Thanks to all our panel and everyone who sent in questions
Many thanks to our panel and readers for taking the time to join us today. We hope you got as much out of this discussion as we did and we look forward to exploring the points raised in further depth over the coming weeks.
If you want to get in touch to send transport story ideas, please email olivia.boyd@theguardian.com or tom.levitt@theguardian.com
Let's get philosophical
Final question for the panel:
Many predicted the death of the book thanks to tech advances but that is yet to happen. Are predictions around take-up of driverless cars sufficiently taking into account habit, taste and human nature?
Here is a selection of replies from our panel:
A very important questions that gets asked only infrequently.
My answer is no. This is in part because the developments around autonomous vehicles are strongly 'supply' driven. It is actors from industry -- vehicle manufacturing, IT and robotics -- and government that are the driving force behind the development. In the process, the full diversity in the way that people use and relate to cars is not taken into account.
No question some people will still buy cars; driving can be fun and I think if you are a Ferrari fan you'd not want to give that up. But for the daily grind in crawling traffic through our cities, most people will choose to share AVs and spend the time they'd devote to driving to something more productive. Reading that book perhaps?
Safety – are we asking a lot of our computer systems?
Back to the subject of safety, Nick from TRL says:
Good question. I'll start by saying that human drivers are a contributory factor in virtually all crashes. 1.3m people dying on roads around the world each year is far too many.
An automated vehicles that is never distracted, tired, impaired by drink/drugs and will make better risk judgements is likely to greatly reduce that number.
However, the average distance driven between fatal collisions in the UK is 180 million miles. Expecting a complex computer system to do better than that is asking a lot - but we will get there...
Fiona adds that Heathrow ran a successful trial at its head office earlier this year, where AVs interacted safely with passengers and other road users.
Yes autonomous vehicles still have a way to go to achieve their aspired safety levels; however there is already evidence that technology such as driver assist is already making the roads safer. In an airside environment we would expect to see improved safety as a result of autonomy. In fact earlier this year we trialled autonomous vehicles at our head office, interacting with passengers and other road users safely.
Infrastructure costs
GSB reader Adele White asks:
what does the panel estimate the costs of infrastructure will be to enable driverless cars to operate on our roads? How will this money be raised in the current climate of austerity?
Nick Reed from TRL says:
It depends - there might be some infrastructure developments that facilitate automated vehicle deployment but these are likely to be costly and may get overtaken by technology in short order.
The other option is to focus more on the vehicle technology so that infrastructure changes are minimal. This may make vehicles initially more costly but that will rapidly change as numbers scale up.
Worth also mentioning the importance of the digital infrastructure (3D maps) and communications infrastructure (5G network) that may be even more important than any dedicated physical infrastructure for automated vehicles.
Tim adds:
This is an excellent question and ne of the elephants in the room. I am not aware of published academic research that has arrived at plausible estimates of the infrastructure costs involved. However, there must be studies in the policy community that have looed at this issue. Perhaps some of the other panellists can fill us in.
Two key reasons why it is difficult to estimate the infrastructure costs are that:
a) this will depend on which set of technologies will become dominant and we are not yet at that stage of development (i.e. there still are competing technological designs at the moment)
b) it is unclear what the distribution of responsibilities will be between government agencies -- both national and local levels -- and the private sector. Government is increasingly looking at the private sector to make substantial contributions, but private sector investors are often strongly risk aversive and may well prefer to wait with making substantial investments in new infrastructures. Private sector actors may be looking at government to take on the early risks. So it is not unlikely that we'll end up in a period of finger pointing at each other.
Building liveable cities
Key question for the panel:
A report by McKinsey, due out today, notes that as driverless vehicles make it more pleasant to travel by car they could increase demand for mobility and even encourage low-density development. What can urban planners and others do to make sure AVs genuinely make our cities more liveable?
In response, here are some highlights from our panel:
Guardian contributor This is key. More consideration needs to be given to the integration of AVs and our cities – the street layout, people walking and cycling and public transport. But the starting point needs to be the liveability of the city and the health and happiness of the people in that city - not the technology that can permit driverless vehicles.
Making it easy, economic and pleasant to *share* transportation seems the key. The economics are reasonably obvious but there could be a taxation policy that penalises single occupancy vehicles at certain times. Beyond that, seems hard to believe that people would welcome longer journeys, most people I know are rushing from thing to thing and I don't think that will change!
Like Stan, Tim and Susan highlight the role of policy:
AVs would need to be complemented by elaborate policy packages to minimise 'rebound effects' such as increased demand for vehicular travel and low-density developments. Examples of measures in such packages would be pricing car use rather than car ownership and tighter regulation & pricing of parking, and on the positive side, more (financial and regulatory) support for various forms of car sharing -- from conventional car clubs to informal peer-to-peer sharing -- and further investment in attractive infrastructures for cycling and walking.
Guardian contributor Major policy changes will be required to prepare and respond to driverless cars - but this has not been the focus of attention. We do not want to make the same mistakes that we did when cars first arrived - when it took another 40 years for the driving test to be introduced during which time there were significant road fatalities and casualties.
Where's the evidence?
