SKAI Intelligence Joins ABB, NVIDIA Physical AI White Paper as Main Contributor, Emphasizes Importance of "Talent"

The competitiveness of the physical AI industry is expanding beyond technology development to include talent acquisition and the ability to deploy solutions in real-world settings. SKAI Intelligence contributed to a global white paper on industrial physical AI, jointly published by ABB Robotics and NVIDIA, highlighting the importance of talent development and establishing training and validation systems.

SKAI Intelligence said on Oct. 21 that it had joined AWS, AsiaInfo, Deloitte and others as a Main Contributor to the second global white paper on industrial physical AI, jointly published by ABB Robotics and NVIDIA. Titled "Preparing Talent for Industrial Physical AI," the white paper was officially released on Oct. 20.

Focusing on manufacturing, the white paper defines industrial physical AI as AI systems that perceive and make decisions about real-world physical environments, then perform tasks using robots and automated equipment. It also sets out a systematic approach to the skills, education and training, and validation methods needed to deploy the technology in industrial settings, proposing a global framework that can be adapted to the industrial and educational environments of different regions.

The white paper highlights the gap between the pace at which physical AI technology is spreading and its level of adoption in real-world industrial settings. According to Deloitte's "State of AI in ER&I: The Untapped Edge 2026," cited in the white paper, a survey of 3,235 senior business leaders in 24 countries worldwide in August and September 2025 found that 72% of the 591 respondents in the energy, resources and industrials (ER&I) sector were using physical AI at least to a limited extent.

By contrast, only 19% said they were using it at a moderate or higher level. Meanwhile, 91% expected to be using it at least to a limited extent within the next two years. This indicates that while technology adoption itself is expanding rapidly, the capabilities needed to apply it to actual industrial operations and scale its use remain relatively underdeveloped.

A similar gap was evident in talent development and job design. In the ER&I sector, 53% of respondents said they were training more employees to improve their AI skills, while 50% said they were designing or implementing upskilling and reskilling strategies. However, 84% had yet to redesign jobs to incorporate AI capabilities.

The white paper therefore emphasizes the need to build capabilities that can be applied in real industrial settings, rather than focusing solely on AI training. Since physical AI combines AI models with sensors, robots, control systems, data infrastructure, production processes, and safety and quality management, it calls for talent with skills spanning development, system integration, and on-site operations.

In particular, it identifies "Translational Capability" as a core skill: the ability to retain expertise in one's own field while understanding and connecting the technical language, constraints, and validation standards of adjacent fields such as AI, robotics, and manufacturing processes.

In Part 3 of the white paper, "Building a Scalable Learning Ecosystem," SKAI Intelligence presented its expertise in establishing physical AI capabilities and learning systems that connect real industrial environments with simulation. Josh Liu, Chief Technology Officer (CTO), said, "Industrial physical AI programs must be built around demonstrable capabilities," emphasizing the need for a learning system that moves from understanding to deployment and improvement through real-world assignments and hands-on practice.

Digital twins, simulation, and synthetic data—areas in which SKAI Intelligence has developed expertise—were also covered in the white paper as key technologies. It identified simulation and digital twins as essential capabilities that can support design and operational validation, the generation of synthetic and rare-event data, operational decision-making, and talent development.

The learning process consists of five stages: "understanding → simulation → physical validation → workplace application → improvement." A key feature is its emphasis on a "Simulation First" approach: using simulation as the initial step, followed by physical validation in a real-world environment, then on-site application and improvement.

The evaluation of training outcomes also goes beyond checking whether participants have completed a course or passed an exam, extending to verifying whether they can perform tasks in real industrial settings. Practical capabilities are assessed through system demonstrations, the development of validation plans, and simulation and physical testing.

Lee Jae-cheol, CEO of SKAI Intelligence, said, "For industrial physical AI to spread to real-world settings, we need not only the technology but also people with the capabilities to understand, integrate, validate, and operate it." He added, "With our participation as a Main Contributor to this white paper, we plan to expand technical discussions with the global industrial community, drawing on our physical AI capabilities in areas such as simulation and digital twins, and contribute to building a physical AI ecosystem that can be used in actual manufacturing settings."

Meanwhile, SKAI Intelligence also contributed to the first white paper, "Industrial Physical AI in High-Precision Manufacturing." That paper presented a method for training and validating robotic vision AI before its deployment on production lines, using synthetic data that reflects key variables in high-precision manufacturing, including industrial tolerances, reflections, and occlusions.

By contributing as a Main Contributor to two consecutive physical AI white papers jointly published by ABB Robotics and NVIDIA, SKAI Intelligence has expanded its collaboration beyond industrial physical AI technology to include on-site deployment and talent development.

This content was produced with the assistance of AI translation services.