There’s a lot of talk about self-driving cars being able to outperform human drivers, but this is likely the first time that has been proven. The test car was driven in the middle of the night at Ford’s Arizona Proving Ground not only without headlights, but without onboard cameras to help it “see.” They were rendered useless by the lack of light.
Instead, the car relied on a combination of lidar and detailed 3D maps to guide it around the course. The maps included everything from topography to road markings, as well as landmarks like signs, buildings, and trees. The car used lidar laser pulses to orient itself, and compared its location to information from the maps in order to avoid driving off the road.
Ford engineers monitored the test from both inside and outside the car, wearing night-vision goggles to see. That meant they could also see the lidar system sending out laser pulses around the car, which took the form of a grid. The system sends out 2.8 million pulses a second, in fact. Riding in the back seat, Ford engineer Wayne Williams said he could feel the car moving, but couldn’t see anything to indicate movement out the window.
Ford has tested autonomous Fusion Hybrid prototypes since 2013, but plans to expand the scope of testing this year. It will add 20 cars to the fleet, bringing the total to 30. Cars will continue testing at locations in California, Arizona, and Michigan, and the larger fleet will allow for a greater frequency of tests.
The company is widely rumored to be in talks with Google about pooling resources on autonomous cars. Google is actively seeking an automaker to partner with on commercializing the technology, as it isn’t very interested in getting into the car business. However, neither Google nor Ford will confirm plans for an alliance.
- Lockdown couldn’t keep Waymo from testing self-driving cars … in a fake city
- Tesla Model S vs. Lucid Air
- The best commuter cars for 2020
- The best cargo vans for small businesses in 2020
- Every upcoming electric car