ABB Robots Showcase Physical AI at 2026 China WAIC and Release White Paper

en.Wedoany.com Reported - At the 2026 World Artificial Intelligence Conference (WAIC) in China, ABB Robotics showcased its latest achievements in the field of Physical AI and jointly released a white paper with NVIDIA, focusing on the transformative role of industrial-grade Physical AI in precision manufacturing.

Marc Segura, President of ABB Robotics, stated that Physical AI is reshaping the capabilities and application scenarios of robots. Through the "Autonomous Versatile Robot (AVR™)" technology, which integrates machine vision, perception, and decision-making, robots are endowed with the ability to understand their environment and engage in continuous learning. James Han, Senior Vice President of ABB Group and President of ABB Robotics China, pointed out that the Chinese market has a strong demand for high precision and flexible deployment, and the company is collaborating with ecosystem partners to promote the implementation of Physical AI technology.

The white paper jointly released by ABB Robotics and NVIDIA elaborates on the potential of industrial-grade Physical AI to enhance the speed and flexibility of precision manufacturing, and proposes a replicable standardized engineering implementation process. Co-authored by AsiaInfo Technologies, Deloitte, and SKAI Intelligence, the white paper emphasizes the need to front-load risk identification through digital twins, synthetic data, and AI validation before physical debugging.

This white paper release represents the latest progress following the strategic cooperation announced by both parties in March 2026, aiming to integrate ABB's RobotStudio with NVIDIA Omniverse™ library technology to address the technical bottleneck of "simulation to reality."

At the conference site, ABB Robotics demonstrated the Physical AI toolchain and AVR™ technology through two demonstration workstations.

The Sim2Real digital twin workstation replicates product models and robot behaviors in a high-fidelity virtual environment, using synthetic data to cover long-tail cases, thereby closing the loop between virtual and real workflows.

The AI-guided snap-fit assembly workstation utilizes synthetic data generated by digital twins to pre-train visual AI models, reducing reliance on physical prototypes and real-world testing, and enhancing generalization capabilities.

Additionally, ABB Robotics signed a strategic cooperation agreement with SKAI Intelligence to collaborate in areas such as simulation scenario development, synthetic data generation, and visual AI training, jointly promoting the validation and adoption of Physical AI technology in industrial manufacturing.