Will human behaviour hold back autonomous vehicles?

Jamie Stevenson
London 51° 30' 23.112" N, -0° 7' 37.956" E" E

What do you really know about autonomous vehicles? Many people, and this is no insult, probably know little beyond 'they drive themselves'. According to a recent study, this lack of understanding may prove decisive, and act as a key barrier to the adoption of autonomous vehicles, which could slow what has so far seemed an unstoppable juggernaut.

Human beings aren't typically known for their love of change. Indeed, even the most subtle of disruptions to the status quo can cause quite an upset -- just ask Coca-Cola. So it's no surprise that something as fundamentally different as autonomous vehicles may be met with resistance.

The study by TRL looked at how human drivers would adapt to a world of autonomous vehicles and, unsurprisingly, it wasn't a seamless transition. If we couldn't adapt to a new soft drink, after all, a driverless car was always going to be a struggle.

Yet, there were signs of progress.

Can't adapt, won't adapt

The first few years of autonomous adoption will see traditional, manually driven vehicles sharing the roads with fully- and semi-autonomous vehicles. Of course, a future has been envisioned in which the roads are solely occupied by fully autonomous vehicles but, in order for this to work, human drivers will need to adapt to this transitional period.

Being human, we are subject to human error -- we drive erratically, cutting into lanes quickly or overtaking at times when it isn't ideal. This, however, could cause problems for autonomous vehicles -- if they're programmed to rigorously obey the rules of the road, they could grow confused by drivers treating them the same way they would another human.

The study used a series of driving scenarios in a simulator, placing participants in a 3D virtual replica of the Greenwich Peninsula to see how they'd drive on the roads, with a changing number of autonomous vehicles in each scenario, which included junction driving and overtaking tasks.

TRL found that drivers struggled to anticipate how an autonomous car would behave, and often couldn't recognise them. This of course, meant drivers simply continued to behave in the same way that they would if the road was entirely populated by other human drivers.

This lack of adaptation is a problem, not only for the autonomous vehicle, but for drivers themselves, who could benefit from understanding autonomous vehicle behaviour to ensure a smoother drive. Indeed, by recognising an autonomous vehicle and understanding the way it manoeuvres the roads, human drivers could more confidently traverse the roads.

It would also have wider repercussions - If human drivers fail to adapt, the risk is that widespread autonomous adoption will remain elusive for a few more years, as road safety and traffic flow could become compromised.

If this is the case, and more accidents occur, consumers will be less likely to trust autonomous vehicles -- and trust is one of the major barriers to autonomous adoption, according to a recent HERE survey.

Thankfully, however, the study did find tentative signs of progress.

The more the merrier

The study found that some drivers' behaviour changed as more autonomous vehicles were introduced into the simulator, pulling out into smaller gaps, or moving through traffic more smoothly.

This, while tentative, is encouraging -- the coexistence of autonomous and manually driven vehicles is absolutely imperative in ensuring the transitional years, when more autonomous vehicles are gradually introduced, aren't hugely disruptive.

While some companies have attempted to adapt the autonomous vehicle to better suit drivers and pedestrians (check out the smiling car, for example, or how HERE is looking to humanise the autonomous vehicle), perhaps the more pressing need is for humans to better understand a technology that's set for ubiquity, even if there are growing pains.

How do you think drivers can better understand the autonomous vehicle? Let us know in the comments.

Topics: Automotive, Autonomous cars, Autonomous World, Features

Comments