Follow
Subscribe

General Motors Invests in Oculii, a Startup Developing Advanced Radar Software for Autonomous Vehicles

Home > News > Content

【Summary】​General Motors has made another key investment in the autonomous driving space, as the automaker looks to add more advanced self-driving driving capabilities to its model lineup. GM’s venture capital arm invested “millions of dollars” in San Francisco startup Oculii, according to its co-founder Steven Hong.

Eric Walz    Sep 13, 2021 10:30 AM PT
General Motors Invests in Oculii, a Startup Developing Advanced Radar Software for Autonomous Vehicles

General Motors has made another key investment in the autonomous driving space, as the automaker looks to add more advanced self-driving driving capabilities to its model lineup.

GM's venture capital arm invested "millions of dollars" in San Francisco startup Oculii according to its co-founder Steven Hong. GM's investment in Oculii was first reported by Reuters.

The investment (from GM) is a "fantastic signal they're serious about the technology and bullish about radar in general," Hong told Reuters. He declined to disclose the financial details of GM's investment.

Hong is a Stanford University graduate who founded Oculii with his father, Lang Hong, an engineering professor at Wright State University in Ohio.

Oculii is working to improve the resolution of radar data with advanced AI software that cleans up images rendered by a vehicle's radar sensors. The technology can help improve the perception systems of autonomous vehicles.

The company developed what its calls "virtual aperture imaging software" which it says increases the resolution of any radar platform by up to 100x. The technology can lead to more cost effective radar solutions for the auto industry as the resolution of lower cost radar sensors can be improved using software.

Oculii's technology scales from partially automated vehicles and fully self-driving vehicles, Hong told Reuters in an interview.

The radar also supports Simultaneous Localization and Mapping (SLAM) using precise doppler and angle information to accurately map the environment and localize a vehicle relative to where it's been. This is a critical function in both robotics and autonomous vehicles. 

Radar is an important sensor for autonomous vehicles. When combined with cameras and lidar sensors it's used for perception and navigation. Radar can help identify other vehicles and road users, as well as measure an object's speed and direction. Radar also works better in low light conditions, giving it an advantage over cameras-based systems in detecting objects at night.

Despite the advantages of radar for self-driving vehicles, automaker Tesla opted to use a camera-based perception for its Autopilot driver assist system. Both Tesla Chief Executive Elon Musk and the automaker's artificial intelligence director, Andrej Karpathy, believe that radar has its shortcomings and feel that vision-based autonomous driving systems are more accurate for navigation, which is why Tesla has doubled down on use of cameras.

Hong however, said he expects Tesla to eventually embrace radar perception as an extra layer of safety as prices decline.

"Traditional radar is very low resolution and very noisy," Hong said. But high-resolution radars are a key backup to cameras and other sensors when they fail, thus providing "extra safety," he added.

GM is getting ready to roll out its "Supercruise'' autonomous driving feature to more models, which means that radar sensors will be installed on many more models in the near future. Last year, GM said it plans to offer Supercruise on 22 vehicles by 2023, so the automaker will need access to low cost automotive grade radar sensors. 


Prev                  Next
Writer's other posts
Comments:
    Related Content