By Hyunjoo Jin

San Francisco (Reuters) –Toyota Motor Corp unit Woven planet is trying to follow the Tesla (NASDAQ) Inc lead in developing self-driving technology using low-cost cameras.

Woven Planet said to Reuters it can use cameras to capture data and efficiently train its self driving system. It called this a “breakthrough”, which it hopes will reduce costs and increase the scale of the technology.

It said that gathering diverse data from a large fleet of vehicles is crucial to develop a reliable self-driving vehicle system. However, this is expensive and difficult to scale up to allow for autonomous cars to be tested with costly sensors.

Woven Planet uses data from sensors such as radars or lidars to train and long-term deployment.

Tesla is betting on camera data to track over 1,000,000 vehicles to help develop automated driving technology. Alphabet (NASDAQ) added lidars and expensive sensors to only a few vehicles.

We need lots of data. In an interview with Reuters, Michael Benisch, Woven Planet’s vice president of Engineering, stated that it is not enough to have a limited amount of data from a very small number of expensive autonomous vehicles.

Benisch, an ex-engineering director at, stated that “Rather we are trying to prove that we can unlock Toyota and large automakers’ advantage, which would be access to a large corpus data with lower fidelity.” Lyft Toyota’s autonomous division (NASDAQ:), which was acquired by Toyota last year.

Woven Planet’s cameras are 90% more affordable than previous sensors and can easily be fitted in passenger vehicle fleets.

According to the company, using most of its data from low-cost cameras improved the performance of its system to an extent comparable to when it was only trained on data from high-cost sensors.

However, he said that Toyota will still be using multiple sensors, such as radars and lidars, to develop robotaxis.

Toyota is also partnering with Aurora in testing an autonomous ride-hailing fleet https://www.reuters.com/business/autos-transportation/toyota-aurora-test-drive-autonomous-ride-hailing-fleet-texas-2022-03-23 based on the Toyota Sienna minivans, equipped with lidars, radars, and cameras.

He said that it was possible for camera technology to catch up with more sophisticated sensors in the future, but this could take many years.

“The real question might be about how and when it takes to achieve safety and reliability. This is something I doubt we have a clue about.

Tesla’s CEO Elon Musk said it can achieve full autonomy with cameras this year https://www.reuters.com/article/tesla-robots-idCNL1N2U702E after missing his previous targets several times.