Light Detection and Ranging (LiDAR) is one of the most important technologies underpinning the development of driverless cars. LiDAR systems feed raw sensor data to the car’s brain, creating a 3D picture of the surrounding environment and enabling autonomous vehicles to navigate it.
But LiDAR has also been controversial because it’s expensive, sometimes fallible, and almost nobody is producing it at scale. There are competing sensor technologies, such as radar and cameras, and each has advocates trying to push them to market.
Ouster, a San Francisco-based startup developing LiDAR units, is trying to change perceptions of the technology. It says its LiDAR units are cheaper, smaller, and—perhaps most importantly—ready for commercial-scale production right now.
“Our LiDAR sensing technology is available today—that’s the most important thing,” says co-founder and CEO Angus Pacala. “Ouster is trying to turn the industry on its head. The entire [autonomous vehicle] market is constrained by issues with LiDAR sensor supply.”
Ouster is emerging from stealth this week armed with $27 million in funding from Cox Communications, which led the Series A round, with participation from Fontinalis Partners, Amity Ventures, Constellation Technology Ventures, Tao Capital Partners, and Carthona Capital. Ouster plans to use the new capital on further research and development, as well as beefing up manufacturing. Ouster anticipates producing tens of thousands of sensors in 2018 based on current product designs, Pacala says.
Pacala has been interested in the sensor technologies powering autonomous vehicles for a long time, he says. He and Ouster co-founder Mark Frictl, a classmate from Stanford University, previously worked together developing sensing tech at Quanergy and have talked a lot about the production and manufacture of sensors from a practical standpoint. Their goal when they started the company was to “volumize” high-performance LiDAR units and get the price down enough to make them widely attractive to automakers and suppliers.
“We knew all the big players needed to ramp up [production of autonomous vehicles] quickly and the sensors needed to come from somewhere,” Pacala explains. “We knew there was a massive need.”
Ouster says its initial 64-channel LiDAR system, called the OS1, is five times cheaper than units produced by the competition. Pacala says the company can achieve this reduction through innovations in semiconductor engineering, enabling Ouster to create custom microchips that cost much less.
The LiDAR space has been hot this Fall. GM announced its acquisition of LiDAR startup Strobe in October, and in September, two LiDAR companies—Innoviz Technologies and LedderTech—raised $65 million and $101 million in venture capital, respectively. Another Bay Area LiDAR startup, Luminar, raised a $36 million seed round earlier this year.
Chris Thomas, a partner at Fontinalis, says his firm spent a year doing a deep dive on more than 50 companies developing sensing technologies, and found Ouster to be ahead of its peers. Calling LiDAR a “bedrock sensor technology,” Thomas says Ouster has the unique combination of a “premier team and a premier approach,” along with a laser focus on manufacturing at scale.
“Legacy companies couldn’t meet the demand,” Thomas says, describing the market’s limitations. “The knock on LiDAR has always been that the price is too high to be volumized. Ouster is going to solve that problem, and its ability to come into the market and immediately produce at volume is 100 percent going to change the LiDAR debate. There’s a big difference between a science project and an incredibly robust and highly volumized technology.”
In 2018, Pacala plans to more than double the 40-person Ouster team and manufacture 1,000 LiDAR units per month by the end of January. Eventually, he adds, the company expects to open an office in Detroit. The company is currently in talks with automakers and suppliers about potential partnerships, but Pacala declined to share specifics. “You can’t sell high-performance sensors without partnerships,” he notes.
