Self-Driving Cars’ Success Will Hinge on Psychology

As much as we may want to let go of the wheel, we still need a bit more convincing that it’s really okay to do so.

Chris Rockwell

May 13, 2016

6 Min Read
Self-Driving Cars’ Success Will Hinge on Psychology

Self-driving cars are picking up speed and they’re giving all-new meaning to the term automobile.

But don’t hang up your driving gloves too fast. Although new technologies promise safety, productivity, fuel savings and the freedom to truly take your eyes off the road, human interaction still will be at the heart of these future mobility experiences.

People, not technologies, ultimately will determine the successful adoption and use of self-driving systems.

Technology isn’t the main obstacle in the path of self-driving systems. It turns out that psychology – specifically, our ability to trust these autonomous technologies – might be the biggest roadblock.

At a recent TU-Automotive telematics conference in Detroit, industry attendees agreed overwhelmingly that driver experience will be the most important hurdle to the adoption and success of assisted and fully self-driving systems.

According to a new survey from AAA, only one in four U. S. drivers would trust an autonomous vehicle to drive itself with them in it.

And while advanced-driver-assistance systems are desirable, a report by the University of Iowa found that most drivers are uncertain about them and how they work.

Forty percent of drivers surveyed reported their cars had acted in ways that startled or surprised them when such systems activated.

As much as we may want to let go of the wheel, we still need a bit more convincing that it’s really okay to do so.

For the foreseeable future, we’re not talking about putting the car on 100% autopilot. People still will play a key role behind the wheel, and, in the short term even may need to be more involved.

Current autonomous Level 2 (Combined Function Automation) and early Level 3 (Limited Self Driving) classes of assisted-driving technologies allow drivers some degree of less-than-perfect attention to their driving as the vehicle is monitoring closure rates and braking or lane-keeping. Ironically, this means the driver still needs to stay engaged with the act of driving and monitor the assistive technology to maintain full awareness. In essence, we must monitor the monitor.

In other words, the driver interface won’t go away. Rather, Cockpit 2.0 needs to be re-conceptualized with the human psyche – our past experiences and our future fears and worries – in mind.

It should deliver both on the promise of self-driving convenience and the ability to control the state of the self-driving system in a way that gives drivers the freedom and confidence they are being cared for.

Unlocking Trust

Trust is obtainable. We trust the abilities of people we don’t know when we step on buses, trains and airplanes. We trust the safety of machines when we step on elevators and escalators.

Recent observations in Silicon Valley have shown pedestrians are often more willing to step out in front of a moving unmanned Google Car at a stop sign than in front of a vehicle controlled by a human in the driver’s seat.

Trust is possible. The question is: how can new autonomous-vehicle interfaces be designed to communicate and deliver it? It requires the design of new cockpits that provide control, awareness, familiarity, feelings of protection and collaboration.

When it comes down to it, we trust technology to predictably follow rules, and we trust people to make sound judgments based on incomplete information, which is often the case in situational driving.

Sensors in our cars are very good at accurate determinations of speed, distance, close rate and stopping distance.

We’re getting more comfortable with systems such as smart cruise control, because the function is straightforward: maintain speed, monitor and adjust close rate, brake to avoid collision. Again, rules.

But what about passing on a 2-lane road? This requires contextual judgment. We also need to trust the car is going to operate the way we think it’s going to. When drivers experience stress – the gap between how we think something works and how it actually does – their sense of control diminishes.

Under emergency situations, experience stress is magnified resulting in even more difficult or dangerous control situations for drivers.

New designs can help us with that, foreshadowing the safety of maneuvers before they’re conducted, giving us confidence the technology is secure. But self-driving systems can’t be imposed upon us. At all times, we must sense we have control over when and how much automation technology is being used.

Maintain Awareness

While self-driving and assisted-driving systems give us permission to get distracted (allowing us to shift attention to non-driving activities) we must still be aware of the status of automation and the headway scene.

Because turn-taking is inevitable, it must be predictable. Studies suggest it takes 5 to 8 seconds for the driver to reassume full control in a self-driving car situation. At 70 mph, that’s 800 feet – almost three football fields – without clear control of a moving vehicle.

If we can increase situational awareness and drop that takeover time to 1 to 2 seconds, drivers will be much more likely to adopt and use autonomous or assisted features. Some of the best technologies coming to market are creating smooth, predictable transitions between self-driving systems and user control.

Design of autonomous systems must always communicate capabilities and status to the driver and help drivers maintain an awareness of the headway scene for smooth transitions.

Support Collaboration

Self-driving and assisted driving systems should enhance our driving ability, creating smooth transitions and predicting when we need them to take over. These systems must be in constant communication with us to eliminate surprises. Humans loathe ambiguity and unpredictability in interactions, so if a system acts erratically and is hard to understand, it will frustrate us.

Systems can’t fight our inputs, but instead need to support a symbiotic man-machine pair.

Feel Familiar

Automation must feel familiar and clear. Using known display and control formats and limited new terminology will help drivers in their changing role as sometime-drivers in the future.

Audi’s telescoping steering wheel moves away from the driver and shows a clear transition of driving state. Drivers simply can pull up the steering wheel when they want or when they need to take control back from the vehicle.

Build Confidence in Protection

To instill confidence, systems must make people feel safe. They must look and operate in a way that is protective of the driver, passengers and others.

We can’t afford for drivers to lack confidence in these systems or think they work only sometimes.

The quicker we make it through the white-knuckle phase of autonomous mobility, the better that is for adoption.

Ultimately, it has to be clear that the technology is protecting people. As autonomous systems are better designed to give drivers confidence and security, we will begin to see true adoption of these future vehicles.

In the end, the success of self-driving cars won’t be about technology, it will be about psychology.

Chris Rockwell is CEO and founder of Lextant, a user-experience consultancy dedicated to informing and inspiring design through an understanding of people, their experiences and aspirations. 

You May Also Like