Toyota doubles down on Nvidia tech for self-driving cars

Toyota is deepening its relationship with Nvidia, as the automaker, and its research arms in Japan and the U.S., ramps up its autonomous-vehicle development program.

Nvidia CEO Jensen Huang announced Monday during his keynote at the 2019 GPU Technology Conference that Toyota Research Institute-Advanced Development — the automaker’s Japan-based research arm — is using the chipmaker’s full end-to-end development and production to develop, train and validate its autonomous vehicle technology. The partnership builds on an ongoing collaboration with Toyota and is based on development between engineering teams from Nvidia, TRI-AD in Japan and Toyota Research Institute in the United States.

This new agreement means Toyota will use Nvidia’s platform for training deep neural networks, testing, validation and eventual deployment for its cars. Toyota is also using Nvidia’s newly released AV simulator Drive Constellation, which is now available to customers. Toyota is the first customer to use Constellation, a cloud-based platform that enables autonomous-vehicle developers to test their technology in the virtual world. 

In short, Toyota is going to use Nvidia’s technology for the entire workflow or process to develop autonomous vehicles.

“Close collaboration is really our business model,” Danny Shapiro, Nvidia’s senior director of automotive said Monday. “It’s our way of developing jointly and building the Nvidia drivers platform.”

Nvidia and Toyota have been collaborating for several years now. Toyota announced in 2017 it would use Nvidia’s Drive PX supercomputer, a platform with a processor called Xavier, to power the autonomous driving systems inside its future cars.

Toyota, and its research arms TRI and Japan-based TRI-AD, are taking a dual approach to autonomy.

Toyota intends to eventually deploy fully autonomous cars that would serve elderly and disabled people under its so-called Chauffeur system. The automaker is also working on “Guardian,” a system for production vehicles that will operate in the background and step in when needed. The driver is always driving, but Guardian is watching, sensing and anticipating problems. 

It’s unclear if Toyota will use Nvidia’s platform for the development of the Guardian system or fully autonomous vehicles.

“Our vision is to enable self-driving vehicles with the ultimate goal of reducing fatalities to zero, enabling smoother transportation, and providing mobility for all,” TRI-AD CEO James Kuffner said in a statement. “Our technology collaboration with Nvidia is important to realizing this vision. We believe large-scale simulation tools for software validation and testing are critical for automated driving systems.”

Toyota is just one of several automotive partnerships Nvidia has locked in since 2015, when it introduced its original architecture for autonomous vehicles, a supercomputer platform called Drive PX. The original platform was designed to process all of the data coming from the vehicle’s cameras and sensors and then use an AI algorithm-based operating system and a cloud-based, high-definition 3D map to help the car understand its environment, know its location and anticipate potential hazards while driving.