BMW i Ventures invests in autonomous truck technology company Kodiak Robotics – TechCrunch

0


BMW’s Silicon Valley-based venture capital arm is investing in Kodiak Robotics, a company that develops autonomous trucking technology.

Kodiak will use the funds to establish a safety case for its stand-alone technology stack so that it can be brought to market faster. It will also work on hiring new talent and expanding its truck fleet, with a stated goal of at least doubling the number of vehicles it operates each year. The startup currently has 10 trucks rotating between its trade route in Texas and its test pilot in Mountain View, California.

Terms of the deal were not disclosed. BMW i Ventures typically invests in companies that can provide solutions for BMW’s current and future business, but Kodiak CEO and co-founder Don Burnette told TechCrunch that BMW’s investment was purely financial and not strategic, so which means that there is currently no technical partnership between the two.

This new investment comes just one week after the tire manufacturer Bridgestone announces minority stake in Kodiak. The financial data behind the deal was also not disclosed. To date, Kodiak has publicly announced $ 40 million in funding for its Series A, and Burnette says the startup has made several additional investments since then.

Burnette also shared the company’s plans to conduct large-scale driverless operations for less than 10% of what Waymo has raised publicly to date – $ 5.7 billion – and less than 25% of the collection. of TuSimple’s total existing fund – roughly $ 1.94 billion, including the money raised from its IPO. That leaves us with a figure of about $ 500 million.

“This is the total amount we think we need to go driverless, and it’s because we think we’ve been a much more efficient business so far, and we will continue to be much more efficient in the process. ‘future,’ said Burnette.

The BMW i Ventures financing will eventually become part of Kodiak’s B series. With this latest investment, the company is not trying to further develop its autonomous driving capabilities or functionality, but rather to develop its safety record and prove that its system can handle the road without a driver on board, says Burnette.

“We’re moving towards a Level 4 range system, but we still have a driver in the seat who is actually monitoring our system at all times,” Burnette said. “Today we are technically a level 2 system, which is true of just about everyone.”

Vehicles with a human driver overseeing operations such as steering, braking and throttle assist, as well as items such as lane centering and adaptive cruise control fit under the Society of Automotive Engineers (SAE) definition of range level 2. Level 4 means that the vehicle can handle all aspects of driving under certain conditions without human intervention.

Kodiak says he’s made progress. In January, he announced that his pilot Kodiak had made “no-disengagement deliveries” (meaning the autonomous system did not need to be shut down for safety reasons) during a trade route between Dallas and Houston. The company has operated this route from its test and operating facility in Dallas for two years, and says it has now reached a level of maturity where the system can handle whatever the freeway throws at it.

“We do some really complex and advanced maneuvers, not only to deal with obvious things like mergers and cuts and heavy traffic, but also more nuanced challenges like identifying vehicles that are stopped on the side of the road.” said Burnette. “Our system can automatically identify that and then slow down as required by law, or move away from that object to give it more space. He may also consider changing lanes if a lane is available, a way to give the disabled vehicle even more room to the side, and that’s exactly how humans drive.

To get to the point where Kodiak can prove that its technology is in fact safer than a human driver, and therefore suitable for large-scale commercial operation, the startup must accumulate the total number of kilometers driven in simulated environments, tests structured on a private closed track, and in real driving.

Burnette says Kodiak is the only company that doesn’t designate one type of sensor as “primary,” and instead takes a holistic approach, meaning it’s not a company that prioritizes lidar or vision. . Tesla chief AI officer Andrej Karpathy recently revealed the company’s new supercomputer that takes a visual-only approach, but Burnette fundamentally disagrees with that method.

“We believe that each of these different modalities has strengths and weaknesses, and our goal is to build on those strengths and cover weaknesses with other modalities. So we created a sensor fusion algorithm that allows us to consider which sensors are advantageous when they give us the most usable information, ”said Burnette.

Kodiak also doesn’t use HD mapping, so their trucks see real-time on the road, allowing the Kodiak driver to be flexible when it comes to changing road conditions or environments. The system is trained using data collected by Kodiak trucks, as well as scenarios designed by its engineers, and this data is automatically labeled using Scale AI, which is one of the ways Kodiak is able to cut costs, says Burnette.

The Kodiak team came from the original self-driving team of Google, Uber, Lyft, and other notable tech companies. Burnette says BMW i Ventures’ investment in the business came after a thorough vetting process in which the company sent in its autonomous driving experts and tapped into the team’s expertise and technology.


Leave A Reply

Your email address will not be published.