Article Image

Evolution of Lidar: What's the Next Step?

by Sam Chase
If you were driving around California or Michigan, and you passed what you suspected is a self-driving car, the first thing that might tip you off would be a spinning, semi-conical device atop the car. Known as Lidar (light detection and ranging), this sensor technology serves as the primary way that autonomous vehicles "see" their surroundings.

Every second, Lidar sensors send out millions of laser pulses that bounce off nearby objects, allowing a self-driving car's software to build a constantly refreshing three-dimensional map of its environment. Camera sensors then help an autonomous vehicle interpret the objects that the Lidar is gauging, and the vehicle navigates itself accordingly.

But while many consider Lidar to be the key to ultimately making self-driving cars a reality, it is also one of the biggest barriers. The reasons are twofold and interconnected. First, Lidar technology isn't good enough yet to sufficiently guide fully autonomous vehicles. Second, Lidar technology isn't cheap enough yet.

The challenge presented by the former compounds the difficulty of the latter.

A car traveling at highway speeds approaches upcoming objects at a faster rate than a car traveling at 25mph. To safely travel at speeds of 65mph or faster, a car would need to be equipped with high-powered Lidar that extends its lasers to great distances, giving the autonomous driving system a sufficient amount of time to process the information and react.

Velodyne's forthcoming Velarray LIDAR
Velodyne's forthcoming Velarray LIDAR

"Low-speed applications may be more affordable more quickly than higher-speed ones," Oxbotica CEO Graeme Smith told MIT Technology Review. "If you want a laser that's operating over 250 meters, you need a finely calibrated laser. If you're working in a lower-speed environment and can get by with 15 meters' range, then you can afford [to use] a much lower-cost sensor."

Of course, no viable autonomous vehicle will only navigate at school zone-level speeds. And cost remains a limiting factor even in the low-end tech. Lidar manufacturer Velodyne sells its least powerful sensor for $8,000, a figure that would be prohibitively expensive for mass production of autonomous vehicles.

Progress, however, is being made both with regards to Lidar capabilities and price points.

Velodyne touts its forthcoming Velarray sensor as ensuring "safe operation in L4 and L5 autonomous vehicles" at "a target price in the hundreds of dollars when produced in mass volumes." Quanergy, one of Velodyne's competitors, announced that it would begin production on a $250 Lidar system later this year.

Lidar is not the be-all and end-all for autonomous vehicles. Industry leader Tesla doesn't even use the sensors. But nearly every other self-driving car OEM does, so the technology's capability and affordability will undoubtedly shape how the industry progresses in coming years.

Webinars

ARCHIVED

Blockchain for Connected Vehicles: Driving Toward A More Automated Future

Blockchain technology, the same system that backs cryptocurrency transactions such as Bitcoin, has the potential to revolutionize connected vehicles and autonomous driving. Russell Vegh, principal member of the technical staff at A&T Internet of Things Solutions, will provide an in-depth overview of blockchain technology touching on key concepts such as mining, smart contracts, and decentralized applications. He will focus on how blockchain could play a key role in autonomous vehicles, V2I and V2X. He will also discuss key challenges such as device identity, information privacy and adoption.

Like us on Facebook