce to a reasonable level
Yes! Although I suspect that number will be lower than I infer you think. Mass production does wonders for cost reduction.
cars are generally only replaced every 3 to 5 years, but have a large secondhand value, so the average life of a car is 10 to 15 years. On that basis, it is going to take 10 to 15 years after the price settles for driverless cars to reach saturation level
Agree. This is why I said “their mass market place will come very quickly” rather than their share of the total accumulated stock of cars on the road.
The same applies to Taxi’s. You may be able to save the drivers wages, but that has to be offset against the cost of buying the vehicle in the first place. That means 3 to 5 years once these vehicles are proven and reasonably priced, but that means a reasonable adoption in the general market first, so probably 10 years. Also, you cannot use driverless cars 24/7 because there isn’t the demand. If you have a fleet of 5 cars in London, you will be using all 5 during the rush hours, 3 or 4 during the rest of the day and 1 overnight.
OK – so let’s have a look at what these numbers might translate to (I agree, btw. That’s why I said “multiple drivers”, deliberately leaving the question of how many per vehicle vague.)
Let’s assume your peak time means 8 hours a day (6-10am, 4-8pm), your overnight means 6 hours (from midnight to 6am) and your off peak means the remaining 10. If you had a fleet of 5 cars, you’d get 40 car hours in the peak (8 hours times 5 cars), 35 in the off peak (10 hours times 3.5 cars) and 6 overnight. That total of 81 hours averages to just over 16 hours per car per day. For simplicity, let’s assume employment costs for drivers average out at £10 per hour. That means a car’s driver employment costs in a day are £160. Multiple that by 360 and you get an annual employment cost of £57,600. Compare that to IHS Automotive’s 2014 prediction that driverless technology will add between $7,000 and $10,000 to the price of a car in 2025.
And that’s before you think of the marginal journeys where demand exists to pay for the car and the fuel but not for the driver, during peak hours too but especially off peak where the car isn’t a marginal cost.
As for delivery vehicles, the driver isn’t just the driver, he is also the loader/unloader and often customer services rep. You are going to have to automate more than just the driving
Agree. Maybe in 15 years companies will figure out a way of doing this on a widespread basis, maybe they won’t. Maybe, as Amazon plans, they’ll use drones. Who knows? Time will tell. We don’t have to worry about every single type of vehicle use now.
@Harry Crayola, 1 May 2016 at 14:43
For (small) local deliveries, how about autonomous ground vehicles no bigger than a pram or stroller moving at walking/cycle pace along local residential roads, pedestrian pavements and cycle paths? For example https://www.starship.xyz/. Much safer than airborne drones but probably easier to catch and steal too!
Harry Crayola et al: Whilst comparisons with car usage based on ‘hire plus fuel’ are probably reasonable *now* the use of electric vehicles — where you can recharge at each end — would appear to be more interesting and, dare I note, far cheaper. I have a friend with a fully-electric car (BMW i3) and their marginal-cost-per-mile is *very* low indeed[1]. On an allied point, he has already had ‘issues’ over the ease of hacking the car’s eletronics (ie. very insecure) [2].
When I’ve used a Zip car I’ve driven to the location and parked _my_ car near the ‘zip space’. I’m wondering if this is not as intended! Driving is *very* much more convenient if you aren’t travelling station to station. As soon as you have local journeys (at each end) you are wasting the time oyu’ve gained by a fast train.
I feel that people “who pay through the nose for peak tickets just because they don’t like driving?” actually abhor trying to find a parking space (also a price-contingent decision)
[1] The car tweets the details (yes, direct from the vehicle!) https://twitter.com/edent_car
[2] https://shkspr.mobi/blog/2015/11/reverse-engineering-the-bmw-i3-api/
Mark Townend “pedestrian pavements” – I sometimes daren’t walk along local pavements because they are in such disrepair. I doubt that a small vehicle would manage it. Plus what’s to stop some person of nefarious intent just picking the thing up and walking off with it? Stick it straight inside a metal box (aka faraday cage) and it won’t even be trackable!
@ Alison W – apparently they have cameras / CCTV and are designed to photograph and instantly transmit the image to the “control room”. When it was shown on BBC London the creators were very keen to emphasise that “interference” was a futile endeavour. Won’t stop people doing it though. Create some new techno whizz product and the first thing people do after the initial excitement is work out how to break it. This will apply to any endeavours about “driverless transport”. When someone succeeds in so doing then we are instantly into reputational damage territory. Personally I think these robo-vehicles should be designed by a Bond villain and be designed to kill / incapacitate anyone who dares to interfere with the robo-vehicle. 😛 😛
@Alison W
“the use of electric vehicles — where you can recharge at each end — would appear to be more interesting and, dare I note, far cheaper”
I’ve seen some interesting opinion on EVs very recently (not least, this). I haven’t mentioned it because I don’t know enough and I didn’t want to confuse the point I was making about comparing the financials of a train journey with those of a driverless car with another point. Partly this was because I’m not convinced it will have any effect on driverless versus manually driven cars, but they do look like they may well be another technology that is going to meaningfully weaken the fundamentals of non-metro rail transport.
Agree on the parking issue. People who can afford to take taxis don’t have this problem, nor do people who can afford chauffeurs. Driverless cars will effectively give all cars chauffeurs (at a very low cost) while slashing the cost of taxis. How attractive will rail travel be when you can buy a 24/7 chauffeur to your car for a few thousand pounds?
The big problem with cars (and many other things) will be hacking, not just by criminals but also by government agencies and corporate interests. But I don’t see this as being particularly a problem specific to driverless cars. I don’t see any reason why driverless tech requires vehicles to be connected to a network (except for software updates). They should be capable of autonymous operation reacting to their surroundings rather than relying on a control centre. That said, I think they will be connected (as will other cars, which will probably all have at least some of the technology) and so I agree this is something for which we need to maintain vigilance.
The key attribute for driverless cars to reduce traffic congestion in towns and cities must surely be inter-car communication. Having mentioned roundabouts in partial jest I realise that traffic lights could possibly be eliminated and replaced by mini roundabouts if all cars approaching were deciding in advance which should give way to which. What to do with those few retro human controlled vehicles that remain? Do we keep ‘lights on sticks’ just for them? Sell your shares in the traffic light industry now!
Re Harry C,
Electric vehicle may not look so cheap when usage increases and the distribution network needs upgrading to cope (creating lots of roadworks and delays…)
HMT will also lose a huge amount of revenue from a shift to EVs so expect changes to vehicle taxes at some point.
@ Ngh – given that various commentators are predicting the “lights will go off” in the UK due to power generation shortages there are surely serious issues related to the expansion of electric vehicles? TfL are already on record that if the bus fleet was to move to largely electric vehicles that additional generation and distribution capacity would be required to allow thousands of buses to be charged overnight. It has already been the case that getting the infrastructure into places for the very limited trials of such buses in London has been problematic. I don’t see mass charging of tens or hundreds of thousands of “cars” being any more practical unless we get national energy policy sorted out.
Fandroid refers to mini-roundabouts, and driverlessness.
The exact protocol to be used to manage a given junction is a technical road-design decision, independent of the driver species. Some junctions are already mini-roundabouts, and for those that are not, it is because designers believe that a different protocol (e.g. traffic lights) works better.
If all cars are driverless, however they communicate they still need a protocol. Maybe the lights on sticks are there no longer, but some junctions might still operate as if they were. Others might operate as if one or more mini-roundabouts, without needing the paint on the road.
What all-driverlessness might permit, however, is completely new protocols, impractical for homo sapiens. For instance, simultaneous right turns by vehicles approaching in opposite directions might be flexibly managed as non-conflicting if and only if vehicle sizes and the rest of the traffic allowed. (Which kind of happens already in Indian cities, though more anarchically).
But as someone said, this would require a very high level of software quality assurance, ground already expensively prospected by railway signalling and control systems today.
Re WW,
Exactly, DfT and DECC have known this is and it will be an issue for a years (though with staff rotation and turnover how often will they have to relearn?).
For example lots of the existing grid transmission and distribution substations are designed to cool down over night when loads are lower and this won’t happen with EVs charging…
Theoretically the market will of course respond except it won’t in practice.
@Malcolm – only if you eliminate all other non-automated vehicles – bicycles, agricultural vehicles, horses etc – which will still need to have Mk 1 signalisation to enable them to interact with other traffic. Oh, and in all towns and much of the countryside, that includes pedestrians, too. Likely?
If power infrastructure is a problem, TfL could locate its bus charging sites to make use of underused railway power supplies overnight. For driverless cars, overnight charging would be simplest if individual ownership was retained as the dominant model. All those driveways and domestic garages would retain a degree of usefulness.
While we all dream and speculate about extremely smart vehicle technology, it remains true that a lot of our existing street technology is about as unsmart as it can get. Take most light controlled pedestrian crossings as an example. If the pedestrian button is pressed, then the arrival time of the green man depends on when the button was last pressed. It has no relationship with actual traffic conditions. The simplest of smart signals would detect if there were any vehicles within range and, if not, start the vehicle stop cycle straight away. If the road was empty I would just cross on the red man signal without hitting the button, but there are plenty of people with mobility problems who would like the flaming lights to get a move on. It won’t be much of a Brave New World if the cars are exceptionally clever but those on foot have to make do with very dumb lights.
.
@Fandroid – “overnight charging would be simplest if individual ownership was retained as the dominant model. All those driveways and domestic garages would retain a degree of usefulness.”
