Tech
NVIDIA Builds Full-Stack Safety System for Autonomous Machines
NVIDIA has outlined Halos, a safety architecture for autonomous vehicles and robots operating alongside people.
What happened
NVIDIA says physical AI safety must cover hardware, software, AI behaviour, operating environments and the deployment lifecycle. Its Halos architecture combines safety-engineered computing, operating-system and middleware controls, AI models, simulation, synthetic data and validation tools for autonomous vehicles and robotics. The company cites ABI Research’s projection of 49 million level 3-5 autonomous vehicles by 2035 and Omdia’s estimate of roughly 60 million industrial robots deployed between 2026 and 2035. NVIDIA says partners including Geely, Nissan, Einride, Agility and KION Group are developing systems using parts of the platform. Certification bodies have assessed or are inspecting several components.
Why it matters
As autonomous machines enter roads, factories and warehouses, developers need evidence that systems can respond safely to changing conditions, failures and AI-specific risks. NVIDIA’s approach treats safety as an ongoing engineering and validation requirement, rather than a final pre-deployment check.
Source: NVIDIA Blog