In the DMV area especially, our cities shut down at the mention of inclement weather; but if autonomous cars are to be successful, they will need to learn to navigate challenging weather. That’s why in September of last year, Waymo, an autonomous car developer, announced they would begin testing their self-driving cars in Michigan.
This is a necessary step in the autonomous car process. Snow, rain, ice, wind, and fog all affect human drivers: our visibility, the way the car handles the road, and weather can even force us to pull over. Self-driving cars will need to handle similar situations in order to maintain their promise of safety.
These vehicles are using state-of-the-art technology to navigate, and understanding that technology will help to explain why weather will be difficult for them to navigate. Most companies testing autonomous technology use a combination of radar, lidar laser scanners, and cameras. You may be thinking that cars should be fine as they have radar technology, technology we use to map out storms and follow inclement weather patterns, but unfortunately while radar does function well in bad weather, the data it gives the vehicle is not detailed enough to allow a car to navigate by itself.
Lidar laser scanners (light detection and ranging), offer a more detailed view of the surroundings to the autonomous car. The sensor sends a multitude of infrared lasers around the vehicle every second, mapping and building a 3D picture of the area by measuring the time it takes each beam to come back after bouncing off the nearest object. Think echolocation for cars. You’re essentially riding around in a Batmobile. Unfortunately, bad weather can make this process inefficient as the lasers can mistake rain, snow, and dust in the air as objects and stop the vehicle from moving forward.
Some companies, like Ford, are attempting a different tactic to allow autonomous cars to navigate poor weather conditions, creation of 3D internal maps. Essentially, Ford has allowed its vehicles to create maps of the area on clear weather days and store the information for use on poor weather days. The cars would record road markings, nearby landmarks, signs, geography and topography, then use above-ground landmarks as a way to pinpoint its location when weather disrupts its vision, following the created map to successfully resume its route.
It’s a neat idea, but one that needs a lot of work, especially if a car is not driving on a local road when bad weather hits as it may not have that particular road mapped out in its RAM.
Autonomous car developers are continuing to work on these issues related to weather. Some have windshield wipers on their sensors and other’s are trying to formulate new ways to ensure cars can navigate snow-covered roads. The answer could be in smart infrastructure, roads and signs that can communicate electronically making vision less important. But roads and signs only go so far, the human factor must always be considered as well.
A graduate of George Washington University of National Law Center, Thomas J. Curcio is an author, speaker, and personal injury attorney. He has been named one of DC Best Lawyers Personal Injury Litigation and co-author of “Evidence for the Trial Lawyer." His experience includes the successful representation of injured passengers of the 2009 WMATA (DC Metro) Fort Trotten train crash.