Except that most houses built before about WW2, even if they have off street parking at all, have garages too small to accommodate anything much bigger than an Austin 7. And we are not going to downsize – almost anything that small is not going to pass modern crash tests. Even with a resdients’ parking scheme, I consider it a result if I’ve managed to park within 100 yards of my house (with bonus points for being on the same side of the street).
timbeau
almost anything that small is not going to pass modern crash tests.
OK – I’m prejudiced, but – so how come the “Smart” is supposedly crashworthy?
Those things strike me as inherently unsafe, anyway, for a large range of other reasons, as well.
Generally.
I am very suspicious/cynical of the Brave New World of all-automated (& putatively electric) vehicles. I can only re-iterate:
Pedestrians / cyclists / country roads outside towns / failure of updating of mapping etc.
That last about failure of update is a guaranteed certainty I’d have thought.
Somehow I don’t see HM The Queen, the PM, or the prison service whilst moving prisoners around every using autonomous vehicles. The risks of ‘unsavoury interaction’ abound.
@Harry Crayola:
“We’ll be on 5G tech at least by then.”
“5G” doesn’t actually exist in physical form yet. There are the usual bunch of competing technologies. One might become an outright winner, or we might see a combination of technologies combined to produce the final product. [Snip]
“But I’m not sure why you need them all trying to communicate with each other, or indeed any control centre. Why do you need a control centre? They’re being designed to be autonomous. They only need to be able to communicate with cars near enough to them that they may come into conflict.”
While the actual driving part will be built into each vehicle, autonomous vehicles only really make sense if you can make the most effective and efficient use of them. If, as is usually suggested, the goal is for them to be operated in fleets, [Snip]
If there’s a burst water main on the A232, or a derailed train affecting services via Herne Hill, a computer needs to collect all that information and decide how best to reroute your journey.
[Snip]
There’s more, but this post is already
long enoughtoo long [and has therefore been severly snipped Malcolm].@GH
Spot on.
@GT
I think there will be a significant problem until you caan eliminate shaaring of road space between humans, whether in vehicles or not, and completely driverless vehicles. As long as there is even a part time or remote driver (cf drones) I think this can be workable.
@AW
And don’t forget the police.
@Fandroid:
“Overnight charging would be simplest if individual ownership was retained as the dominant model. All those driveways and domestic garages would retain a degree of usefulness.”
I think I’ll buy some shares I extension lead manufacturers. And the companies that make those rubber strip things you see in older offices to keep people from tripping over wires and cables that have to cross the floor.
Most of the UK’s cities already struggle to cope with extensive private vehicle ownership, as that common view of a typical Victorian London residential street demonstrates. Never mind the National Grid’s generation and distribution capacity; where are you going to put all the high-Ampere sockets in that street? What design will they have? Will they require a three-phase industrial-grade supply?
Providing the sheer quantity of electricity needed to recharge entire roads of on-street parked electric vehicles within a reasonable time is going to be a mammoth task in its own right.
This is why I haven’t assumed electric vehicles will be mainstream at the same time as computer control. The UK has too much legacy infrastructure that will need sorting out first, not to mention the red tape and politics that will doubtless be involved. Other countries will likely make the switch to all-electric long before the UK can catch up.
I suspect that the UK will first see computer control creeping into conventional cars in the form of some kind of “autopilot”, most likely initially for use on motorways only.
@Fandroid, 2 May 2016 at 14:39
“. . . overnight charging would be simplest if individual ownership was retained as the dominant model. All those driveways and domestic garages would retain a degree of usefulness.”
That could work with hire vehicles as well. By encouraging customers to take a car home with them, space at an operator’s depot for overnight stabling and charging could be reduced. The customer could even be incentivised to do some light maintenance and cleaning whilst it was in their care. Whilst the customer is at work the vehicle would go into the general daytime hire pool and another car would be made available for the customer to take home at the end of the working day (very unlikely to be the same vehicle). As well as the car being available for the customer’s own leisure use in the evening the car could also make a number of third party local trips based out of the customer’s drive or parking space if the customer was agreeable, and that could be incentivised by the commuter getting a commission on any such local fares. Such a facility (and the corresponding daytime third party hire service whilst at work) could also be offered for owner drivers. With robocars we could all become our own Uber operator in the evening without even leaving the house!
As to charging, another way of looking at the problem is to think of the cars collectively adding storage to the grid. The more electric vehicles are utilised however though the less opportunity there would be for intelligent timing of the charging cycles. Note large scale grid storage of many different kinds generally is seen as a most promising way of managing the fluctuations of demand and renewable production. There is some traditional use of pumped storage hydro to bank off-peak nuclear capacity that can’t be turned off easily or quickly and would be wasted otherwise.
As far as self driving delivery vehicles (OK, partly self driving) are concerned, these were in common use up to the 1950’s. My big brother has told me that, even when he was a boy, the milkman would still take a hand held crate from house to house, delivering milk and collecting empties. He would return to the float every few doors to swap empties for fresh bottles.
The milk float moved by itself to keep up with him. This was because the horse knew what to do.
Of course the necessary infrastructure provision by the Metropolitan Drinking Fountain and Cattle Trough Association is no longer functional.
I also understand that the horses providing power for beer deliveries were in charge of the journey back to base because the draymen were no longer fit to drive unassisted.
@ngh
“Electric vehicle may not look so cheap when usage increases and the distribution network needs upgrading to cope (creating lots of roadworks and delays…)”
Perhaps. I know very little about them and I am somewhat less confident about my opinion on their prospects than I am about driverless cars. I wonder if NIMBY pressure against new power generation might be too intense to keep prices reasonable, too?
“HMT will also lose a huge amount of revenue from a shift to EVs so expect changes to vehicle taxes at some point.”
This is an upside. Fuel duty at its current levels is a very bad tax that is exceptionally poorly targeted at its objectives (it is far, far in excess of any estimate of the cost of carbon and it is hopeless at internalising congestion externalities). Revenue raising, meanwhile, should be carried out as neutrally as possible and taxes on specific activities or consumption goods violates this priciple to the extent that those taxes exceed an objective of internalising an external cost. If EVs prompt government to shift from taxing fuel to pricing scarce road space, so much the better.
Harry Crayola says “Fuel duty at its current levels is a very bad tax …“.
This discussion has already moved so far away from Sian’s interview, that it seems rather odd to work to prevent yet further digression. However, I am going to try, so I am requesting that we just note this view of Harry’s, note that other people may disagree with it, and do not otherwise discuss it.
Anomnibus “All those driveways and domestic garages would retain a degree of usefulness.”
If only! I, like possibly the majority of people living in zones 1 and 2 (at least) have neither from living on the fourth floor (in my case) of a building. Just as many new builds are intentionally without parking / permission for cars, then lots of central-ish older properties in all cities don’t either. Unless the car is small enough to fit in a lift*
* which I don’t have either, sadly.
AlisonW: just to note that Anomnibus did not suggest that everyone has a driveway or a domestic garage. Just that there are, in total, quite a lot of those things in London as a whole, which, I think, is fairly indisputable.
@Alison W
This is actually the most difficult part of the problem. The parts of London where demand for EVs are highest are in inner London where journeys tend to be much shorter (on average) and so range anxiety is not a problem. But, as you say, 90% of residents in inner London do not have off-street parking that they can control and just use an extension lead plugged into their own supply. In much of outer and suburban London, householders are far more likely to have these but here demand is much lower, partly because of range anxiety and partly because fewer households can get away with using small cars (as most EVs are) except, perhaps, as second cars. I have yet to see a solution to this paradox.
quinlet: Are private car journeys shorter for residents of inner London? I would have thought there would be nowhere local to go (except a paid or otherwise constrained car park). Furthermore an inner London car owner would be more likely to have a pressing need for a car, or a high income, either of which would point to longer journeys. Whereas a car owner further out may still have the chance of going to nearby shops, church, friends etc by car, and may have a slightly more moderate income. (Though a high proportion of car use throughout London is for shopping, and I concede that travelling further to a preferred shop is more feasible further out).
“Range anxiety”. I know it is the standard term, but I wonder if I am the only one to think that this phrase sounds rather patronising – those silly anxious car users overly concerned that their car may leave them stranded. A similar concern to keep a fossil fuel tank topped up is considered normal sensible behaviour, rather than being labelled with an almost medical term.
Re Fandroid,
The railway substations aren’t designed for a high continuous 24hr duty cycles either! They need to cool down too…
Re numerous others,
EV power supply – for fast charging is ideally 3 phase 63A supply*. The (european) EV connectors are based on the IEC 60309 (industrial plug standard) 63A ones but with added smaller pins to allow communication between car and charger so the cable is only live when charging, limiting charging rate based on cable and supply and with the pins closer together to make the connectors smaller. The cables and connectors are rated for 70A.
*this is double the power (45.3kW) available from a typical domestic supply (24kW)
Losses on a 100m extension lead won’t be good and the average person won’t be able to carry it weighting in at circa 340kg!
Most charging cables supplied are based on a single phase 16A or 32A connections at the moment.
