After 11 autopilot engineers left Tesla in May amid department restructuring, Tesla found itself in need of new software talent.
On October 1, Tesla announced its acquisition of Deepscale, a Silicon Valley, California-based startup that will help actualize Tesla CEO Elon Musk’s desire for its vehicles to become self-driving robotaxis. On LinkedIn, Deepscale’s CEO Forrest Iandola announced his new role as a senior staff machine learning scientist at Tesla.
CNBC reported that Tesla bought Deepscale “outright,” but no financial details have surfaced. Deepscale is Tesla’s sixth acquisition, including Maxwell Technologies in May 2019. According to Pitchbook, Deepscale raised $21 million of venture capital, funded by companies like Autotech VC and Bessemer.
DeepScale has the technology to improve computer vision using low-wattage processors, which are standard in most cars. Using sensors, cameras and mapping, Deepscale’s technology can help cars understand and decipher their surroundings, a key component to deploying driverless cars.
Deepscale’s focus on computer vision to give cars “sight” is different from light detection and ranging (LIDAR), a popular technology that many automakers have invested in for autonomous prototypes.
In April, Musk announced that all Tesla Model 3, X and S vehicles contain a custom chip, which gives Tesla vehicles complete self-driving ability.
“All cars being produced all have the hardware necessary – computer and otherwise – for full self-driving,” Musk said. “All you need to do is improve the software.”
Musk aspires towards the Level 4 designation by the Society of Automobile Engineers (SAE), which means the vehicle can fully operate without human drivers. Right now, with the latest chip, the cars operate at Level 2, not fully autonomous, but more advanced than most cars on the road.
Tesla’s ambition is evident with this acquisition. Musk announced in April his hopes to have completed Tesla’s full self-driving system by the end of 2019 and for the U.S. government to be processing approval by the end of 2020.
