This month, the Government published evidence on issues around the potential deployment of self-driving services in the UK. Our Head of Research and Service Design, Ed Houghton, who presented to the Transport Select Committee in March looks at how far we have come in terms of public attitudes – and where we are going next with the DeepSafe consortium.
Self-driving technology has the potential to radically change how we move around our towns and cities. It shows so much potential that the UK Government considers deployments possible by 2025. In August 2022, it stated that it intends to move forward with defining a regulatory, legislative and safety framework to make deployment a reality. Industry is waiting for these developments to enable rapid commercialisation, and as Government looks to define its place on the global stage as an AI safety superpower, it stands to reason those self-driving services – based on AI – should for the basis of this next leap forward.
When we have worked with industry on the development of safe self-driving services, we have seen how UK companies see the potential value of the technology to our neighbourhoods. Our work has always focused on bringing in under-represented public voices into the discussion about the future of our transport system – and our work with industry leaders such as Oxa and DRISK have enabled us to explore, in detail, the opportunities and challenges the public sees when they consider self-driving tech. This has given us a deep understanding of where we think development should move next.
“Societally, we place driving licences, gaining the freedom of driving, as almost a part of our identity... The vehicle becomes part of how we gain freedom. The challenge that industry faces is that you have to disconnect the vehicle, the object, from somebody’s identity.”
One key issue we pick up on is the view that technology-based solutions are too often built without broad engagement. As researchers we know that without effective engagement with diverse audiences, new technology solutions can be severely limited in their utility to the end user. For something as potentially transformational as self-driving technology, we think the risks are too high to deliver technologies to market technologies that have not been extensively validated with the public.
That’s why, when the UK Parliament Transport Select Committee sought evidence on the evolution of safe self-driving services, we were keen to share our insights from our public engagement work. The Committee’s report, recently published, highlights the outcomes of their inquiry, and rightly showcases the challenges and opportunities facing the development of commercial services.
Our research during trials, which have included live public engagement, have shown overwhelmingly that the public considers safety a key priority, but many still lack knowledge of the reality self-driving technology, and what it might mean in practice, on the road:
26.8% would feel confident using an AV tomorrow if it were possible to do so. Over half would not (55.1%). The remainder are undecided (18.1%).
3 in 10 (29.9%) believe that self-driving vehicles will be safer than traditional vehicles, whilst 44.2% disagree. A quarter (25.9%) are undecided.
But our study showed that demonstrations can make a big difference in reassuring and shifting public opinion.
Live trials improved perceptions of safety by 15 percentage points: before the trial, 68.3% agreed that AVs would be safer than human driven vehicles, whilst after the trial 83.6% agreed, an improvement of 15 points.
Trust in self-driving vehicles is low, but a large minority is yet to be persuaded: findings from our national survey show almost a third (32.5%) think self-driving vehicles will be trustworthy, whilst two in five (43.8%) do not. Almost a quarter (23.6%) are undecided.
It’s important that safety and demonstrating trust are key outcomes for future services. If industry is to drive adoption and acceptance, designers must prioritise these factors in their practice. How we can do that en-masse with the wider self-driving ecosystem? That is the topic of our new study with the DeepSafe consortium, which kicked off recently.
Through DeepSafe, we are working with experts at drisk.ai, Claytex, RF Pro, and Imperial College London to test and validate AI responses to hard-to-predict edge case scenarios, and using these to demonstrate the potential of the technology to the public. We believe this process will not only prompt engagement and discussion on the value and potential of the technology, but also surface insights that can help self-driving AI developers ensure the technology they’re developing is human-centred.
We think this will go some way to supporting some of the findings from the UK Parliament Transport Select Committee, as they rightly called for a focus on safety as a priority from government. As the report highlights, “Safety must remain the Government’s overriding priority as self-driving vehicles encounter real-world complexity.” Understanding this complexity, and engaging the public in validating self-driving AI responses to it, is exactly what the deep safe project is looking to do.