Alison, Malcolm, all: a few edge case, anecdotal tales from this weekend. My lovely daughter & her SO moved in together on 1/5. Dad’s jobs included: pick up a “Rug Doctor” on the Old Kent Road for return within 24 hours by flatmate via Uber; shift stuff from (a) Putney to Brockley; (b) New Cross to Brockley; (c) Battersea to Brockley. The SO opted to hire a van to shift from Mile End to Brockley, picking up a Gumtree sofa on the way. The young un’s now have a cyclable commute to Denmark Hill & Greenwich; and friends within walking distance.
I live in Z2 with on-street parking; I logged ~55 miles between the various venues. Time spent? Don’t ask.
Could I have done it with a Zip van? Absolutely.
Would it have been easier with fewer traffic lights and speed bumps and 20 mph zones & 6 ft 6 inch barriers? You bet.
But it’s not just about us three. It’s all about the system.
@Malcolm: I think the difference is that with petrol you know that you can refuel the car in a few minutes and be on your way, whereas if you run out of charge you really are stuck, since that means that the car will be unusable (and unmovable) for several hours. The ability to refuel quickly means the “range” of a petrol car is effectively unlimited in any area where petrol stations are located more frequently than every few hundred miles, and so there is no real need to be anxious about it. But if you drive more than X miles from home in your electric vehicle, you are literally going to get stuck. That is quite reasonably anxiety-inducing, though less so the larger the number X is.
Various solutions to electric range anxiety have included battery swapping (hasn’t caught on) and fast-recharge stations (so far proprietary to one brand of vehicle).
On the effects of vehicle charging on the electricity network: there’s a good reason that Tesla’s other product line are home batteries for electricity storage (designed to be charged during the day from a solar panel – perhaps more viable in California than London…)
@Fandroid: It strikes me that parking might be transformed. If only pool cars were allowed in a car park, then they could just fill the entire space leaving no open roadways
You can automate a car park without having to automate the vehicles – systems like this are a logical development of automatic warehousing systems. Simpler paternoster-style systems are also quite common in Japan – you just drive into a space at ground level and the parking cells are rotated as needed.
@IanJ – Other systems may also be available; the Autostacker in Woolwich and the Zidpark in Southwark worked on the basis of cars being placed on dollies and these dollies then being stacked and retrieved automatically. Don’t know what the retrieval times were but the technology couldn’t be made to work commercially in 1961, when the Woolwich site opened. Zidpark’s a splendidly dated term though…
OB
…. & 6 ft 6 inch barriers …. Some people’s private cars can’t get under those, either. Or the Epping Forest Ranger (Verderer) I saw drive under one of his own employer’s similar barriers, only to lose all his roof-lights, oops.
[ 6ft 6″ = 2 metres – my car 2.04 metres, oops again ]
@IanJ
The cost of mechanical car parks is very high and the commercial case only works in areas of very high land value. These days, though, these tend to be the areas where cars are most actively discouraged and few new car parks are built. There was the other problem with the automated car park in, I think, Westminster, many years ago. This worked by putting the cars into a free slot from one side and retrieving them from the other. Unfortunately the software of the day failed and it tried to put more than one car into one slot!
@IanJ
Rapid chargers are intended to give 80% recharge within 20 minutes (still a long time compared to filling up at a petrol station), but regular use of these will wreck the battery. Proper battery management relies on trickle charge most of the time. The DfT is aiming for a network of rapid chargers along the motorway system and this may help to reduce range anxiety, because you may be able to get home and will certainly be able to move the car. I’m not sure it’s the answer though and long journeys will continue to be a problem for EVs.
@quinlet – there was talk fairly recently of battery exchange points where a tired battery could be swiftly replaced by a re-energised one;whatever became of that? [I can foresee that it would require heroic degrees of cooperation between manufacturers, tho’]
@Malcolm
I have been trying to recover a report I saw some time ago about journey lengths in inner and and outer London which showed clearly that outer London residents make, on average, longer car trips. However I have not been able to retrieve this and TfL’s documents now tend to concentrate (quite rightly for normal purposes) on trip numbers rather than distance travelled. The best I could find is that outer London residents make an average of 0.8 trips per day as car drivers compared to inner London residents who make only 0.4 trips per day, on average, as car drivers.
However, I don’t think your logic really works. It’s true that for most people in inner London, their local journeys will not be made by car – except during evenings and weekends where car use for local journeys is far more prevalent. The school run also takes place in inner London and these are generally very local journeys. High earners also make local journeys and they are more likely to use a car than others.
However one of the very big differences between inner and outer London is that car use for journeys to work is much higher in outer London and these can involve relatively long trips. Radial commuting journeys may well be made by rail but so many outer London work journeys are not radial.
The term “range anxiety”: I quite agree that potential purchasers or users of electric car should take this into account. (And therefore, with current technology, in many cases, go for the more traditional fuels, while that is still an option).
I was trying to suggest that the very real phenomenon currently known as “range anxiety” should have a better name. “Anxiety” suggests (to me, at least) a feeling which one ought not to have, as in “Don’t look so anxious, everything will be fine!”.
Whereas, as made clear above, the feeling that one might be left stranded requiring emergency assistance is a perfectly rational response (to the fact that one might be left stranded requiring emergency assistance).
Anomnibus 19:08 ‘“5G” doesn’t actually exist in physical form yet. There are the usual bunch of competing technologies. One might become an outright winner, or we might see a combination of technologies combined to produce the final product. [Snip]’
I’m a telecoms journalist and editor and as such I am following the 5G story fairly closely. The industry is cooperating remarkably well in its programme to develop 5G — far better than it did with 2G, 3G and even 4G. It may all fall apart, but so far I’d expect a global standard that can serve whole range of needs, including connected and autonomous vehicles.
@Malcolm. Clearly you don’t like the phrase “range anxiety”, but it has become an industry standard term in respect of electric vehicles.
@Island Dweller. Absolutely. I will therefore stop going on about it.
Solution to “range anxiety” (and yes, it may be the accepted term but!) – reintroduce motorrail!
When doing a longer-distance journey, drive your electric car to the local on-boarding facility, get transported to facility near destination, drive off and carry on. Charging facilities provided on-board. Sorted!
Alison. Methinks the capacity of motorrail to take cars off motorways (while the same tracks already take people and freight), is so small as to be a luxury choice. I don’t know what the best electric car ranges are, but I suspect that the top end manufacturers have worked extremely hard to extend them to be close-ish to those of conventional fuel cars. Isn’t Tesla investing in a massive battery factory?
Motorail. As Fandroid says, there is insufficient rail capacity. Particularly as the trains would have to run at least half-hourly, to satisfy car users accustomed to travelling when they are ready, rather than on someone else’s schedule.
Fandroid’s other point, though, about increasing electric car ranges, rather misses the point made earlier that, in countries with sufficient fuel stations, the distance you can travel in a normal car without being obliged to stop for several hours is effectively infinite.
Greg they were actually 6’6″ wide not high. Most of the height barriers round my way are OK for a 1990s ‘monospace’.
@AlisonW, Malcolm: I agree motorrail is unlikely to make a comeback, but I wonder if the Channel Tunnel operators have considered fitting fast charging points in their car shuttle wagons?
@Graham H, quinlet: As you allude to, there isn’t really much of a business case for building new car parks at all in central London and there is probably more of a case for demolishing existing ones and selling the land for other uses. Local politicians used to talk about “the parking problem” and assume the problem was lack of supply. Once they switched to reducing the demand for parking via pricing, residents-only schemes etc., the “problem” went away. There’s probably a lesson there for all kinds of transport policy that would fit with Sian Berry’s overall philosophy.
In all the discussion above, I haven’t seen how these driverless cars or other vehicles would deal with zebra crossings (of which there are of course still many in London), where folk intending to cross stand and wait on the pavement until the driver of the leading vehicle of approaching traffic decides to have the decency to stop.
How will such a situation work with driverless vehicles, possibly being unable to distinguish between a group of pedestrians hovering around a zebra crossing to have a natter and those intending to cross the road or indeed anyone who has just crossed to the other side?
The Highway Code has this for pedestrians: “Rule 19
Zebra crossings. Give traffic plenty of time to see you and to stop before you start to cross. Vehicles will need more time when the road is slippery. Wait until traffic has stopped from both directions or the road is clear before crossing. Remember that traffic does not have to stop until someone has moved onto the crossing. Keep looking both ways, and listening, in case a driver or rider has not seen you and attempts to overtake a vehicle that has stopped.”
For drivers, this applies: “Rule 195
Zebra crossings. As you approach a zebra crossing
•look out for pedestrians waiting to cross and be ready to slow down or stop to let them cross
•you MUST give way when a pedestrian has moved onto a crossing
•allow more time for stopping on wet or icy roads
•do not wave or use your horn to invite pedestrians across; this could be dangerous if another vehicle is approaching
•be aware of pedestrians approaching from the side of the crossing.
A zebra crossing with a central island is two separate crossings.”
For the avoidance of doubt, zebra crossings are those equipped solely with Belisha beacons and no other traffic light control for road traffic. Perhaps the ‘manual’ equivalent that is present during school times is the lollipop lady or man, not necessarily controlling a zebra crossing.
And there are those that can believe that a driverless vehicle would be able safely and effectively to cope with all that on a public highway?
I can anticipate one answer to this but I’m not going to suggest it.
If the things were to extend out into country lanes, I can well imagine how they would(n’t) cope with the single track lanes with occasional refuges to pass one another and farm tractors and the like emerging from obscure gates part-way (one tends to give way to the latter on single track lanes, no matter in which direction you are travelling and perhaps needing to reverse some good distance to avoid any unpleasant dispute). BTW, yes, such still exist within the Greater London area – try the area between New Addington and outer Bromley/Biggin Hill, for example.
