This is Upstate New York, so it’s too early to call game over with respect to snowfall. It’s only late February, after all, and we’ve been known to have snow in May. Still, the corner has been turned. But when it comes to handling snow, the corner hasn’t been turned quite yet for self-driving cars. As I learned when I saw an article on Bloomberg a couple of weeks back.
Volvo does some car testing north of the Arctic Circle, where the winter weather can make Syracuse look like Palm Beach. One of the cars Volvo’s testing up there is a self-driving SUV. What they found was that frozen snowflakes were getting to the sensors that were trying to read the road. As a result, the SUVs went snow blind. Volvo engineers solved the problem by placing the vehicle’s radar sensors behind the windshield. The windshield wipers keep things clear, so the sensors can do their thing.
Anyway, it seems like self-driving cars aren’t quite weather for winter weather.
“There’s been a lot of hype in the media and in the public mind’s eye” about the technology for self-driving cars “being nearly solved,” said Ryan Eustice, an associate professor of engineering at the University of Michigan who is working with Ford on snow testing. “But a car that’s able to do nationwide, all-weather driving, under all conditions, that’s still the Holy Grail.” (Source: Bloomberg News)
Ford thinks that they’ve solved one of the driving-in-snow problems: visibility of the lane lines. Using lidar (which does sensing with laser light), they’re scanning roads and creating high-def 3-D maps “that are much more accurate than images from global-positioning satellites, which can be 10 meters (33 feet) off.” What Ford is trying to do is replicate what human drivers do when the line lanes are no longer visible: we drive in the ruts left by the car ahead of us.
Lidar is one of several technologies used to let driverless cars “see”. Cameras and radar are also used. High-speed processors crunch the data to provide 360-degree detection of lanes, traffic, pedestrians, signs, stoplights and anything else in the vehicle’s path. That enables it to decide, in real time, where to go.”
Different types of sensors are used in combinations, capturing data on the same objects in different ways, rather than relying on just one source of information. AI also comes into play.
Winter conditions get in the way of all this. Just as they get in the way of us human drivers. Just like us, a driverless car can become disoriented.
Maybe the best advice for driverless cars will be the same advice that we have to take every winter. There are some times when you just need to stay off the road.