Building an autonomous vehicle system requires many parts, but one is more important and controversial than the other. This important component is the lidar sensor.
This is a device that perceives the surrounding 3D environment by emitting a laser beam to the surrounding environment and receiving the reflected beam. Self-driving cars being tested by Alphabet, Uber and Toyota rely heavily on lidar to help them locate on detailed maps and identify pedestrians and other vehicles. The best sensors can see details of a few centimeters from 100 meters away.
In the race to commercialize self-driving cars, most companies see lidar as essential (Tesla is an exception because it relies only on cameras and radar). Radar sensors don’t see much detail in low and bright light conditions. Last year, a Tesla car crashed into a tractor trailer, killing its driver, largely because the Autopilot software failed to distinguish the trailer body from the bright sky. Ryan Eustice, Toyota’s vice president of autonomous driving, told me recently that this is an “open question” — whether a less advanced self-driving safety system can function properly without it.
But self-driving technology is advancing so fast that the nascent industry is suffering from radar lag. Making and selling lidar sensors used to be a relatively niche business, and the technology wasn’t mature enough to be a standard part of millions of cars.
If you take a look at today’s self-driving prototypes, there’s one obvious problem: lidar sensors are bulky. That’s why vehicles tested by Waymo and Alphabet’s self-driving units have a giant black dome on top, while Toyota and Uber have a lidar the size of a coffee can.
Lidar sensors are also very expensive, costing thousands or even tens of thousands of dollars each. Most of the vehicles tested were equipped with multiple lidars. Demand has also become an issue, despite the relatively small number of test vehicles on the road.
Post time: Apr-03-2022