TECH – NVIDIA and Hexagon showcased a fleet of humanoid robots capable of operating industrial machinery, inspecting components, and scanning assets, powered by advanced AI and simulation training. Developed primarily for factory and logistics environments, the AEON robot integrates NVIDIA’s comprehensive robotics suite, enabling autonomous task performance in real-world settings.
Unveiled at Hexagon LIVE Global, AEON was built through collaboration between Hexagon’s robotics division and NVIDIA. According to Arnaud Robert, President of Hexagon Robotics, “Our goal with AEON was to design an intelligent, autonomous humanoid that addresses the real-world challenges industrial leaders have shared with us”. The robot is engineered to perform precision tasks—such as manipulating parts and inspecting assets—while reducing human error under variable conditions.
Training for AEON occurred almost entirely within simulation environments using NVIDIA’s Isaac platform. Employing Isaac Sim and Isaac Lab, the humanoid learned navigation, locomotion, and manipulation capabilities virtually before physical deployment. This method trimmed development time drastically: what would typically take months to train in labs was achieved in just two to three weeks.
Read More: Huawei Pura 80 Series Delivers Unmatched Battery and Charging Breakthroughs
At the core of AEON’s capabilities is NVIDIA’s robotics stack. It currently operates on Jetson Orin for real-time processing and is slated for an upgrade to the Jetson Thor platform, which offers more power and safety features. Deepu Talla, NVIDIA’s Vice‑President of Robotics and Edge AI, proclaimed, “The age of general-purpose robotics has arrived, due to technological advances in simulation and physical AI”.
AEON also leverages Isaac GR00T foundation models and Mimic tools. These systems enable the robot to learn from human demonstrations and generate synthetic motion—sharp increases in dexterity and adaptability follow. For example, path planning efficiency improved by up to 80%, while AI-driven perception achieved zero-shot capabilities. AEON can autonomously scan parts and conveyor lines, uploading 3D models to Hexagon’s Reality Cloud Studio via Omniverse, which facilitates collaboration and digital twin creation. Lucas Heinzle, VP of R&D at Hexagon Robotics, explained this feature as enabling “integration of reality data capture with NVIDIA Omniverse, streamlining workflows…moving us closer to making digital twins a mainstream tool for collaboration and innovation”.
Pilot deployments are already in motion at factories and warehouses, signaling a broader shift toward smart, automated operations. AEON marks a major step in transitioning humanoid robots from the lab to the factory floor, representing a powerful combination of high-performance hardware, simulated training, and AI-driven autonomy.
