# Nvidia introduces Halos safety system for autonomous vehicles and robots

Digest AI · Hardware & Compute · published 2026-09-21T16:00:46Z

Canonical: https://digestai.news/story/nvidia-introduces-halos-safety-system-for-autonomous-vehicles-and-robo

## Summary

Physical AI is moving from research to large‑scale deployment, with ABI Research forecasting 49 million level‑3‑5 autonomous vehicles by 2035 and Omdia estimating 60 million industrial robots will be deployed between 2026 and 2035. As these machines share roads, factories and warehouses with people, safety must be built into every layer – hardware, software, AI behavior, operating environment and the deployment lifecycle.

Nvidia’s Halos safety foundation combines a decade of AV safety work with new hardware and software. For autonomous vehicles it bundles DRIVE AGX Thor compute, the Hyperion platform, Halos OS (AS‑ILD certified), Halos Core and Middleware, the Alpamayo vision‑language‑action model, and a Safety Evaluation Framework that links cloud simulation to in‑vehicle evidence. For robotics it offers IGX Thor modules, Halos Core for IGX, the Holoscan Sensor Bridge, Isaac Lab and Omniverse simulation tools, and an Outside‑In Safety Blueprint that extends awareness beyond onboard sensors. Independent assessments by TÜV SÜD, TÜV Rheinland and ANAB certify the stack for functional‑safety standards.

Automakers and mobility providers such as Geely, Nissan, Uber, Grab and Lyft, as well as robotics partners including Bosch, Infineon and KION Group, are already integrating Halos components to meet certification and trust requirements for scalable physical AI.

## Key points

- ABI Research projects 49 million level‑3‑5 autonomous vehicles by 2035; Omdia expects 60 million industrial robots 2026‑2035.
- Nvidia Halos offers a full‑stack safety system covering hardware, OS, AI models, simulation and certification for AVs and robotics.
- Companies such as Geely, Nissan, Uber, Bosch and Infineon use Halos components like DRIVE AGX Thor and IGX Thor.

## Why it matters

A unified safety stack lets manufacturers certify and scale physical AI systems, reducing risk as autonomous vehicles and robots proliferate.

## Sources

1. [Why Deploying Physical AI at Scale Demands Safety at Every Layer](https://blogs.nvidia.com/blog/physical-ai-halos-safety) (NVIDIA Blog, 2026-09-21, primary source)

Part of the developing story: [Nvidia Expands Autonomous Vehicle Safety Ecosystem](https://digestai.news/thread/nvidias-robotaxi-stack-powers-global-driverless-fleets) (2 stories)

## Cite

Digest AI, "Nvidia introduces Halos safety system for autonomous vehicles and robots", 21 September 2026, https://digestai.news/story/nvidia-introduces-halos-safety-system-for-autonomous-vehicles-and-robo

---

Written by Digest AI's editorial model from the linked sources; the sources are the record. Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse
JSON: https://digestai.news/story/nvidia-introduces-halos-safety-system-for-autonomous-vehicles-and-robo.json