“It will be business class travel”

I think autonomous driving has the potential to completely transform our relationship with cars. Your commute to work, for example, could become closer to the experience of business class on a plane, where you have the ability to work or enjoy yourself in your own personal space without worrying about getting there.

Our Drive Me trials – which will put real families behind the wheel of autonomous cars – will focus mainly on commuter routes, because we want to find out how people interact with a self-driving car in daily life. The commute to and from work – where traffic is often slow-moving or stationary – is also where there is the greatest potential to turn wasted time into productive time, or relaxing time. We want to get a deeper knowledge of what people want to do in a vehicle when they don’t have to drive it, so we’re looking at services that support working and entertainment.

“Autonomous driving will change transport as a whole”

The relationship between autonomous cars and other road users is a fascinating one. With the Drive Me trials we are not going to have special markings on the cars because that would mark them out as a novelty – it might make other road users behave in a different way. The whole point is to see how autonomous cars work in everyday situations on real roads. In the research we’re looking at the secondary effects – new risks that aren’t easy to foresee. It always pays to expect the unexpected.

“A unique perspective”

Across the auto industry a huge amount of work is being done on autonomy, but we’re deliberately taking a different path with Drive Me. What’s unique about Drive Me is that it’s about how customers use autonomous cars in the real world. That means encountering the kind of everyday problems that might not be very exciting, but which we’re going to have to solve alongside the big ones. How do we keep the car’s sensors clean when the roads start to get dirty? How do we make the hardware as effective as possible? They’re not dramatic things, but they’re important – and Drive Me is giving us a unique perspective on how autonomy will be used by real people.

We call it ‘human-centric automation development’. This means testing developers’ assumptions of customer experiences with autonomous driving. It helps the developer’s learning curve and optimises it around how users can, want, and need to use the product, rather than asking them to change. We see it as multi-stage problem solving.

Basing our work on people means that we study both subjective and objective experiences. Subjective experiences like the opinions and descriptions of their experiences. Objective experiences like how they use the interior space, how much attention they give to driving, how much hovering on the steering wheel and pedals, how much they want to be involved in complementing the driving behaviour.

“We need a solution that works for everyone”

We’ve learnt a lot already and one of the surprises has been the variety of ways people respond to autonomous driving. There are many different attitudes and different levels of trust. Some people are very hesitant – they don’t want to relinquish control, even in a situation like slow-moving traffic where the car can operate itself without any drama. More surprising is the fact that some people will just jump in and straight away be too trusting of the system, assuming that it can do more than it actually can at this stage. These are different challenges, but we have to create a solution that works for everyone.

The important thing is to make everything predictable, to know when you are in control and when the car is in control. When we look to unsupervised automation, where you delegate control to the extent you’re free to do something else and the car does the driving, we have to make sure the car is very predictable. It will be naturally cautious, like a good driver, but not timid. Once you’re comfortable that the car knows what it’s doing then you’ll be able to relax and use your time in different ways, working or doing something else.

“The important thing is to make everything predictable, to know when you are in control and when the car is in control.”

“The right way is to put people, not technology, first”

When we talk about autonomy we mean different things. At one end there is piloted driving, the assistance that you can get in a Volvo right now. At the other is what’s called Level 5 autonomy, the point where a car can drive itself over any road that a human can drive on without any driver input.

Obviously it’s going to take time to get to that stage. Look at the way the cell phone evolved – it started off being a luxury item, and one that only worked in big cities and densely populated areas. Then it spread out as coverage improved, and prices fell, until it became almost ubiquitous; now we notice when it doesn’t work rather than when it does.

I think it will be the same with this higher-level autonomy – it will start in cities and other areas where it is needed most, and then it will spread outwards. Eventually I think it’s likely that every road will be capable of supporting autonomous driving, but there isn’t going to be a big bang where it suddenly happens. The important thing is that we are approaching things in the right way by putting people, not technology, first.

Creating the future