“You have way more agency to shape these things now, whilst markets are not yet developed, than you do when you have big, entrenched, multi-billion-pound-a-year businesses that you're trying to shape,” says James Hutt, co-founder of the Institute for Driverless Transport (IfDT), which launched in April. “It is much, much harder once something has already reached scale.”
That is one of the key messages behind the institute’s new research roadmap, published in Westminster on 28 October, to “improve the quality of conversation in the UK about the impacts of autonomous vehicles (AVs)”.
The research comes at a pivotal time for driverless cars in the UK. In London, customers can now take supervised rides in a small fleet of Wayve vehicles via Uber, with a licensed driver riding along. Waymo and Baidu are also preparing services in the capital, while Uber is collaborating with autonomous mobility firms Pony AI and Verne on trials in Manchester.
The future shape of AV fleets is far from clear, however. Earlier this month (1 November), London mayor Sadiq Khan called for a cap on the number of ‘robotaxis’, telling The Guardian that the consequences of AVs have “not been fully grasped by politicians”. Last month, a report by the London Assembly Transport Committee said there was “considerable uncertainty around many of the claimed benefits of autonomous passenger vehicles”, including over safety.
There has been “loads of really good work” to get AVs on the roads, Hutt tells Professional Engineering, including from the government – but he says “good debates, good policy discussions, good commercial discussions” are now needed around the wider consequences of mass adoption.
The roadmap explores open questions from five key areas: economics and productivity, social change, job displacement, security and geopolitics, and speed of rollout. Informed by six months of conversations with operators and regulators – as well as potential users, academics, logistics businesses, unions, local authorities, security specialists and civil society organisations – each question considers the potential consequences of AVs, and the work needed to secure theoretical benefits.
One question asks when AV rides might be cheaper than using a taxi or driving yourself, with current predictions in the 2030s. But, as Hutt says, “the difference between 2031 and 2039 is a really long time”. The institute wants to identify the types of modelling that could provide a more precise answer.
A debate perhaps noticeable by its absence is whether or not AVs will definitely reach widespread use. The roadmap works from the assumption that they will be used at a “reasonable level of scale” according to Hutt. Instead, it asks “what would need to be true for that to come to pass and what does the world look like in that scenario, rather than asking the question ‘Is this something we want to happen?’”
He continues: “The starting point is probably that there are really strong economic drivers and really strong safety drivers, which probably say something like we are already at the point where under operating conditions, these vehicles are provably less likely to cause injury and death. And lots of people, when presented with the choice to get somewhere without driving themselves, will choose to take that option.”
Costs are already roughly comparable to taxis, he says – but that is “definitely not the end game”, with further drops expected.
One major question, focused on the speed of AV rollout, ask which public bodies need to be considering the potential effects, and which need to be leading the response. “They will have consequences in other domains,” Hutt says. “For the built environment, and how you want to design towns and cities. They'll have security consequences, how you think about the resilience of our critical national industries. They will have economic consequences… It'll change access to work and what the labour market looks like.”
Hundreds of thousands of driver jobs could be at risk, for example, including many part-time and zero-hours roles that can offer convenient stopgaps between other income streams.
Labour, security and built environment decisions are made by national and regional governments, which Hutt says need to consider the implication of AVs. “We're really trying to push the wider public sector to go, ‘Oh, hang on, this does actually have quite significant impacts.’”
TfL is a great asset in the conversation, Hutt says, but not every city has equivalent expertise – and, he continues, the window for meaningful engagement could shrink, especially if widespread AV deployment extends to public ownership.
“Once you're in the widespread public ownership world, TfL do not have the regulatory power in order to determine whether people accept this. And they, in some senses, have a really good window now to shape market evolution,” he says.
“Regulators, transport operators and civic society have a lot of opportunity to shape things at the moment. We're at moments of maximum leverage for determining how these things evolve… London Assembly have gripped that really quickly. That's not necessarily true across every element.”
Congestion is another key topic for consideration, with opposing views on whether it will increase or decrease. AVs being selected for otherwise non-car journeys, combined with the miles driven without passengers – known as ‘deadheading’ in the US, where deployments are more mature – could lead to an increase, whereas the potential for less private ownership and AV integration with public transport networks could do the opposite. Clarity is clearly needed.
The next step for IfDT will be finding collaborators to fill gaps in the research, to build an evidence base for politicians, regulators and potential investors. Engineers will have a key role to play, Hutt says, thanks to the profession’s use of systems thinking.
“Part of what has made technology – including autonomous vehicle software in there – able to scale so substantially in recent decades has been this real focus on modularity and an almost complete disregard for everything that exists outside of your own function,” he says.
“Engineering, my sense is, has a slightly different mindset, which is understanding how a function and a component fits within a wider system… this is a systems engineering problem and we need to think about how the effects we can create locally with that component create impacts in the system more broadly, and then decide whether we want that and design in our tolerances, work out what we're designing for. That feels to me like a conversation that I'd hope more engineers were keen to be part of.”
Want the best engineering stories delivered straight to your inbox? The Professional Engineering newsletter gives you vital updates on the most cutting-edge engineering and exciting new job opportunities. To sign up, click here.
Content published by Professional Engineering does not necessarily represent the views of the Institution of Mechanical Engineers.