@Graham Feakins
I don’t understand the difficulty. Which of those Highway Code rules you quoted are you suggesting would be difficult for a driverless car to incorporate into its software (or hardware – are you suggesting that it couldn’t detect the crossings – or the pedestrians, or their locations?)
Seems pretty simple to me. Driverless car must behave as a human driver would when they detect people near a crossing who might be about to cross: ensure speed is not above a safe stopping speed so that vehicle can be stopped in the event that the pedestrian steps onto the crossing. Now, the rules mean that cars do not have to stop unless someone steps onto the crossing, so they would travel at an appropriate speed. People would need to step onto the crossing to stop the cars, but they wouldn’t stop if they didn’t. Perhaps I’m misisng something, but this seems pretty simple to me.
Similarly, I imagine people would opt for software which replicated current conventions for farm vehicles on narrow lanes etc, too. Why wouldn’t they? What might the problem be?
@Malcolm
My apologies for wandering off into optimal taxation theory – and more so for doing so normatively and without citing the literature to substantiate.
@Harry Crayola – As a pedestrian, one is taught to stop and wait on the pavement at zebra crossings until an approaching car slows down *and stops* and not to try and step onto the road at all in front of a moving car before it has done so. Perhaps that message isn’t taught so much these days but it seems sufficiently clear to me from the Highway Code today and always has been.
And just how does one of ‘your’ vehicles detect whether one or more pedestrians standing or walking on a pavement beside a zebra crossing detect whether any one of them is facing the road with a look that says “I want to cross the road” and not one that says, turning away from companion, “Oh, look over there!” or indeed one who just happens, as I said above, to be just standing there nattering.
From what I understand you suggest, the software would automatically slow down every vehicle just because there happens to be somebody on the pavement just adjacent the crossing. That’s not terribly bright, is it?
As for farm vehicles, it seems apparent that maybe you haven’t experienced the surprise of negotiating a sharp blind bend with tall hedgerows/cuttings between irregular passing places along sharply twisting, narrow, single track lanes, those passing places themselves being well out of sight of one another, let alone the turns and obstacles in between, only to come across e.g. a tractor and trailer with hay, or a flock of sheep/cows in front of/following a farmer in transit between parts of a farm – try asking any farmer how often he may decide through any year to stop up/open up any link between his fields and land which the lanes divide. Oh, and it’s even more exciting driving at night along such lanes. The important aspect of horses, whether on country lanes or not, has been mentioned by others above.
I suggest that the best that the software you mention can do is to keep vehicles equipped solely with such well clear of country lanes.
It strikes me that Graham F’s observations about people standing near zebra crossings could well be used as an argument against introducing zebra crossings, if we did not already have them. They ought not to work at all, given the fundamental contradiction in their rules: vehicles need only stop if a pedestrian is on the crossing yet pedestrians are advised not to go on the crossing unless vehicles have stopped.
Yet they have worked fairly successfully for over three quarters of a century. Whether this success will persist into the era of driverless cars is evidently a matter on which different views are held. But one, I suggest, where real life experiment will fairly soon provide an answer.
Malcom
in countries with sufficient fuel stations, the distance you can travel in a normal car without being obliged to stop for several hours is effectively infinite.
I’d dispute that. One should not, really drive for more than 2 hours, without at least a 20-minute break between sections.
[Greg, the key part of Malcolm’s comment is “to stop for several hours”, meaning having to recharge the car. LBM]
Pedestrian crossings
Then there are the groups of pedestrians having an empty chat, right by the crossing … and you stop your car – & they just look at you ….
Or there’s a pedestrian waiting to cross & you stop & he/she waits for over 10 seconds, to make sure you’ve really stopped …
And, of course, car-drivers who attempt to bully pedestrians, by clearly intending not to stop, even if your toes are out onto the crossing ….
Um, err ….
@Graham H
Battery exchange would have been a far better operational choice for EVs than recharging a fixed battery as the ‘recharge’ time would have been cut to a few minutes, much like filling a tank with diesel or petrol. Moreover, petrol stations could have been adapted to become battery exchange points. But this would have required an industry standard which was never likely to happen and certainly won’t now, at least not in the near to middle future.
Interestingly battery exchange formed the basis of one of the world’s largest fleet of electric buses, which worked in London in the 1900s. They worked out of what became Victoria Garage (now Sainsburys) and a single battery exchange at lunchtime provided a full day’s work. Sadly the company was bought by what would have later been called an asset stripper who extracted all the value in cash and closed the company down, I think, just before WWI. It was never rebuilt.
@Ian J
Managing parking demand by price/permit is the only sensible way of dealing with parking in a congested area, especially as all surveys show that most motorists put a higher premium on parking availability than on cost. Sadly the political headlines are dominated by price and most of those politicians completely fail to understand that driving the price of parking down, or making it free, is probably about the most motorist and high street unfriendly policy you could imagine.
@Graham Feakins:
Autonomous vehicles will react to pedestrians at zebra crossings in exactly the same way they have to react to them along any other stretch of ordinary road: by anticipating their likely behaviour and driving accordingly. The same way drivers are obliged to do today. The UK has no “jaywalking” laws.
It’s perfectly legal (though not recommended*) to cross anywhere along a UK road, as long as it isn’t explicitly marked with “No Pedestrians” signage. The Highway Code makes it clear that the responsibilities are as follows:
Drivers must drive such that they can safely deal with any hazards that may present themselves while driving. Motorists are expected to drive according to the conditions of the road. This includes driving below the current speed limit if it would be unsafe to do otherwise.
Pedestrians should check the road is safe to cross before stepping off the pavement.
Under ideal conditions, then, a pedestrian will only set foot on the road (whether at a zebra crossing or not) if they deem it safe to do so. Drivers must slow down and / or stop as necessary.
Humans are, however, fallible, so it’s a good thing computer-operated vehicles will also have reaction times orders of magnitude faster than those of any human driver. To a human, the time a pedestrian takes to place their first foot on the road’s surface is very short; to a computer, it’s closer to an eternity. The computer-driven vehicle will be slowing down long before a human driver would have even registered the pedestrian’s movement onto the road. There’s no need for any fancy image processing; you can do this with little more than the sensors already used for adaptive cruise control and detecting obstacles while reversing.
—
* “The Highway Code” is, as its name implies, a set of guidelines, not laws. It recommends pedestrians cross roads at safe crossing points when available, but there’s no law forcing you to do this.
Quinlet: that would be London Electrobus – see http://www.economist.com/node/9465026.
@Graham Feakins
I’m still baffled by your comments, I’m afraid. They seem to be:
1. There are more circumstances where the tech would be less able to detect the intentions of someone at the roadside by a zebra crossing, and therefore the additional number of cases where a vehicle would need to slow down would be a very bad thing.
2. There are long distances between passing points on country lanes, which makes them tricky for human drivers. Only human drivers could cope with the challenges these distances/curves etc present.
The first point is fair enough, I suppose, on its own terms. Althhough when the tech becomes able to detect which way people are facing and whether they are moving or not, it will be restricted to cases where people are facing the zebra crossing and standing still nattering, and where a human can clearly tell they do not intend to cross (and are therefore need not slow down or stop for them). Doesn’t strike me as a gigantic problem. Especially when the faster reaction times and hence increased speed and/or safety are assessed which counter this effect (overwhelmingly, I would suggest).
You say that it’s “not terribly bright” for software to slow down the vehicle automatically when someone’s at the crossing. I don’t think we’re going to agree on this one. I think it’s pretty smart.
Finally, you say we’re taught to wait until a car has completely stopped before even taking a single step onto the crossing. Perhaps. But I can’t remember actually doing this or even seeing anyone do it. People know when a car is traveling at 5mph and decellerating to a stop that they can cross, and they do. Why would this be different in future?
As Malcolm says, we’ll soon have a real life experiments to provide the answers! Perhaps we’ll need to turn them into pelican crossings or tweak the highway code to recommend a single step onto the crossing to remove the possibility of conflicting rules (or anything similar with the same signalling effect. I’m sure we’ll cope.
@Mike
Thanks for the link. Who knows, without the fraudster maybe electric buses would have been the norm for the past century? Although the lessons of GM and streetcars in the US might have meant they would only have survived another few decades.
@HC
The way that zebra crossings currentlty work substantially depends on actual communication between the human on the ground and the human driving a vehicle. This is effected by eye contact, gesture, posture, behaviour and of course the occasional word or smile.
As more and more Mini-Hollnad schemees are introduced by teh elimination of obstacles such as kerbs, lines, signs and barriers, such road sharing is likely to become more prevalent.
I find it difficult to envisage any improvement to safety as long as the computers or robots operating vehicles are bereft of such means of communication.
Another major hazard arises when the pedestrian, cyclist or human driver is unaware that the person sitting in the front of the driverless vehicle has absoulutely no influence on its operation. This is made worse if driverless and human operated vehicles are mixed in traffic.
Tweaking the highway code is one thing. Getting people used to it can take generations.
And it isn’t real life experiments which worry me, it’s real death ones.
