Self-driving cars can teach the world how to drive

In most cases, driver monitoring systems use an infrared camera or cameras on the vehicle’s dashboard, steering wheel, or rearview mirror. Instead of sending images to the cloud like a cell phone, they process the images using software inside the car to track head, eye or hand movements. These systems have been trained on millions of images of real and artificial people behind the wheel, their makers say, and can tell whether a driver is looking at the road ahead or at the cellphone on their lap.

Depending on the automaker, a system that detects that the driver is not paying enough attention when using the automated driving feature may issue a warning, change the color of interior lights on the dashboard or steering wheel, sound an alarm, buzz the steering wheel, or some combination of all these tactics. If the driver does not react, the system will bring the car to a slow stop.

Gabi Zijderveld, marketing director of Smart Eye, a Swedish company that sells eye-tracking systems to automotive suppliers, calls driver monitoring in cars today “pretty simplistic.” The company says its driver control products are in more than 1 million cars on the road and will be in Volvo’s upcoming electric luxury Polestar 3. But in the future, a car equipped with a driver control system may be capable of more subtle feats, says Siederveld.

Data from cars can be used to train more sophisticated algorithms to recognize when a person is, say, safely scrolling through a car’s infotainment system during an easy drive on a sunny four-lane highway, or riskily fiddling with a playlist while navigating snowy city streets. Alerts may sound, flash or buzz in the latter case, but not in the former.

When a driver tries to multitask, researchers who study the psychology and mechanics of driving tend to judge their distraction based on how often and long enough they turn to the road to regain a sense of where their car and other vehicles are, cyclists and pedestrians are in space. Driver monitoring systems will eventually be able to combine information from the vehicle’s many sensors to, for example, determine if a driver is not paying enough attention when their vehicle is about to buckle up and tighten their seat belt.

Drivers who have already fallen on the wrong side of existing driver monitoring systems know that their warnings and voices can be annoying, and sometimes they cry. Automotive engineers choosing when and how systems buzz or beep must find a tricky balance.

Experts say the key to creating great driver monitoring systems is to create software that doesn’t just tell the driver when they’re doing something wrong, but keeps them focused. “It’s about defensive driving and collision avoidance versus post-collision crash avoidance,” says Greg Fitch, head of safety research for Android Auto, Google’s car app. This could mean the sound of low but increasing tones, rather than a shrill beep, when it sees you looking away when you may be watching for pedestrians. The system may not completely disable automatic lane keeping when you use the steering wheel to hug a curb, but instead shares the control.

Source by [author_name]

Exit mobile version