Tim Schwanen from the University of Oxford says there’s “very little robust evidence” to substantiate claims that AVs will trigger a widespread shift away from car ownership, not least as a result of the involvement of conventional vehicle manufacturers in AV development:
This is another area of uncertainty. The expectation is that autonomous vehicles (AVs) will trigger a widespread shift away from car ownership towards various forms of car sharing. However, there is very little robust evidence to substantiate these claims.
And I am quite sceptical. One of the things that is happening with AVs is that conventional vehicle manufacturers are becoming increasingly central to AV development, and their business models are premised on individual/household car ownership. Given the huge path dependencies and sunk costs in the car industry I don't see this change very rapidly.
Safety first
Stan Boland, CEO of FiveAI, says that safety will shape how the AV industry evolves:
Safety is the gating factor. It'll take a broad sensor suite and a lot of compute to achieve necessary safety levels. That means the additional cost per vehicle will be high - a problem for an owned model but not for a shared service model. In fact the cost per mile could be 1/10th of that of an Uber driver. So we think safe urban AVs will accelerate the trend to shared private mobility.
How will autonomous vehicles (AVs) change human behaviour?
Depending on how things evolve, Susan Claris from Arup says AVs could cut the number of private vehicles on the road. Equally, they could see a reduction in use of public transport and in active transport.
Guardian contributor This comes back to uncertainty again. If there is shared ownership, this could see reduced use of private vehicles - with benefits in terms of reduced need for parking etc as vehicles will be better used. But if there is not shared ownership, there could be some unexpected consequences, such as increased demand – as journey time could become productive time journeys could get longer. And there could be reduced patronage on public transport and a reduction in active transport such as walking and cycling.
Fiona Carleton, head of operational strategy at Heathrow, adds:
We fully expect that in megacities such as London which generates more than 50% of Heathrow's demand that the sharing economy will absolutely impact on models of car ownership and how our passengers and colleagues get to and from the airport. However, we are also expecting populations to continue to rise, especially in and around London, and so we need to be aware of the potential balancing effect of population growth negating falling individual car ownership.
Uncertainty or inevitability?
Susan Claris, associate director at Arup, echoes Tim’s point about uncertainties. However, Nick Reed, academy director at TRL, says:
Yes, I would say it is now inevitable that automated vehicles will play an increasing role in transportation.
This will give us more options in how we move people and goods leading to greater efficiency and easier mobility.
Driverless vehicles are already a reality (DLR, Heathrow POD) - the only question is the complexity of the environments in which they operate. Within two years, I think we will start to see commercially viable automated bus services in specific locations.
The debate has begun. First up...
Is widespread use of driverless vehicles now inevitable? How soon will we see them? And where?
In response, Tim Schwanen, director of the Transport Studies Unit at the University of Oxford, says:
I don't think widespread use of driverless vehicles is inevitable. There are many uncertainties that may derail current trajectories and defy current expectations.
These uncertainties are partly technological in nature but extend into legislation, insurance and public support.
Timescales are also uncertain. What we have seen over the past 2-3 years is a dramatic shortening of the horizon of rapid diffusion from, say, 30 to 10-15 years. Many developments are speeding up indeed, but it will take quite some time before we will see fully autonomous vehicles being adopted widely. It is difficult to indicate time scales but my hunch is at least 15-20 years from now.
As for location, this is an interesting point. Use of driverless vehicles is easiest on the motorway network, but much of the development and government interest is now focusing on urban mobility. This is also the most difficult type of traffic to automate because of its huge complexities.
Some background reading to kick off the debate
Check out our recent interview with Paul Newman, the co-founder of UK driverless vehicles startup Oxbotica.
Updated
See what our columnists have to say about driverless vehicles
Guardian columnists Jackie Ashley and Paul Mason explore the impact of driverless vehicle technology on society.
Updated
What we'll be discussing
The technology and interest around driverless cars is developing fast. The co-founder of UK startup Oxbotica has said the use of autonomous vehicles for shopping, short trips, airports and warehouses is imminent.
By 2025, auto manufacturers are predicting that at least half of today’s drivers are unlikely even to want to own a car.
But how close are we to seeing widespread use of driverless vehicles? Can they cut congestion and make our roads safer? Will they make our lives easier, and driving more accessible? And how and where will their use emerge first?
Join us and our panel of experts in the comments section of this page on Tuesday 20 June, 12 noon-1pm (BST), to discuss these questions and more.
Updated
The panel
- Nick Reed, academy director, TRL
- Susan Claris, associate director, Arup
- Tim Schwanen, director, Transport Studies Unit, University of Oxford
-
Fiona Carleton, head of operational strategy, Heathrow
- Stan Boland, CEO, FiveAI
Updated
How to join and ask a question
Make sure you’re a registered user of the Guardian and join us in the comments section below, which will open on the day of the live chat.
You can send questions for the panel in advance by emailing olivia.boyd@theguardian.com or tweeting @GuardianSustBiz using the hashtag #AskGSB
Updated
A good question to end on Olivia. I think the human element of driverless cars has not been fully explored. Perhaps an example of the future being over-sold and under-imagined. Our future view of cities has remained roughly the same for many years and that imagined future is typically a Metropolis/Bladerunner future. The reality is - or should be - a focus on healthy streets, vibrant centres and happy people.