Agility Robotics is the primary to make use of NVIDIA Halos for Robotics to construct security into its humanoids working in factories, warehouses, and logistics operations for patrons together with Amazon, GXO, Schaeffler, and Toyota Motor Manufacturing Canada. | Supply: NVIDIA
NVIDIA at present launched NVIDIA Halos for Robotics. This can be a full-stack, complete security system for robotics and bodily AI that unifies AI compute and security.
NVIDIA mentioned autonomous robots will want AI basis fashions, accelerated compute, and distributed sensors to function in dynamic environments alongside people. Scaling these programs requires a full-stack security structure.
NVIDIA mentioned Halos permits firms to depend on a standardized, unified security structure that connects AI compute, system software program, sensor knowledge, security purposes, and inspection for robotic programs.
“Bodily AI is reworking how factories, warehouses, and logistics operations work, and robotics groups want a unified security structure to scale autonomous programs into these environments,” mentioned Deepu Talla, the vp of robotics and edge AI at NVIDIA. “With NVIDIA Halos for Robotics, builders and system builders can harness NVIDIA’s confirmed autonomous car security basis to develop safer robots sooner and convey them into industrial operations alongside staff with better confidence.”
NVIDIA Halos Core for NVIDIA IGX is on the market in early entry for registered builders in Linux and Linux plus QNX configurations. The open supply NVIDIA Halos Exterior-In Security Blueprint, a part of the Halos Purposes layer of Halos OS, is now out there in early entry on GitHub.
NVIDIA brings security to the complete software program stack
To create Halos, NVIDIA drew on over 18,600 engineering years of autonomous car security growth. Halos for Robotics gives builders with a typical security structure for constructing, validating, and deploying bodily AI programs.
The system spans the important thing layers wanted for robotic security. NVIDIA IGX Thor and NVIDIA Holoscan Sensor Bridge present industrial-grade AI compute, built-in security, and sensor connectivity for real-time robotics and security workloads.
NVIDIA Halos OS gives the software program stack for robotics security. This consists of Halos Core to assist safety-related working features and security purposes constructed with the NVIDIA Halos Exterior-In Security Blueprint. NVIDIA mentioned this program extends robotic notion utilizing exterior cameras and AI brokers to dynamically management robotic conduct in industrial settings.
The NVIDIA Halos AI Methods Inspection Lab is an ANSI Nationwide Accreditation Board (ANAB)-accredited program for bodily AI purposeful and AI security. It helps companions put together Halos integrations for third-party certification by main certification our bodies. These embody TÜV Rheinland, UL Options, TÜV SÜD, exida, SGS, and CertX.
NVIDIA builds Halos with scale in thoughts
The NVIDIA Halos for Robotics ecosystem brings collectively companions throughout software program, programs, sensors, and silicon, industrial purposes, and certification our bodies to assist security from growth by deployment:
- Software program: Acontis, Amazon FreeRTOS, and QNX assist the real-time working surroundings, security communications, and embedded software program layers wanted for purposeful security growth.
- Embedded programs: Advantech and NexCobot ship safety-designed IGX-based programs for robotics deployments.
- Sensors and silicon: Infineon, NXP, SICK, STMicroelectronics, and Texas Devices contribute sensor, security microcontroller, and semiconductor applied sciences.
- Industrial purposes: FORT Robotics, Inventec, KION Group, and Neurealm are growing purposeful security brokers utilizing the NVIDIA Halos Exterior-In Security Blueprint.
- Certification Our bodies: TÜV Rheinland is inspecting NVIDIA IGX Thor, Halos OS, and Holoscan Sensor
- Bridge for purposeful security certification readiness, constructing on TUV SuD’s inspection of Thor SoC and certification of Halos Core for ISO 26262
The NVIDIA Halos AI Methods Inspection Lab consists of greater than 40 firms throughout producers, certification our bodies, and security distributors working to maneuver protected bodily AI programs from design to real-world deployment.
TÜV Rheinland, TÜV SÜD, UL Options, exida, SGS, and CertX all acknowledge the NVIDIA Halos AI Methods Inspection Lab as a part of their certification course of.
Agility incorporates Halos in its humanoids
Humanoid robots are designed to function in dynamic environments alongside staff, gear, and different robots which are always in movement. That requires security engineered for each layer of the stack.
Agility is teaming with NVIDIA to combine NVIDIA IGX Thor and Halos Core into its proprietary protected human detection system for its humanoid robotic Digit, which is designed for industrial work in logistics, manufacturing, and warehouse operations.
For Digit, NVIDIA IGX Thor delivers industrial-grade AI compute with built-in security capabilities, whereas Halos Core helps the software program layer for safety-related working features.
Agility will even take part within the NVIDIA Halos AI Methods Inspection Lab. Collectively, Agility and NVIDIA will use the lab to make sure Digit’s safety-related software program, AI elements, and cybersecurity protections meet requirements reminiscent of IEC 61508, ISO 13849, and ISO/IEC TR 5469 earlier than ultimate third-party certification.
“For humanoids to ship worth at scale, security needs to be constructed into the robotic and validated throughout the complete system,” mentioned Peggy Johnson, CEO of Agility. “Partnering with NVIDIA to implement and optimize the Halos for Robotics system extends our management in accountable automation, which is a non-negotiable requirement for bringing humanoids safely into industrial workflows. This collaboration unlocks true human-robot teamwork, driving the long-term returns that can energy next-generation manufacturing and logistics operations.”
Submit your session concept for the 2026 RoboBusiness
