Skip to content

TechnologyCalifornia1 min read

NVIDIA unveils Halos for Robotics safety architecture for physical AI

NVIDIA introduced Halos for Robotics, a full-stack safety system designed to help physical AI developers prepare for third-party certification.

Share

Topics

Two white humanoid robots high-five a man in a red jacket in a modern showroom or lobby setting, with a wooden wall behind them.

SANTA CLARA, Calif., June 27 (San Francisco News Online) — NVIDIA on June 22 introduced Halos for Robotics, which the company calls an industry-first full-stack safety system for physical AI. The architecture is designed to connect AI compute, system software, sensor data, safety applications and inspection in one standardized framework.

The company said the system includes NVIDIA IGX Thor and Holoscan Sensor Bridge for compute and sensor connectivity, NVIDIA Halos OS for software safety functions, and the NVIDIA Halos AI Systems Inspection Lab for inspection and certification preparation.

NVIDIA said Halos for Robotics draws on more than 18,600 engineering years of autonomous vehicle safety development. In its release, the company described the inspection lab as the first ANAB-accredited program for functional and AI safety for physical AI and said it is intended to help partners prepare for third-party certification from organizations including TÜV Rheinland, UL Solutions, TÜV SÜD, exida, SGS and CertX.

The release did not provide pricing or a general availability date for the full system.

Share

Topics

Andrew Brexton

Andrew Brexton is a senior technology analyst for Times Media Service, based in the Washington bureau. Brexton covers technology, artificial intelligence and emerging innovations, as well as energy, explaining complex technical developments for general readers. Brexton holds a master's degree in electrical and computer engineering and grew up in Galloway, Ohio.

Write to Andrew