NVIDIA Emphasizes Safety Across All Layers for Physical AI Deployment
NVIDIA highlights the importance of safety in the deployment of autonomous vehicles and robots, emphasizing that safety must be integrated across hardware, software, AI, operating environment, and deployment lifecycle.
NVIDIA announced the critical need for safety in the large-scale deployment of autonomous vehicles (AVs) and robots. According to the announcement, as these AI-driven machines take on more complex tasks in the physical world, safety must extend across every layer of the system, including hardware, software, AI, operating environment, and deployment lifecycle.
ABI Research projects an installed base of 49 million level 3-5 autonomous vehicles by 2035, while Omdia estimates that approximately 60 million industrial robots will be deployed between 2026 and 2035. As these machines enter roads, factories, warehouses, and other environments shared with people, safety must scale with them.
NVIDIA's Halos is the first full-stack safety system for physical AI, designed to help developers engineer safety across every layer of design, validation, and deployment. The system includes hardware such as NVIDIA DRIVE AGX Thor and NVIDIA Hyperion, as well as an operating system and middleware built on ASIL-D certified DriveOS, providing a unified software foundation for system isolation, monitoring, and deterministic communication.
Source: nvidia-newsroom
