UWindsor study finds drivers distracted while driving semi-autonomous cars

A study by a kinesiology professor at the University of Windsor has found what many people have already suspected about the driving habits of people when they get behind the wheel of a semi-autonomous car: They don't pay much attention to the road.

"We suspected this would be the case, but it was much worse than we anticipated," said associate professor Francesco Biondi.

Biondi hooked up special headgear to 30 volunteers as they drove a 2022 Tesla Model 3 on Highway 401 to Chatham and back in both the manual and the L2 autonomous modes.

The headgear tracked eye movements, pupil dilation and blink rates. They also wore heart monitors and a device that sensed their hand-eye reaction time. A trio of cameras recorded the drivers' head movements, as well as what was happening on the road in front of and behind the vehicle. 

"They started paying less attention to the road, more attention to the massive touch screen, and they started looking elsewhere, not at the forward roadway, but side glance and looking at the other sides for much longer than in the manual mode," he said.

He said some even fell asleep while in the semi-autonomous mode.

In the semi-autonomous mode, the vehicle maintains its positioning in its lane. It maintains a constant speed unless there's a slower vehicle ahead, in which case it slows down to follow at a safe distance. Drivers are expected to keep their eyes on the road and their hands on the wheel.

• Fatal Tesla crash in Texas had a driver behind the wheel, probe finds

• Speeding Tesla rolls off road in southeast Edmonton, killing 3, injuring 3 others

Biondi says the drivers were accompanied by a research assistant who told them not to text, and woke them if they fell asleep, but he wonders what would happen if she hadn't been there.

"Imagine what they would do if they were left free to do whatever they want with their phones or with the touch screen," he said. 

Bondi says he believes the semi-autonomous vehicles are dangerous and have not lived up to the hype about how much safer they are.

He points to data from the National Highway Traffic Safety Administration in the U.S. that shows there were 736 crashes by Tesla cars with 15 deaths since 2019.

"We've been promised for many, many years that these cars would be much safer than human drivers. But so far, the evidence is not showing that," said Biondi.

Tesla publishes safety information and quarterly crash statistics on its website. Crash rates while drivers are in autopilot — measured as miles driven per accident — are at a fraction of the U.S. average, Tesla's data shows.

"While no car can prevent all accidents, we work every day to try to make them much less likely to occur," the company says on its website, noting that it can use software updates to make safety upgrades and enhancements to its connected cars long after they are purchased.

Meanwhile, the executive director of the mobility partnerships and innovation with Invest WindsorEssex said technology exists that can keep the drivers focused on the road.

"There could be vibrations within the seats, or within the steering wheels themselves. There could be full modes where the vehicle can just pull over to the side of the road until that individual's been rested well enough to get back onto the road itself," said Matthew Johnson.

John Komar, the executive director of the Automotive Centre of Excellence at Ontario Tech University, says engineers creating new autonomous technology now have to factor in what people are going to do with it.

 "This isn't just an engineering problem. This is a human problem," said Komar.

Biondi is calling on more study in Canada, as has been done in the U.S., and also better training for drivers.

"I'm hoping that this going to sort of give a wake-up call to regulators but also to automakers and to the driver population to say these systems are not as safe as we're told," he said.