Bridging the trust gap for autonomous cars

Autonomous cars are very exciting – particularly to the sort of tech-savvy early adopters who read HERE 360 – but among the population as a whole, there’s still a long way to go before most people will trust a machine to do their driving for them.

In recent HERE research with 2,000 car users in the US and Germany, we found that acceptance varies a lot among different groups of people. 49 percent agreed that they’d like to use autonomous vehicles, whether owned or as a service, when they become available. And just 19 percent expressed that agreement with certainty.

Looking into the reasons given for reservations about autonomous cars, the main issues are around retaining control. People want to feel that they are in control of their own destiny and, safe or not, they think autonomous cars will reduce that control.

This anxiety about control is rooted in several factors.

No one is average

People consistently overestimate their own driving ability. In our survey, 73 percent of drivers said that they’re better at it than most. Of course, that can’t possibly be true, but mathematical possibilities do not always have a bearing upon people’s beliefs. Given this uncanny faith in their own abilities, it’s not surprising that many of our confident drivers can’t imagine that a machine could drive as safely or anticipate hazards as well as themselves.

Linked to this, around half (48 percent) of our research participants said they don’t even feel comfortable being a passenger when another human is driving. Given the low level of trust drivers give to other people, the chances of an untried machine reaching an acceptable level of trust thus becomes rather slim.

Skip version one

People are also concerned about using first-generation technology. There’s a feeling that early iterations of any technology have many flaws that need to be ironed-out over time. That they’d be safer waiting for autonomous car version 2.0 to arrive before making the jump.

Other concerns around technology also centre on this fear of losing control. Around 30 percent of drivers said that anxiety about someone potentially ‘hacking in’ to their vehicle and taking over its control was among their top reasons for not using an autonomous car. For autonomous cars hired as a service, it was the most popular reason people said they would hesitate before stepping into one.

So what can be done? Well, we know that experience builds trust and so, as more and more automated features are built into new cars, the more people will discover that there’s little to worry about when your vehicle does something to help you. A phased introduction to automation is already happening.

Indeed, the research showed that drivers who own cars with at least one ADAS feature (e.g. parking assist, adaptive cruise control, proximity detection, etc.) are, by far, those most likely to embrace the idea of an autonomous vehicle. And the majority of people who have experienced ADAS features wish those features would do even more.

The number of cars that have ADAS features is already large (11 million in Europe, USA and China in 2016). So, the simple passage of time and the inevitable spread of technology, as more people get to experience some form of vehicle automation in action, is likely to swing opinions towards a positive reception for autonomous vehicles when they are ready.

There’s also a role for education. Very few of us have seen an autonomous vehicle on the roads, let alone experienced one in action. Their capabilities and specifications are still in a state of rapid evolution, so it’s hard to say very much that’s factual about their capabilities. However, we can be reasonably certain about the safety, speed and convenience they will bring to almost everyone, and that message deserves to be amplified as loudly as possible. And further, we believe that as established and trusted car brands come to unveil their autonomous offerings, their pre-existing reputations will be transferred to their new vehicles.

Time on our side?

It’s also true that there are non-technical issues around autonomous cars that need to be settled before wider trust can be expected. The regulatory frameworks that will be required to allow autonomous cars to navigate our roads at scale are still in their infancy – and may take many years before consensus is reached. Similarly, the ethical dilemmas which arise when autonomous cars need to choose between bad outcomes haven’t yet received compelling answers from experts.

Ultimately, in a way, there’s a happy coincidence here. We can see that time, learning and first-hand experience of automation in action will be required to establish trust in autonomous vehicles. Simultaneously, while the fundamental technology is at a well-advanced stage, there still exist sufficient hurdles to ensure that it will be a little while yet before people will be asked to exercise that trust.

What do you think it will take to bring the masses around to the idea of accepting autonomous vehicles as a force for good?