The initial goal of self-driving cars is to make them as good as human drivers — but Ford has another trick up its sleeve to nudge things to the next level. By using Lidar, Ford recently showed that its autonomous vehicles can drive in complete darkness without human intervention. In the process, they remind us that the self-driving car is about a lot more than as-good-as-humans: It’s about harnessing tech to make cars far safer and smarter than human drivers.
Self-driving cars aren’t a new phenomenon, but one key advantage that took a long time to sink in is the realization that by letting a robot drive a car, you aren’t trying to create a human to drive a car for you. It goes far beyond that: The field of robotics has a lot more toys in its tool chest than those of a mere human.
This jumped into crisp relief especially well in Ford’s recent test at its Arizona Proving Ground. The company rigged up one of its self-driving cars with Lidar technology, which combines lasers with 3D mapping and distance-sensing technology. Lidar is, in essence, radar done with laser.
“Thanks to Lidar, the test cars aren’t reliant on the sun shining, nor cameras detecting painted white lines on the asphalt,” says Jim McBride, Ford technical leader for autonomous vehicles. “Lidar allows autonomous cars to drive just as well in the dark as they do in the light of day.”
“Inside the car, I could feel it moving, but when I looked out the window, I only saw darkness,” recounts Wayne Williams, a Ford research scientist and engineer who rode in the car with the computer at the proverbial steering wheel, a feat that shows extraordinary faith in your own technology.
Driving a car around in the pitch dark would be perilous in real life, of course: Headlights are as much about being seen as they are about seeing the road in front of you. Having a chance to see the technology in action is quite magical, and to me, the key innovation is that the car is taking a step into the realm of augmented reality. Well, to the degree that a robot has a concept of a non-augmented reality, that is.
By using available technologies like computer vision, Lidar, radar, GPS, traffic reports and other data sources, seeing autonomous cars combining all its resources for increased safety is very exciting. Whereas a human driver is limited to his eyes and ears when driving, an autonomous car can — and should — use whatever sensory data it has available.
The video in particular highlights that it is still early days, but it also serves as a reminder that we are rolling into the future, autonomous inch by autonomous inch.