@Harry Crayola – I take the point that the software will improve over time (but perhaps not before it has wiped out a number of footpunters) and even get to the stage where it will tell the prisoners in the selfdriving car that the person who has just stepped out in front of their vehicle that the victim is old Mrs Soandso and that she has just been shopping for butter at Tescos, but that is only slightly to the argument. Software engineers will respond (as engineers do everywhere) with a patch that manages the risks conservatively. Manually driven cars can (and hopefully do ) make emergency stops which can be quite damaging to the passengers; automatically driven cars will be programmed to avoid such stops and as such will take more time to stop. This will absorb road capacity and slow the traffic down.
It will not take long for pedestrians in busy locations to realise that if the stream of them crossing a road is unfettered the driverless cars will all stop and hey presto we have a pedestrianised area with a load of static cars going nowhere.
Nameless
“I find it difficult to envisage any improvement to safety as long as the computers or robots operating vehicles are bereft of such means of communication.”
Why do accidents happen? This is the first link I found when I Googled “causes of road accidents”. It seems pretty clear to me that many among these look like great contenders for improved safety by driverless tech.
@Graham H
You seem torn between the choice of whether driverless tech will be programmed conservatively, reducing speeds as software adjusts speed for potential hazards, or aggressively, killing pedestrians because it can’t stop the cars fast enough (perhaps in jest? It’s hard to tell humour on these forums sometimes!), despite having reaction times phenomenally superior to human drivers.
I note you have identified yet another benefit of driverless tech that hasn’t yet been mentioned: fewer uncomfortable emergency stops!
@Nameless (and others):
“I find it difficult to envisage any improvement to safety as long as the computers or robots operating vehicles are bereft of such means of communication.”
What makes you think they will be?
Face recognition is already a common feature in photo album software right now, and that’s the really hard part. Once you can ‘see’ faces, you can then refine the algorithm to work out which way they’re facing.
There is also already an extensive body of research into cognition and interaction design, including robots that can already interact with humans using emotions. (It wasn’t all that long ago when basic voice recognition was considered “too hard”!)
Vehicle automation is a field that ranges across a lot of branches of science, from Physics, via Cognitive Science and Interaction Design, to Philosophy and Ethics.
A key issue with autonomous vehicles is how they get their information. It’s a common misconception that any self-driving vehicle must work out how to get around in exactly the same way humans do. Not so: computers can make use of all sorts of sensors, not just human-audible sounds and visible light. Infra-red sensors could detect a small child between two parked cars, who would otherwise be hidden from view from a human’s perspective.
There’s also no reason all those sensors have to be mounted on the vehicle itself: infra-red sensors mounted on each lamppost could relay their images in real time to nearby vehicles, thus providing an additional aerial perspective to that provided by on-board systems. Even human-driven vehicles can benefit from some of these technologies during the transitional period.
@HC – I simply don’t know and nor can the rest of us, alas. What is clear is that the benefits of driverless cars aren’t all self-evident and that there may well be equal and opposite downsides.
@Kit Green – and as others have noted earlier, it won’t take long (about 5 minutes is my guess) for children (and robbers!) to discover that they can bring traffic to a complete halt by bouncing a few balls across the road.
@HC
While not disagreeing with everything on that link, I would point out that it has been prepared from a US perspective where:
a) The road system gives almost total priority to motorised vehicles and “jaywalking” is illegal
b) Their road layouts are much simpler
I have no doubt that totally driverless roads would be very safe. However there are problems at the wetware/hardware interface which are not simply addressed by a software patch. There is a fundamental problem in communication between people and machines. This is especially so when the people might not know that they are trying to interact with a self driving vehicle and not with its visible occupant.
On second thoughts, are you just trying to engage us in the Turing Test?
@GH
And for cyclists to discover that they have absolute right of way.
@Anomnibus
And there is a huge gulf between what can be achieved in clear bright dry conditions at close range and the real damp gloomy world of the British urban road at dusk in November.
And the communication you have exemplified is one way. How does a human ascertain the driverless car’s intentions? Can it nod or wave you on?
Err …
Putatively electric small personal transport, with autonomous controls, as per “google” car?
THIS, then?
Note that it is lab-scale only at present.
@Nameless, Anomnibus
Re nods and waves – I was taught as a bus driver (on the left side of the Atlantic) not to take a nod or wave into account, as legally the intention of such a gesture can be construed to be completely different to what was intended. Software and cameras in a self-driving vehicle will be able to record the gesture, but interpreting it correctly is just as fraught with error for machines as it is for humans. And as Nameless states, bad weather, rain on the windshield, lights in one’s eyes/camera can obscure the gesture. The legal aspects and liabilities will drive (sorry) how the self-driving software is programmed in such situations methinks.
Anomnibus “It’s perfectly legal (though not recommended*) to cross anywhere along a UK road”
Actually, not always ‘perfectly’. Year’s back I was taught on a safety course during a section run by the B’ham Police that whilst it might be ‘legal’ if you knowingly cross the road at the ‘wrong’ location and cause an accident you will be held liable. Their example concerned a woman who stepped out in front of a bus near New Street station and caused the driver to slam on the brakes. The pedestrian wasn’t hurt but a person fell down the stairs. Because there was a zebra crossing within near reach the pedestrian was held liable for the accident and charged accordingly.
I think it is unlikely the software will be programmed to stop the car if a person is crossing at a zebra but run them over if a person isn’t crossing at a zebra. It could be the software makes no distinction at all. I don’t think zebras are universally used in all countries so the manufacturer will probably want to make the vehicle programming as generic as possible.
So the issue is more likely to be how the distinction is handled legally rather than by the technology.
As has been mentioned, will the car manufacturers push for jaywalking laws? Or deny or reduce the manufacturers liability if there is a collision with a person at an “unapproved” crossing?
@ Reynolds 953 – oh I can see a great deal of perfectly normal pedestrian and cyclist activity becoming “unapproved” in a future filled with robots on wheels. We will all end up bowing our knees to the technology and its suppliers.
AlisonW,
Are you sure you are not confusing criminal with civil law? I still don’t believe a pedestrian can commit a traffic offence with one exception touched on below. The pedestrian may be liable for damages caused by their action but that is nothing to do with criminal law. The police would have been duty bound to investigate and that investigation may have been available to decide damages in the civil case. I strongly doubt that the police (or technically the CPS) “charged” the pedestrian with anything.
So, I believe it is perfectly legal for a pedestrian to cross anywhere on a UK road unless they should not have been there in the first place (e.g. Motorway) but they have to take the consequences financially of being liable in a civil court for the consequences of their actions (“liable” rather than “charged”). This is no different from any other aspect of life.
@PoP
Pedestrians are, IIRC, forbidden to cross the road where there are zig-zags on either side of a pedestrian crossing.
@ quinlet – the Highway Code says “do not cross at the side of the crossing or on the zig-zag lines, as it can be dangerous” but this isn’t law. I could only see one instruction to pedestrians covered by law* and this was “You MUST NOT loiter on any type of crossing.”
* The Zebra, Pelican and Puffin Pedestrian Crossings Regulations and General Directions 1997
Just to complicate matters there is also the common law offence of committing a public nuisance which can apply to just about anything including traffic matters. More typically in this context it is used for blocking a road without lawful authority or allowing smoke from your bonfire to blow onto a main road. The weakness with using this for jaywalking is you have to show a sufficiently large portion of the public were potentially adversely affected by the action and not just a few individuals. This suggests there must be a level of persistence or repetition and it not be just momentary event that occurs only once.
With all the doubts and doubters about autonomous vehicles it’s interesting to see that the government in Dubai has now adopted a strategy with a target for 25% of the vehicle fleet to be autonomous (and used as such) by 2030. OK it’s Dubai so it’s got, reputedly, pots of money and the embarrassment factor of a failed target is probably less than in the UK, but, just the same, they’re not entirely stupid.
@quinlet, I’d also imagine that there’s a lot fewer pedestrians/cyclists in Dubai, where it is a little warm for outside exercise!
@Graham H – “it won’t take long (about 5 minutes is my guess) for children (and robbers!) to discover that they can bring traffic to a complete halt by bouncing a few balls across the road.”
You’ve been reading about the habits of London’s Edwardian children, haven’t you! It didn’t take them long at all to discover that, by dropping their metal hoops into the conduit slot of the LCC Tramways, they could easily halt nearby trams by shorting out the conductor rails!
@Si
Yes, but not none. And the issues that some have raised with respect to pedestrians and cyclists would occur with any pedestrians.
I note also that Lyft has just announced a real trial of driverless taxis in the US to start ‘within a year’.
@quinlet: a quote from the WSJ report about the self-driving Lyft trial:
In an effort to ease regulatory concerns, Lyft will start with autonomous cars that have drivers in the cockpit ready to intervene—but the driver is expected to eventually be obsolete
In the case of the Underground, the gap between automatic driving and “eventually” having unstaffed vehicles in public service looks to be at least 60 years – it’s not the simple step that Lyft would like the stock market to think it is.
@Ian J/quinlet – a classic case of all the difficult parts of the change being waved away with “expected eventually to be obsolete”. About as useful an explanatory tool as those recipes which begin “Take” (or worse still “add”)
The Underground has run unstaffed passenger vehicles for many years, although they were all originally manned. However, each one runs in its own dedicated “track”. I am. of course, referring to lifts.
The first manned passenger lifts on an subterranean railway in London were, I think those on the Tower Subway (opened 1870) but that was never part of the Underground Group. Unless Wapping and Rotherhithe got them earlier, I think the first stations on what was to become part of the Underground were those on the CSLR, which opened in 1890. These were manned. I think it took at least sixty years from then before there were any passenger-operated or fully-automatic ones – anyone know the date? – and I believe they are all unstaffed now. Aldwych was, I think, manned to the very end in 1994, but it may not have been the last.
