Services giant Capgemini has announced a joint research project with the University of Oxford looking at trust and safety in driverless cars and other autonomous vehicles.

Specifically, the new collaboration will examine the safety and human elements of our interactions with artificial intelligence (AI) and autonomous systems as they become more commonplace in cars, vans, trucks, and other transport systems.

According to an announcement from both organisations this week, the aim is to create thought leadership, assets, and future services that answer the question, ‘How can technology help solve the key challenges of a more intelligent industry in society?’

In other words, what are the trust and safety implications of systems that we have previously trusted becoming more intelligent and self-determining?

Initially, David Jackson, CTO for Product & Systems Engineering at Capgemini Engineering, will work with a team of researchers led by Marina Jirotka, Professor of Human Centred Computing at the University of Oxford.

Their project will be called, ‘Being safe, feeling safe: Designing, measuring and evaluating underlying factors determining safety and trust in autonomous vehicles’.

“The impacts of novel technologies on societies and individuals can sometimes get lost in the excitement of new tools and innovations,” explained Jirotka.

“We believe this project with Capgemini gives us a great opportunity to really examine how we can keep human needs and interests at the forefront of research and development.”

“The wide adoption of innovations such as autonomous vehicles and aircraft needs people to trust them, so we are proud to tackle this challenge with the University of Oxford,” added William Rozé, CEO of Capgemini Engineering.

“This new initiative will further strengthen our strategic research program on Intelligent Industry.”

 

Sector briefing

So, what’s behind the project? Here’s Transform Industry’s concise primer on this sector.

The race is on worldwide to develop better driver-assistance systems and, in time, full autonomy in road vehicles – as well as in shipping and aerial transport systems, such as drones and air taxis.

In the UK, for example, self-driving buses are being tested in Scotland and Milton Keynes is hosting driverless car trials.

The big-picture aim is to shift transport from a driver/ownership model to one centred on on-demand passenger services, with a broader context of improved safety and sustainability.

Studies have consistently shown that human error, rather than technical failure, is the overwhelming cause of road deaths. For example, a 2016 study by the US National Highway Transportation Safety Administration found that human error accounts for up to 96 percent of all car accidents, while road crashes are the leading cause of death among American teenagers.

However, this article in the Atlantic cautions that blaming human error for most accidents is an oversimplification.

This is because citing driver error often ignores a broad range of other factors: poor visibility, incorrect signage, bad weather, dangerous road conditions, and more, some of which would equally apply to autonomous vehicles.

Meanwhile, the millions of mainly petrol or diesel-fuelled vehicles on our roads daily have a massive environmental impact. In the US, for example, road vehicles account for one-fifth of all carbon emissions.

Technologists believe that smart, electric autonomous vehicles, summoned when needed, would mean far fewer cars on the road.

However, this may be a logical fallacy. According to UK motoring organisation the RAC, the average British driver-owned or operated car is stationary – ie parked – 90 percent of the time.

So, while a possible future shift to autonomous transport at scale would undoubtedly reduce the number of vehicles overall (once car ownership stops being a necessity) there may be more vehicles in our cities, not fewer, because autonomous cars would be in service much of the time.

The safety, trust, liability, cultural, and employment aspects of a possible large-scale shift to autonomous vehicles are extremely complex.

This is particularly true in the US, where being a driver is the most common job in some states, and vehicle ownership is both a necessity (with long distances between many towns and cities that are poorly served by public transport) and a status symbol.

The immediate future of driver-operated cars, vans, and trucks mixing with autonomous vehicles on our roads and in our city centres is likely to be the most challenging and risky.