Texas Instruments is accelerating the “secure deployment of humanoid robots into the actual world” with Nvidia. By combining TI’s real-time motor management, sensing, radar and energy applied sciences with Nvidia’s superior robotics compute, Ethernet-based sensing and simulation applied sciences, robotics builders can validate notion, actuation and security earlier and extra precisely.
TI connects Nvidia bodily AI compute to real-world purposes with deterministic management, sensing, energy and security at each joint and subsystem. This partnership will assist builders transfer sooner from digital improvement to production-ready, scalable and safety-compliant programs.
As a part of this collaboration, TI designed a sensor fusion resolution by integrating its mmWave radar expertise with Nvidia Jetson Thor utilizing Nvidia Holoscan Sensor Bridge to allow low-latency, 3D notion and security consciousness for humanoid robots. TI will showcase the answer at Nvidia GTC, March 16–19, 2026, in San Jose, California.
Giovanni Campanella, basic supervisor of commercial automation and robotics at TI, says: “The subsequent era of bodily AI requires extra than simply superior compute – it calls for seamless integration between sensing, management, energy and security programs.
“TI’s complete portfolio bridges the hole between Nvidia’s highly effective AI compute and real-world purposes, enabling builders to validate full humanoid programs earlier in improvement. This built-in method will assist speed up the evolution from prototypes to commercially viable humanoid robots working safely alongside people.”
Deepu Talla, vice chairman of robotics and edge AI at Nvidia, says: “The secure operation of humanoid robots in unpredictable environments requires an enormous leap in processing energy to synchronize advanced AI fashions with real-time sensor knowledge and motor controls.
“The mixing of Texas Devices’ sensing and energy administration applied sciences with the Nvidia Jetson Thor platform supplies builders with a useful safety-capable basis to speed up the deployment of next-generation bodily AI.”
Enabling safer humanoid robots with real-time sensor fusion expertise
TI’s mmWave radar sensor, IWR6243, related by way of Ethernet to Nvidia Jetson Thor allows scalable low-latency, 3D notion and security consciousness for bodily AI purposes.
By fusing digicam and radar knowledge, the answer improves object detection, localization and monitoring whereas lowering false positives for assured, real-time decision-making in humanoid robots.
This resolution allows human-like notion that works reliably in difficult situations – from low gentle and shiny glare to fog and mud indoors and outside – and addresses a essential security hole that has restricted real-world deployment of humanoid robots.
For instance, whereas cameras might not reliably detect glass doorways or reflective surfaces, radar supplies constant detection of those clear obstacles, enabling clean navigation in locations like workplace buildings, hospitals and retail environments.
TI at Nvidia GTC
TI will current its applied sciences at Nvidia GTC, from March 16-19, in sales space 169 on the San Jose McEnery Conference Heart. TI and D3 Embedded’s reside demonstration, “Actual-time sensor fusion for dependable robotic notion with Holoscan”, showcases how TI’s mmWave radar expertise integrates with Nvidia’s Jetson Thor and Holoscan ecosystem utilizing an end-to-end software program processing chain and visualization from D3 Embedded.
On Wednesday, March 18 from 3:00-3:40pm PT, TI’s Giovanni Campanella, will take part in a lightning discuss, “The Fringe of the Edge: Redefining GPU-Enabled AI Sensor Processing”. Campanella will talk about how tight integration of sensing, networking and GPUs is enabling real-time bodily AI on the edge of commercial programs.