@Ian J / quinlet
Unless it is actually made illegal to drive a motor vehicle on a public highway, there will always be some cars around that are not equipped for automatic driverless operation. Only last week I saw a car being driven on the public highway that was at least 80 years old. Any driverless cars’ software will have to be designed to make allowances for such things being present.
The railway environment is different – since Network Rail own the infrastructure they can make the rules and, for instance, insist that steam locomotives are retro-fitted with TPWS. On the public highway MoT tests do not, and cannot, insist on retro-fitting equipment that was not required when the vehicle was built.
timbeau says “MoT tests do not, and cannot…”
I’d agree with “do not” but disagree with “cannot”. It seems to be a matter of policy not to so insist, but that policy could be changed at any moment, and as regards safety improvements, there would be little outcry. There exists no inalienable human right to operate vehicles with brakes now-considered-inadequate, say, on the public roads, only a current practice of permission.
And of course on special roads, such as motorways, there are plenty of limitations as to what vehicles may use them.
@Malcolm. You’re right, government can always impose new regulations. But, in respect of private cars, I can’t think of much precedent for change in the way you suggest here. Regulations that have mandated safety features (from introduction of safety belts in the 60s to requirement for ABS/stability control more recently) have been imposed on new cars registered after a certain date, but I can’t recall any that have been applied to vehicles already registered. Even now, low volume manufacturers (such as Caterham) can choose not to fit some safety features (such as stability control).
My suspicion is that attempting to limit who can use ‘ordinary’ roads just to make life easier for ‘self driving’ vehicles would not be any easy legislative change to get through.
Malcolm & ID
New construction/modification of existing vehicles.
There is a set of precedents to follow here.
If the modification is easy & obviously beneficial, then everybody, or almost everybody is expected to retrofit the add-on.
“Blinking” direction indicators & seat-belts are the striking ones here.
If it’s a matter of more expensive, but also beneficial improvements or changes, such as anti-pollution measures, such as “Cat” converters or diesel emission &/or filtration, then there is usually a “cut-off point” (Often dated a year or two previous to the legislation, so that all vehicles built after $_DATE are expected to comply.
The expectation is, of course, that those older non-compliant vehicles will, (or 99% of them will) rust away, become uneconomic for other reasons & therefore their
undesirable qualities will “waste away” in time, apart for a special few, which are so thin on the ground as not to matter “in the great scheme of things”.
I fully expect that the introduction of automated or semi-automated vehicles will follow this pattern
@Greg etc
“If the modification is easy & obviously beneficial, then everybody, or almost everybody is expected to retrofit the add-on.
“Blinking” direction indicators & seat-belts are the striking ones here.”
Most people do, but there is no compulsion, and some enthusiasts want to keep vehicles in authentically original appearance. (I also knew of an owner of such a vehicle, finding his retro-fitted indicators/seat belts not to be MoT-worthy, removed them for the test rather than mend them)
Did I not read recently that some legislation to do with classic cars (VED-exemption, perhaps) actually requires them to be in more or less as-built condition?
Here it is
http://cars.aol.co.uk/2012/12/21/eu-backs-off-from-anti-modification-mot-rules/
@timbeau
of course there will always be some cars around that are manually driven and not automated. I think the problems of mixed mode – some automated and some not – are much less severe than the problems of dealing successfully with pedestrians and cyclists. Neither Google Car nor Lyft rely on roads dedicated to automated vehicles. Indeed, the main (and perhaps, only) reason why Lyft is using a ‘driver’ is that current motoring laws require this. I would expect that the day after the laws change, all other things being equal, Lyft would drop the driver.
@Nameless
“The milk float moved by itself to keep up with him. This was because the horse knew what to do.”
When the intelligence in a car is as good as that of a horse, we may be on to something. (Horses will generally stay out of the way of oncoming vehicles (whether propelled by an engine or another horse) and will instinctively stop short of an obstruction (my father-in-law owed his life to the horse he was riding spotting some piano wire strung across the path)
@quinlet
“Battery exchange would have been a far better operational choice for EVs than recharging a fixed battery as the ‘recharge’ time would have been cut to a few minutes, much like filling a tank with diesel or petrol. Moreover, petrol stations could have been adapted to become battery exchange points. ”
For most of history the entire motive power unit was exchanged when it ran out of energy – that was the function of the coaching inn. Even in the days of steam, when water could be collected on the fly, it was quicker to change locomotives than wait to re-coal it.
It is only with the advent of liquid fuel, which can be pumped typically at rates of 50 litres per minute, that a quick refuelling stop became practical. (The calorific value of petrol/diesel is around 36MJ/litre (slightly more for diesel than for petrol, so a petrol pump is delivering 30MJ/second i.e 30 MW. To charge a battery at that rate from a 240V supply would require a socket with a 125,000 amp fuse! (For comparison, Lots Road Power station, which powered most of the London Underground, was rated at 50MW)
And that is what so-called “range anxiety” is about – its not how far you can go on one charge, but how long you are going to be stuck there before you can carry on, or go back. Fine if you can get to your destination on one charge, and can be sure that charging facilities are available at your destination. Not so good if, for example, you are going to a remote campsite, or just making a delivery. In contrast, with petrol, it usually takes longer to pay for the stuff than to actually pump it.
@timbeau:
Supplying sufficient electricity to the recharge points is going to be the main obstacle. Converting a petrol station to handle all that additional energy isn’t trivial: railways are very efficient at shifting people around in bulk quantities, but they still require big, hairy sub-stations to provide all the electricity needed, and there has already been mention of how energy-hungry modern trains with air-con and sliding doors are these days. A Class 444 (“Desiro”, DC 3rd rail) train can soak up to 2MW of electricity through its motors. That’s per train.
There are a lot more road vehicles than there are trains, and rubber tyres on tarmac are inherently less efficient than steel wheels on steel rails too. While those road vehicles won’t individually require loads of electricity, the total aggregate demand, spread over an entire city of them, will likely be an order of magnitude more than that currently used by the railways, and that means a heck of a lot of new electrical infrastructure will be needed. Charging several hundred cars every single day on a single residential street will require staggering quantities of electricity that isn’t often available from the mains cables running down those kinds of roads. Expect an awful lot of road-digging.
[Paragraphs which drifted out of London & off topic snipped. LBM]
[Snip for usual tangential topic reasons. Ryde in Isle of Wight is outside the LR zone. Malcolm/LBM]
LBM / moderators
Non-trivial problem
“The L-R Zone”
Where is it?
1: Inside the GLA area – TICK
2: Inside the M25 where larger – TICK
3: Inside inner Home counties (*) & directly affecting London through transport – almost certainly. ( *: Kent, Surrey, parts of Sussex – e,g, Gatwick, “east Berkshire”, South Bucks, Hertfordshire, Thames Estuary areas & inner Essex – out to Stansted )
4: further afield in UK- only if directly relevant to topic under discussion, or in passing – probably.
5: Over the water – only in passing – though see recent references to Hamburg – a lot further away than the IoW …..
[As you may have noticed, the LR zone is wherever moderators say it is. Sorry if that seems excessively arbitrary, but that’s just how it is. But your suggestions do give an approximate guide. Malcolm]
I think it was previously covered in the Euston rebuild article namely how far the Scottish sleepers go from Euston!
A number of Tube trains have ended up on the Isle of Wight!
Malcolm Um, as Greg says from time to time. I would suggest that the “Network South East” geographical boundary(*) is appropriate, at least for heavy rail, as that still defines “network card” eligibility for season ticket holders even if the season ticket is wholly within TfL zones. *See here: http://www.network-railcard.co.uk
But although you are quite right not to fetter your discretion, it is helpful to know that Greg’s suggestions meet with qualified approval.
timbeau: indeed they have. In fact that was the subject of Anomnibus’ recent comment, which happened to hit a moderatorial strict patch, thus raising this whole “LR-zone” issue. As I mentioned in an addition to Greg’s comment, those seeking absolute consistency from moderators will have to look elsewhere. We are only human.
Sorry – the network card is a stand alone product but if you buy an annual season from a TOC (even, eg, Putney to Z1) that includes network card benefits and you can get a second network card for twenty pounds off. See: http://www.nationalrail.co.uk/times_fares/ticket_types/46573.aspx
@Nameless
“The milk float moved by itself to keep up with him. This was because the horse knew what to do.”
Your horse obviously worked for the wrong dairy. Our Express Dairy one, from College Farm, Finchley, expected the milkman to keep up with him, not the other way around. Deliveries took longer when, about 1960, the horse was replaced by an electric float as the milkman had to spend time getting in and out of it to move to the next house.
@Timbeau
“When the intelligence in a car is as good as that of a horse, we may be on to something. ”
Our horse was programmed. There was a woman who came out to give the horse some sugar lumps every day. One day she was ill and did not come out. No sugar lumps, no moving horse. Eventually someone else came with the required sugar lumps and the horse would move on. (Some sort of fail safe device?)
Are we now expecting the first commercial use of self-driving vehicles to be for the delivery of milk?
Simple route finding, with a list of customers’ addresses, and a remote control for the milkman to instruct ‘walk on’. My daughter has succumbed to the gentrification tendency in Walthamstow, and now has milk delivered regularly in glass bottles. The market is there!
