News Briefs
- Elon Musk is hiring a whole lot of people in Texas.
- Autos CEOs change their tune on prospects for EV profitability.
- U.K. battery startup BritishVolt seeks a SPAC merger.
The Big Lidar Debate
Hello readers, it’s Gaby in Detroit. If you’ve been paying attention to self-driving car development in the last few years, you’ve no doubt heard about lidar, which stands for light detection and ranging.
Originally used by the military to map terrain, lidar will be crucial for self-driving cars because of its ability to see things cameras and radar sensors can’t. Lidar lasers can, for example, send invisible light waves through snow or fog to create a picture of a car’s surroundings.
In recent years, established companies and dozens of startups have focused on advancing lidar technology, and investors have been happy to give them money. U.S. venture investors poured a record $650 million into lidar startups in 2020, according to PitchBook, and the SPAC boom has only heightened expectations. Six lidar startups have gone public in the U.S. via reverse merger in the past year, and there are more than 70 startups globally, according to Guidehouse Insights.
Most AV geeks — with the notable exception of Elon Musk — agree lidar is critical to full self driving. But the consensus ends there. The jungle of rivals out there insist that various approaches are the best, most accurate and cost-effective way to enable self-driving cars.
In one camp, there are adherents to what's known as frequency-modulated light wave lidar, or FMCW for short. Proponents including Aurora Innovation say the tech is superior for long-range visibility, especially when travelling at high speeds. They also say it uses less power and is cheaper to manufacture than most alternative lidar methods.
Aurora recently acquired a company called OURS Technology that uses FMCW. Intel's Mobileye, which dominates the market for camera-based sensors on the road today, said in January it would build its own FMCW lidar in-house. Recently listed Aeva, founded by former Apple engineers, is another acolyte.
Austin Russell, the 26-year old physics prodigy and founder of Luminar Technologies, is among those on the other side of the debate. Luminar uses a different kind of lidar called time of flight. It sends out discrete pulses of light to judge how far away something is by how long it takes that light to bounce off objects and return, and by how intense it is.
Austin Russell, founder and CEO of Luminar Technologies, demonstrates imaging created by a lidar, which bounces lasers off objects to guide vehicles.
Photographer: David Paul Morris/Bloomberg
Last month, I visited the Orlando headquarters of Luminar. It's home to a manufacturing pilot line, clean rooms with technicians working on chips under microscopes and safety-testing equipment that subjects the company's hardware to extreme temperatures and vibrations.
Russell draws a lot of confidence from controlling the manufacturing process, which he views as key to proving one's lidar will work. Luminar looked at FMCW in its early days, he said, but decided the trade-offs weren't worth it. He's not shy about predicting his FMCW rivals will fail and doesn't fully buy their investor presentations. "I’m pretty skeptical that the physics actually works out behind peoples’ slides," he said.
I've heard the same level of conviction and scorn for the competition from almost every lidar executive I've interviewed. Despite millions of dollars spent and years of development, as with so much else in the world of autonomous cars, it's still early innings in lidar land. Some will prevail. There are simply too many for some not to fail, which brings us to one point on which many concur: consolidation is coming.
No comments:
Post a Comment