Whilst the potential time and productivity savings of autonomous vehicles have been covered in some depth, the first generation of models are likely to require humans to maintain vigilant so that they can regain control of the vehicle if required.
It’s an issue that has encouraged several pieces of research in the past couple of years. I wrote last year about a Stanford study that analyzed the ease with which drivers can regain control, and the results suggest it is far from straightforward. The study placed a number of test drivers on a track to find out, and the results suggest the transition isn’t as smooth as we might think.
“Many people have been doing research on paying attention and situation awareness. That’s very important,” the authors say. “But, in addition, there is this physical change and we need to acknowledge that people’s performance might not be at its peak if they haven’t actively been participating in the driving.”
This was also the topic of a second study, this time from the University of Southampton, which examined just how long it might take for control to go back and forth between man and machine.
The study saw a number of participants in a driving simulator, where they were driving at 70mph. Sometimes they were otherwise engaged in separate tasks, other times they were not. The researchers wanted to test the response time when they both ceded control, and regained it from an automated system, with intervention requests sent at random intervals during their drive.
Perhaps not surprisingly, whenever the participants were engaged in something other than focusing on the road, their response time was much longer. Even then though, the response time was anything from uniform, with some responding in just 1.9 seconds, but others taking up to 25.7 seconds. This represents a huge challenge for designers when building devices that cater for such a wide range.
Asleep whilst not at the wheel
Suffice to say, fatigue can also impact our response times, which makes the findings of a third study conducted by a German team especially worrying.
The study simulated the effects of fatigue on people traveling in both regular and autonomous vehicles. It found that people tend to tire much faster in autonomous vehicles than they do in regular cars, with the fatigue especially rapid if they are already tired to begin with.
This had a big impact on their ability to regain control of the vehicle, with tired drivers considerably slower than bored drivers.
Participants who were sleep deprived were tired just 20 minutes into the autonomous simulation, versus up to 50 minutes in manual vehicles. Similar levels were achieved in 35 minutes among well-rested volunteers, which was still well below well-rested manual drivers, who barely showed any fatigue at all after an hour of driving.
What’s more, when sleepy drivers did regain control, they often did so badly. For instance, they were slower to look at the speed display, whilst they were worse at avoiding oncoming obstacles. The results suggested a general lowering of awareness of the lanes either side of them.
It underlines the challenges the industry still faces when it comes to ensuring that autonomous vehicles are as safe as possible in a wide range of situations and conditions.