@IslandDweller
Not quite right.
The 1967 legislation required the retrofitting of front seat belts to all cars registered from 1965 onwards. For example, my late Dad’s new car was first registered in February 1967 but did not need to have seat belts fitted until that July.
@GT
Surely the L-R zone is in the mind? And in my experience most people don’t have one.
@Old Buccaneer
“I would suggest that the “Network South East” geographical boundary(*) is appropriate, at least for heavy rail, as that still defines “network card”
Since that boundary extends to Shrewsbury and Exeter (and indeed the isle of Wight) I think that is probably further than is appropriate.
“if you buy an annual season from a TOC (even, eg, Putney to Z1) that includes network card benefits ”
With the recent extension of the boundary the cheapest available is now between the two stations in Lichfield (£164), depriving Island Line of a nice little earner as the previous cheapest was between the two stations in Ryde (currently £172). This is more than a Network Card costs, but there is no minimum fare for tickets bought with a GC, as there is with a NC.
Nameless: In spite of moderators’ advice to be concise, your comment may have overdone it slightly. It is not clear to me exactly what it is that “most people don’t have”. If they don’t have an L-R mental zone, that should not be a problem, they can just use one of the ersatz ones suggested. If they don’t have a mind, I am at a loss what to suggest…
@M
I meant figurative space in one’s mind conducive to consideration of such matters.
@Nameless:
Is this a “Left-brain vs. right-brain” reference? In which case, you’re right: I’ve generously given others pieces of my mind so often, I haven’t any left.
On a separate note: shouldn’t there have been a new issue of LR Magazine by now?
The remaining interviews with London’s other Mayor candidates might be a little out of place now, but it’d still be interesting to see what they had to say. (If anything.)
Not at all relevant to London, but it’s worth pointing out that confusion seems to have arisen between the Network Southeast area (as covered by the Network Railcard) and the Gold Card area. The latter was extended in Jan 2015 to include a much larger area. Prior to that the various areas were the same.
@Fandroid
I’m confused now – are the Network Railcard and Gold Card areas not the same now? That could lead to confusion if a Gold Card holder has a wider validity than the discounted Network Card he/she can buy for a partner.
(Although I discovered only today that you can buy ANY (one) railcard at a discounted rate if you have a Gold Card – your kids’ 16-25 railcard, your own 2together card, whatever.
@ Timbeau – definitely different areas.
http://www.nationalrail.co.uk/static/documents/content/routemaps/Gold_Card_scheme_area.pdf (Annual Gold Card area)
http://www.railcard.co.uk/clientfiles/files/map.pdf (Network Card area)
The L – R zone used to be quite clear. It was exactly the same as the A – D, E – K and S – Z zones and an outline map was, I recall, printed on the front cover of each directory.
timbeau @ 1459 on 11 May 16: I was giving an example from personal history, not seeking to score points (whether or not they make prizes is irrelevant, in my view).
The edge cases you cite are rounding errors in the great flow of cash between DfT, TfL, TOCs and NR.
AFAICT neither timbeau nor I are moderators.
Moderators rule!
(may I remind people humbly of my proposed test: ‘what real-world problem is addressed here?)
@Paying Guest – “The L – R zone used to be quite clear…&c.” – Tee hee! Funnily enough, that’s exactly what I thought of the moment I came across the first mention of L-R in the discussion above. Shows my (our?)age, I suppose but there again, I still have and use a functioning dial telephone with proper bell (and no, I can’t “press 1 to hear this message again”). LBM might appreciate this London Area minor distraction.
@Graham F/Paying Guest 🙂 I had a colleasgue who until the late ‘eighties used to refer to the “electric” telephone. [Not sure whether this was because he still cherished the tin can variety, or because he was a devotee of McGonagall – “across the wires the message ran/he is no better/ he is much the same”].
@Graham H
That wasn’t McGonagall. The anonymous lines, referring to the illness of the Prince of Wales, have been attributed to Alfred Austin, Poet Laureate from 1896-1913, but are probably an anonymous parody.
@Paying Guest
The phrase “green issues” seems apt.
I was tempted to wade into this extensive discussion on ITSO*, concessionary fares, cards, and whatnot, but I think I’ll pass.
(Sorry!)
—
* (What are the odds that someone will name their ITSO scheme “FACTO”?)
I knew timbeau could be relied on to dig out the possibilities…
Let’s get the wording right if we are going to quote it.
“Across the wires the electric message came: He is no better, he is much the same.”
@PZT – Indeed,the correct version appears to begin “O’er” rather than across.
If Wikiquote is to be believed (and anything with Wiki in front of it is suspicious), the full reference is:
” – On the Illness of the Prince of Wales (1910):
‘O’er the wires the electric message came,
“He is no better; he is much the same.’
– An 1871 poem on the illness of the Prince of Wales, although there is some doubt that Austin actually wrote this part. That classic compendium “The Stuffed Owl: An Anthology of Bad Verse” (2d ed. 1930; Capricorn paperback 1962) includes a dozen quotations from Austin but attributes this particular couplet (p. 17) to a “university poet unknown.” It also provides a metrically more accurate first line, “Across the wires the gloomy message came,” plus “not” for “no” in the second line.”
I do apologise that my original memory failure has led to this diversion, without, alas, revealing any hard truths.
“anything with Wiki in front of it is suspicious”
1871 rather than 1910, surely. In 1871 the PoW was seriously ill with typhoid, but recovered. He died of a heart attack in 1910 – but by then he was no longer the Prince of Wales.
I would have thought so, although 1902 is also a possibility when the King was ill with appendicitis. See what I mean about Wikix
For those who wish to follow the development of driverless vehicles in the the UK the *GATEway Project has just announced the opportunity to become involved in workshops and vehicle trials on its website.
https://www.gateway-project.org.uk/?utm_campaign=websiteregistrations_email#1
* GATE – Greenwich Automated Transport Environment
@Graham H
We are a long way from Green politics, but on a point of fact, the royal personage who had appendicitis in 1902 was no longer the Prince of Wales – we was the King – albeit uncrowned as his illness necessitated the postponement of the coronation.
@timbeau – my comment was carefully drafted.
@Graham H
The poem’s title was “On the illness of the Prince of Wales”, not “the person who used to be Prince of Wales until last year”
In which case 1871 is the only date
Going back to the discussion earlier – how many of the commentariat (myself included) are skeptical about driverless cars adding anything useful to London’s transport problems.
This story has just been published – clearly shows that these systems have a long way to go. It’s beyond me how Tesla could sign off their system if their cameras are unable to recognise/react to an obstacle as large as that broken down van.
http://electrek.co/2016/05/26/tesla-model-s-crash-autopilot-video/
The system in question was not a self-driving car, it was an extended version of cruise control.
Of course we can still be sceptical about true self-driving cars in London, but an accident resulting from over-dependence on a rather inadequate driver assistance scheme (which was admitted to be “no use in towns”) has little bearing on the matter.
Malcolm. It’s way more than a cruise control system. Not fully self driving as it isn’t linked the sat nav – but it is designed to operate the car autonomously along a road once the driver has got the car onto that road. Tesla advertise that the system will detect other vehicles and obstacles, and react appropriately. Tesla say the system works best in dense traffic with clear lane markings. That’s exactly the conditions in that video, yet the system failed to detect a large stopped van. I think it shows that the system isn’t as developed as Tesla would like us to believe.
Yes, I agree with all that. Clearly it’s a faulty system. And it show the dangers of lack of clarity of “who is in charge”, the human driver thought that the system knew what it was doing, and it didn’t. You are probably right when you say that Tesla is advertising it as doing more than it is actually capable of.
But I still maintain that a faulty system which is acknowledged not to work in towns (and now appears not to work on the open road) has little to tell us about the future viability, or not, of proper self-driving cars. The very limited success of Ader’s steam-powered aircraft (in 1890) was no indication that manned flight would not happen.
A real future problem with “self-driving” cars – people who put temporary signs up & then forget to remove them, or worse, some of them ….
3 years ago, I passed a temporary 30-mph sign on the A20 approaching London & slowed down … about 2 miles later, with no sign of any roadworks & no end-of-limit sign, I re-joined everyone else doing 50 mph ….
About 10 miles for Walthamstow, there’s a country road over the M25 & repairs were needed to the safety rails – temporary 30 signs, all well & good … until the repairs finished & they took away only the “end-of-limit” signs …..
An autonomous car would automatically obey those limits, wouldn’t it?
aGreg
Once we have autonomous cars on the road then part of the signing for road works or emergencies must include putting ‘virtual’ signs on an electronic map. Indeed, this ought to be the starting point as this will be the design for what appears on the road. The autonomous car should be able to rely on the virtual map rather than looking for real signs – though their systems should – pace Tesla – be able to detect and cope with broken down vehicles and the like.
Sadly, some of the OEMs designing autonomous vehicles still think they should rely on real road signs rather than virtual signs and are pushing for standard designs across Europe. This means actually identical, rather than similar. It seems to me to be a very strange and backward approach.
So every highway authority will need a standardised online presence? I can’t see that working. What about emergency works on water mains?
Relying on real signs rather than virtual does have the benefit (while there are still human drivers around) that everyone is singing from the same hymn sheet. But maybe even better, the cars should observe both the visible signs and the virtual ones, obeying whichever is more restrictive. That would be a more failsafe approach. Meanwhile, if there are virtual signs, what they indicate can also be shown on the dashboard of human-driven vehicles.
