Torc Robotics Inc. lately introduced that it’s collaborating with Flextronics Worldwide Ltd. and NVIDIA Corp. to develop a scalable bodily synthetic intelligence compute system for autonomous vehicles. The corporate final week demonstrated its know-how at NVIDIA GTC in San Jose, Calif.
“It was actually thrilling to make our GTC debut this week, together with the announcement of our NVIDIA and Flex collaboration,” mentioned CJ King, chief know-how officer of Torc, on the occasion. “As we get nearer to commercialization, it’s important to share extra of our best-in-class know-how and get our world-class engineers out right here to satisfy and share the story of Torc with others.”
Based in 2005, Torc is an unbiased subsidiary of Daimler Truck AG. The Blacksburg, Va.-based firm affords self-driving car software program and is specializing in commercializing autonomous vehicles for long-haul functions within the U.S.
Torc focuses on bodily AI
Torc Robotics mentioned its bodily AI allows self-driving vehicles to understand, perceive, and carry out complicated actions in the true (bodily) world. The firm asserted that bodily AI is on the core of its software program stack to allow vehicles to autonomously navigate their environment utilizing sensors equivalent to lidar, radar, and cameras.
This permits the truck to make knowledgeable choices about lane adjustments, braking, and impediment avoidance to make sure secure and environment friendly autonomous driving operations. Torc claimed that it’s the first deployment of bodily AI for autonomous long-haul trucking at manufacturing scale.
Along with engineering workplaces in Austin, Texas, and Montreal, Quebec, Torc Robotics has a fleet operations facility within the Dallas-Fort Value space in Texas, in addition to a presence in Ann Arbor, Mich.
Register now so you do not miss out!
NVIDIA DRIVE AGX supplies compute structure
Torc mentioned at NVIDIA GTC that it has collaborated intently with NVIDIA on a multi-chip adaptable structure that makes use of DRIVE AGX utilizing the DRIVE Orin system-on-a-chip (SoC) and DriveOS working system.
“NVIDIA DRIVE AGX has been industry-proven in full manufacturing for automotive real-time functions on the edge,” acknowledged Rishi Dhall, vice chairman of automotive at NVIDIA. “It delivers the excessive compute efficiency, low latency, and multi-sensor connectivity wanted for Torc’s refined autonomous trucking software program, delivering strong notion, prediction, and planning for secure and dependable operation.”
“Torc is on a transparent path to scalable manufacturing for its business launch in 2027 and dealing towards a seamless improve to NVIDIA DRIVE AGX with DRIVE Thor,” he added.
The platform makes use of NVIDIA DRIVE AGX with Flex’s Jupiter compute platform to help productization and scaled market entry in 2027. | Credit score: Torc Robotics
Flex design contributes to ‘software-defined car’
Flextronics is offering its Jupiter compute design platform and superior manufacturing capabilities. This provides Torc a scalable high-performance manufacturing {hardware} and software program primarily based on an autonomous-ready Freightliner Cascadia. That car is provided with superior applied sciences and redundant methods to help future deployment of autonomous driving capabilities.
“Our collaboration with Torc, Daimler Truck, and NVIDIA illustrates how Flex companions throughout the complete ecosystem to allow mobility firms to launch next-generation know-how with higher resilience and pace,” mentioned Mike Thoeny, president for automotive at Flex. “We respect the belief Torc and Daimler Truck have positioned in Flex by leveraging our Jupiter compute platform and superior manufacturing capabilities to ship autonomous long-haul trucking at scale.”
Torc Robotics mentioned NVIDIA DRIVE, paired with Flextronics’ automotive compute design capabilities, adheres to its stringent dimension, efficiency, value, and reliability necessities. The system additionally meets the overall value of possession (TCO) targets of Torc’s fleet prospects, that are pursuing continuous, long-haul driverless trucking integration.
“This work supplies a real software-defined car (SDV) performance,” the corporate mentioned. “It’s adaptable to ever-changing operational design domains, together with new lanes, new routes, new hubs, new {hardware} and sensor configurations, new operational guidelines, new street situations, and extra.”
Torc mentioned its driverless product-acceptance check efficiently validated the platform on a closed course at freeway speeds final 12 months.
