NVIDIA has launched Halos for Robotics, a full-stack safety system designed for physical AI and autonomous machines. The system provides a common safety architecture for robots that operate in dynamic industrial environments alongside human workers.

The Halos system includes the NVIDIA IGX Thor computing platform and the Holoscan Sensor Bridge for real-time sensor connectivity. It also features the Halos OS software stack and an accredited AI Systems Inspection Lab. The lab is designed to help manufacturers prepare their robotic systems for third-party safety certifications.

Agility Robotics is the first company to integrate Halos for Robotics into its Digit humanoid robot. The robot is used in logistics and warehouse operations for customers including Amazon and GXO. Agility is using the system to meet international safety standards such as IEC 61508 and ISO 13849 for its human detection software.

The ecosystem includes partners such as Advantech, NXP Semiconductor, and STMicroelectronics, who provide compatible hardware and sensors. Assessment agencies including TÜV Rheinland and UL Solutions recognise the Halos Inspection Lab as part of their certification processes.

“Physical AI is transforming how factories, warehouses and logistics operations work, and robotics teams need a unified safety architecture to scale autonomous systems,” said Deepu Talla, vice president of robotics and edge AI at NVIDIA.