However, someone somewhere is going to notice that proper advantage is not taken right now of the virtual speed limits that we already have. My sat-nav can tell me, pretty reliably, what the speed limit is on the road it is on. But my car (and everyone else’s) completely fails to enforce that speed limit, which would I think be technically quite simple. (Though political dynamite to supporters of “road freedom”).
It might be helpful for the inventors of selfdriving vehicles to go on a study tour of the installation of ERTMS…
Malcom
the cars should observe both the visible signs and the virtual ones, obeying whichever is more restrictive.
Thus, driving round the roads of Essex @ 30, when the limit is 60, or ditto the A20, you mean … maybe not a good idea.
GH
Quite so
@Graham H
[Minor snip PoP]
“Transport for London’s Traffic Information Management System (TIMS) is a database of live and planned traffic disruptions in London, including congestion, traffic incidents, works, events, and other issues affecting traffic.”
http://tims.tfl.gov.uk/
It was also covered at the TfL data event I went to at the AWS loft.
https://tflnwp.files.wordpress.com/2016/04/aws-loft-event.pdf
See pages 27 (“Planned Road Closures – trialling a standardised way to share data Travel Demand Management (TDM) April 2016”) onwards for details.
@Malcolm
Page 32 of the above-linked PDF points to the DATEX standard, which is a Europe-wide way of communicating and exchanging traffic information.
“With the aim to support sustainable mobility in Europe, the European Commission has been supporting the development of information exchange mainly between the actors of the road traffic management domain for a number of years. ”
http://www.datex2.eu/
Enjoy!
Greg: One hopes that the self-driving cars will have a similar ability to human drivers, to deduce that a particular visible sign is bogus. They will have the additional “clue” from the virtual limit. My “more restrictive” stipulation would only apply when both limits are “plausible” (e.g. when limit has been recently changed between 30 and 40, and one version of signs has not been changed at the same time).
Perhaps Datex will one day be as successful as ITSO!
@Briantist/Malcolm
I commented previously (29 Apr 20:48) about experience with my 5 month old car fitted with sign recognition cameras. In that time I have found them highly effective. The system reacts to mandatory signs while ignoring advisory signs, even when the mandatory signs are temporary (road works, etc). Whilst not perfect the system is very, very close to that goal.
The smart cruise control also does a great job of maintaining an appropriate distance from the car in front, but does not (nor was it designed to) provide steering inputs. If the vehicle ahead changes lanes to overtake another vehicle you need to be sufficiently alert to a) confirm that there is a gap for you to overtake as well, and b) input the appropriate steering commands.
I have no doubt that with additional sensors and processing to detect and act on fixed obstacles the system could provide the necessary safe steering inputs.
To put this in context I have on several occasions driven the length of the M4 from Wiltshire to Jn 1 with no more than a couple of brake inputs from me.
Paying Guest: That sounds like a very satisfactory part-way stage to self-driving, a useful contrast to the Tesla report.
Just to clarify, do I understand that your car’s system does no steering whatever, leaving that to you? (Including the normal staying-in-lane). That sounds like a clearer division of labour. And that, in the case you mention, you have the choice of (indicating and) steering into the lane to the right (in which case the cruise control will remain the suitable distance from the car you were already following), or staying in the lane you are in (in which case the cruise control will slow down and fix its attention on the vehicle which is now in front of you – the one that the other car overtook). Is that correct?
Malcolm: Pretty much so – the driver has to do all the steering. It is a BMW 428i which is fitted with 4 (optional) systems: smart cruise control; lane departure warning; speed and other warning signs detection; and, a head up display (HUD). The first adjusts the spacing between itself and the car in front on a speed dependent basis, accelerating and braking as necessary (right down to stationary if need be). There is a pair of buttons to adjust the spacing to taste, offering 3 steps up and 3 down. The smaller the space the firmer the ride.
Lane departure warning vibrates the steering wheel if the sensors detect that you are veering out of the present lane without indicating. It also vibrates similarly if you start to pull out with another vehicle in the blind spot. Sign detection I covered in the previous post, but worth just adding that the current speed limit and any other restrictions such as no overtaking are displayed on the HUD alongside your current speed and, if you have so selected, distance to go to and direction of the next turn.
If the vehicle in front starts to pull out to overtake the cruise control system remains locked on to that vehicle, but should you remain in lane, after a fairly short interval the system switches lock onto the slower vehicle which can lead to some fairly firm (but not excessive) braking. On the other hand if you indicate and start to pull out there is no sign of the system swapping lock.
I have to say I have been extremely impressed. My expectations were fairly low when I specified the 2 systems, but the BMW engineers seem to have done a cracking job on optimizing the software. I had expected it to be pretty reasonable on motorways, but would not have been surprised if it had occasionally broken lock when encountering much bigger vehicles in an adjacent lane. On the other hand I had felt it might have problems coping with the vehicle in front going round a tight bend on non-motorway roads. Both expectations proved quite unfounded.
All 4 systems have proven their worth to me in making it easier to stay law abiding and in greatly reducing the stress involved in driving in traffic. I would definitely specify them again and see the combination as representing a very significant step towards fully automated driving.
It seems police were left scratching their heads today when a car was found abandoned on the tracks of The Midland Metro in Birmingham .
So if one thinks of Croydon Tramlink trams travel on shared streets then sections of track for trams only but with a road type surface and then on railway track and with the expansion of tram systems or even just existing level crossings then a way to ensure driverless vehicles don’t end up on rail tracks will need to be developed.
Then we have roads that end or run along a waters edge ….
Does anyone have any experience of automatic parallel parking? {I chickened out when specifying my current Golf….].
Melvyn: Interesting possibilities.
But we should remember that driverless vehicles, while they may be slightly more inclined to go “off piste” than the best of human drivers, will probably be much more consistent than homo sapiens, and thus be very much better-behaved than the worst humans. (Of which the one who put his car – or more probably someone else’s car – on the tram tracks was probably an example).
@Graham H
“Does anyone have any experience of automatic parallel parking? ”
Yes. Contrary to the systems described by Paying Guest, the parking system only does the steering. It beeps at you when you get close to the car in front or behind and tells you when to engage forward or reverse gear. It occasionally refuses to identify spaces I think it should get into, (and of course you have to decide whether the gap it has identified is actually legal, and not a disabled bay, bus stop, zebra crossing, or someone’s driveway crossover) but it only really makes a pigs ear of it if you are parking on a sharp curve, as it tries to line up parallel with both the cars fore and aft. As with any parking sensor, you also have to watch out for things projecting at a level above or below the level of the the sensors – skips are a particular problem.
@timbeau -I understood from the VW puff that it did far more than beep at you -my non-self-parking Golf does that (and it would be so much better if it didn’t also beep at you simultaneously when reversing -the in-car cacaphony is very distracting). The option we were offered appeared to envisage hands-off parallel parking.
@ Malcolm – your remark has made me wonder how long it will be before cyber criminals decide “stealing” or hijacking driverless cars and their passengers is a worthwhile endeavour. Given no system is really secure against determined people there must come a point when criminals will be able to track who orders a driverless vehicle, where it will go and whether it is worthwhile taking control of the vehicle so that it delivers the occupant to their waiting clutches so they’re robbed, beaten, killed, taken hostage. And no I don’t believe I’m scaremongering or being silly before anyone accuses me of that. The system security measures will have to be immense to protect users and the assets deployed in these systems.
Graham: timbeau’s description reads to me as if the parking is hands-off (or at least hands off the wheel), but not feet-off.
@WW -they will be hacked,as with drones. And we shall have empty cars laden with explosives directed to targets….
WW: I agree that the security measures will have to be good. But taking over a driverless car will just be one additional way of robbing-beating-and-taking-hostage. Plenty of traditional ways of achieving this already exist, sadly.
A few months ago, there was a serious fire in North Finchley High Road. For a couple of hours, the one-way southbound section was blocked by fire appliances. Traffic from side roads was being directed by police officers using hand signals and, on occasion, by actual spoken conversation with drivers. The police were ordering drivers to perform what would otherwise be an illegal right turn and to drive the wrong way up a one way street. This was also contrary to the existing road signs.
How will the emergency services order driverless vehicles to comply with such instructions? And when could any such systems be expected to be universally effective?
Anonymous at 21:04: There may be some form of override, operated by police or by someone in the vehicle (if any). But if there is not, or it cannot meet the needs of the particular emergency, then the police will have to act as they would now when there is no available diversionary route. That is, the vehicle will have to stay where it is until the emergency is over.
@Malcolm/Graham
Sorry if that wasn’t clear. Yes, the car steers for you. You still do the gears, throttle and brakes. The beeps are to tell you when to stop (as with normal parking distance sensors). A different beep depending on whether it has parked (to its own satisfaction) or whether it is necessary to engage the opposite gear to straighten up. (And another beep if you are going too fast).
All this about driverless cars & the steps towards them boil down to one simple question:
How “strong” do you want your AI?
[ See also a novel of an alternate future which didn’t quite happen, but came very close … “Rule 34” By C Stross. Set in the Edinburgh of a different “now” ]
@Malcolm
There is then the problem of the tailback unable to pass the d.c. (or indeed d.c’s), thereby causing gridlock in the area.
Surely the point is that there should be some universally agreed protocols before fully autonomous vehicles can be allowed in general traffic.