Solomon Integrates NVIDIA Architecture to Advance Autonomous Physical AI Humanoid Robots

The Taipei International Computer Show (COMPUTEX 2026) opened today. Solomon, a supplier of AI 3D vision and robotics technology, announced the integration of the NVIDIA NemoClaw architecture to coordinate multiple AI agents on humanoid robots. Solomon stated in a press release that this allows for the integration of inference, perception, sensor fusion, movement, and manipulation into a single workflow, further advancing the development of Physical AI and autonomous robots. Solomon explained that through NVIDIA's open-source foundation models and Solomon's active perception technology, robots can understand task requirements, actively adjust perspectives, optimize grasping and transport positioning, and make dynamic adjustments in real-time during execution in response to environmental changes. Solomon pointed out that the aforementioned AI workflow and perception capabilities are all executed through the NVIDIA Jetson edge computing platform, allowing robots to complete real-time autonomous operations directly on the local end, enhancing the reliability and autonomy of humanoid robots in complex real-world environments.