SAN FRANCISCO -- Was Autopilot on when a Tesla Model S smashed into the back of a parked Culver City fire truck on the 405 Freeway on Monday in broad daylight?
That's what the driver told police. Tesla Inc. -- which would have such information because it monitors car and driver behavior over wireless networks -- has not yet said yes or no.
The crash highlights a big problem facing the auto industry as evolution toward completely driverless cars continues: Many new cars are being equipped with robot systems that take over functions such as cruise control and steering but still require drivers to pay full attention.
Many drivers -- perhaps most drivers -- don't always focus solely on the road. Distracted by smartphones and other devices, drivers are crashing, killing themselves and others, at increasing rates.
Fully driverless cars, with no need for steering wheels, are likely to prove safer than human drivers, in large part because they are paying strict and constant attention to driving. Such vehicles are beginning to appear on public highways in places such as Phoenix and Las Vegas.
But those systems are expensive and experimental. Semiautonomous systems such as Autopilot are being installed in cars from Tesla, Audi, Volvo, Cadillac, Mercedes-Benz and others. They require humans and robots to share the driving duties.
Researchers with deep experience in human-machine interaction say it's folly to think that won't cause problems. Even if the human-robot team-up leads to safer roads on average, plenty of drivers will abuse the relationship, intentionally or not, and events like Monday's crash will make the news.
"There's something we used to call split responsibility," said Hod Lipson, director of Columbia University's Creative Machines Lab. "If you give the same responsibility to two people, they each will feel safe to drop the ball. Nobody has to be 100 percent, and that's a dangerous thing."
That's also true for humans sharing tasks with robots, he said.
Engineering researchers in the psychology department at the University of Utah are studying whether semiautonomous driving technology will make things better or worse.