Nvidia Corp. announced the launch of Halos for Robotics on June 22, 2026, positioning the platform as an industry-first framework for building, testing, and managing safety systems within AI-powered robotics applications. The system aims to bridge the gap between complex robotic autonomy and the stringent safety requirements necessary for machines operating in dynamic human environments, such as factories and warehouses.

Agility Robotics Inc., a leader in humanoid and physical AI, is the first organization to integrate Nvidia Halos into its operations. The company plans to use the framework to support robots deployed for clients including Amazon.com Inc., GXO, Schaeffler, and Toyota Motor Manufacturing Canada. According to Peggy Johnson, Chief Executive of Agility Robotics, safety validation across entire systems is a non-negotiable requirement for scaling humanoid labor in industrial workflows.

The Halos for Robotics framework comprises several integrated layers designed to manage safety-critical functions. It utilizes the IGX Thor and Holoscan Sensor bridge to provide industrial-grade AI compute and connectivity. The software stack includes Halos OS for underlying safety support and Halos Core, which provides pluggable blueprints that allow robots to extend perception through external cameras and AI agents. These components enable capabilities such as human detection, avoidance, and automated emergency slowing or freezing to prevent injury.

To address regulatory challenges, Nvidia introduced the Halos AI Systems Inspection Lab, which holds accreditation from the American National Standards Institute National Accreditation Board. This facility assists partners in preparing for third-party certifications by entities including TÜV Rheinland, UL Solutions, TÜV SÜD, Exida, SGS, and CertX. These efforts aim to help companies navigate emerging safety standards, such as the European Union Machinery Regulation of 2027, which takes effect on Jan. 20, 2027, and introduces new conformity requirements for machines with self-evolving behaviors powered by AI foundation models. Nvidia intends for Halos to serve as a standardized safety validation platform for the broader robotics industry.

Nvidia Corp. announced the launch of Halos for Robotics on June 22, 2026, positioning the platform as an industry-first framework for building, testing, and managing safety systems within AI-powered robotics applications. The system aims to bridge the gap between complex robotic autonomy and the stringent safety requirements necessary for machines operating in dynamic human environments, such as factories and warehouses.

Agility Robotics Inc., a leader in humanoid and physical AI, is the first organization to integrate Nvidia Halos into its operations. The company plans to use the framework to support robots deployed for clients including Amazon.com Inc., GXO, Schaeffler, and Toyota Motor Manufacturing Canada. According to Peggy Johnson, Chief Executive of Agility Robotics, safety validation across entire systems is a non-negotiable requirement for scaling humanoid labor in industrial workflows.

The Halos for Robotics framework comprises several integrated layers designed to manage safety-critical functions. It utilizes the IGX Thor and Holoscan Sensor bridge to provide industrial-grade AI compute and connectivity. The software stack includes Halos OS for underlying safety support and Halos Core, which provides pluggable blueprints that allow robots to extend perception through external cameras and AI agents. These components enable capabilities such as human detection, avoidance, and automated emergency slowing or freezing to prevent injury.

To address regulatory challenges, Nvidia introduced the Halos AI Systems Inspection Lab, which holds accreditation from the American National Standards Institute National Accreditation Board. This facility assists partners in preparing for third-party certifications by entities including TÜV Rheinland, UL Solutions, TÜV SÜD, Exida, SGS, and CertX. These efforts aim to help companies navigate emerging safety standards, such as the European Union Machinery Regulation of 2027, which takes effect on Jan. 20, 2027, and introduces new conformity requirements for machines with self-evolving behaviors powered by AI foundation models. Nvidia intends for Halos to serve as a standardized safety validation platform for the broader robotics industry.