ABB Robotics and NVIDIA announced a partnership Tuesday at the Computex trade show in Taipei to scale industrial-grade physical AI [1, 2].

This collaboration aims to accelerate the deployment of AI infrastructure for manufacturers. By integrating advanced simulation and power-system software, the partnership allows companies to validate production lines and power distribution before physical construction begins [2, 3].

The integration includes NVIDIA’s Omniverse libraries and DSX Blueprint into ABB's software ecosystem [1, 3]. These tools enable the creation of digital twins that can model complex industrial environments. An ABB spokesperson said integrating the DSX Blueprint will allow data-center operators to model and validate power distribution before breaking ground [3].

Craig McDonnell, Business Line Managing Director Industries at ABB Robotics, said the partnership allows the company to bring AI-driven simulation to customers at an unprecedented scale [2].

The initiative focuses on bringing generalized robot intelligence to the factory floor. This shift is intended to move AI from purely digital environments into physical applications, a concept known as physical AI [1, 2]. Through the use of SimReady 3D digital assets, manufacturers can more accurately simulate how robots will interact with their surroundings [3].

Stephen Engle, a Bloomberg Technology reporter, said the collaboration marks a significant step toward smarter factories and more generalized robot intelligence [1].

“Our partnership with NVIDIA allows us to bring AI‑driven simulation to our customers at unprecedented scale”

The integration of NVIDIA's simulation libraries into ABB's industrial hardware represents a shift toward 'physical AI,' where machine learning is applied to real-world kinematics and power logistics. By reducing the gap between digital simulation and physical deployment, manufacturers can lower the risk and cost of building new facilities and automating production lines.