Nvidia has introduced Halos for Robotics, an end-to-end safety architecture aimed at autonomous and humanoid robots that operate in physical environments with people, especially in factories, warehouses and logistics centers. The company says the system is designed to help ensure a robot does not take actions that could damage its surroundings or injure nearby workers.
The architecture combines computing power, AI models, distributed sensor data and testing processes in a single framework. Nvidia said the goal is to give robotics teams a unified safety structure for expanding autonomous systems into industrial settings where human-robot collaboration is already increasingly common.
Deephu Talla, Nvidia’s vice president for robotics and edge AI, said in a press release that