Torc Robotics Inc. recently announced that it is collaborating with Flextronics International Ltd. and NVIDIA Corp. to develop a scalable physical artificial intelligence compute system for autonomous trucks. The company last week demonstrated its technology at NVIDIA GTC in San Jose, Calif.
“It was really exciting to make our GTC debut this week, along with the announcement of our NVIDIA and Flex collaboration,” said CJ King, chief technology officer of Torc, at the event. “As we get closer to commercialization, it’s essential to share more of our best-in-class technology and get our world-class engineers out here to meet and share the story of Torc with others.”
Founded in 2005, Torc is an independent subsidiary of Daimler Truck AG. The Blacksburg, Va.-based company offers self-driving vehicle software and is focusing on commercializing autonomous trucks for long-haul applications in the U.S.
Torc focuses on physical AI
Torc Robotics said its physical AI enables self-driving trucks to perceive, understand, and perform complex actions in the real (physical) world. The company asserted that physical AI is at the core of its software stack to enable trucks to autonomously navigate their surroundings using sensors such as lidar, radar, and cameras.
This allows the truck to make informed decisions about lane changes, braking, and obstacle avoidance to ensure safe and efficient autonomous driving operations. Torc claimed that it is the first deployment of physical AI for autonomous long-haul trucking at production scale.
In addition to engineering offices in Austin, Texas, and Montreal, Quebec, Torc Robotics has a fleet operations facility in the Dallas-Fort Worth area in Texas, as well as a presence in Ann Arbor, Mich.
NVIDIA DRIVE AGX provides compute architecture
Torc said at NVIDIA GTC that it has collaborated closely with NVIDIA on a multi-chip adaptable architecture that uses DRIVE AGX using the DRIVE Orin system-on-a-chip (SoC) and DriveOS operating system.
“NVIDIA DRIVE AGX has been industry-proven in full production for automotive real-time applications at the edge,” stated Rishi Dhall, vice president of automotive at NVIDIA. “It delivers the high compute performance, low latency, and multi-sensor connectivity needed for Torc’s sophisticated autonomous trucking software, delivering robust perception, prediction, and planning for safe and reliable operation.”
“Torc is on a clear path to scalable production for its commercial launch in 2027 and working toward a seamless upgrade to NVIDIA DRIVE AGX with DRIVE Thor,” he added.
The platform uses NVIDIA DRIVE AGX with Flex’s Jupiter compute platform to support productization and scaled market entry in 2027. | Credit: Torc Robotics
Flex design contributes to ‘software-defined vehicle’
Flextronics is providing its Jupiter compute design platform and advanced manufacturing capabilities. This gives Torc a scalable high-performance production hardware and software based on an autonomous-ready Freightliner Cascadia. That vehicle is equipped with advanced technologies and redundant systems to support future deployment of autonomous driving capabilities.
“Our collaboration with Torc, Daimler Truck, and NVIDIA illustrates how Flex partners across the full ecosystem to enable mobility companies to launch next-generation technology with greater resilience and speed,” said Mike Thoeny, president for automotive at Flex. “We appreciate the trust Torc and Daimler Truck have placed in Flex through leveraging our Jupiter compute platform and advanced manufacturing capabilities to deliver autonomous long-haul trucking at scale.”
Torc Robotics said NVIDIA DRIVE, paired with Flextronics’ automotive compute design capabilities, adheres to its stringent size, performance, cost, and reliability requirements. The system also meets the total cost of ownership (TCO) targets of Torc’s fleet customers, which are pursuing non-stop, long-haul driverless trucking integration.
“This work provides a true software-defined vehicle (SDV) functionality,” the company said. “It is adaptable to ever-changing operational design domains, including new lanes, new routes, new hubs, new hardware and sensor configurations, new operational rules, new road conditions, and more.”
Torc said its driverless product-acceptance test successfully validated the platform on a closed course at highway speeds last year.
Tell Us What You Think!