Physical AI Powering the New Age of Industrial Operations 2025

3How it scales: Technology platforms and partnerships A new physical AI technology stack for robotics is emerging, and strategic partnerships will be essential to enable integration, scalability and ecosystem-wide success. Before intelligent robotics can scale, the right foundation must be in place. Technology alone is not enough – what matters is how systems are built, integrated and evolved through collaboration. 3.1 The new physical AI technology stack As robotics technology matures, a new physical AI stack is emerging that fundamentally differs from the traditional robotic platform. This emerging technology stack is structured around five layers as highlighted in Figure 4. Emerging physical AI technology stack FIGURE 4 Application Interfaces and tools for end-user interaction and system integration, including APIs, connectors and intuitive HMIs for monitoring and control Simulation/training Virtual environment, tools and data platforms for robot development and testing, using synthetic data generation, high-fidelity simulators and digital twins to enable accurate “sim-to-real” transfer Operating system Backbone of the stack for hardware abstraction, process coordination and component communication; manages task scheduling and integrates frameworks such as a robot operating system (ROS) to ensure standardization and interoperability Edge hardware On-device processors for real-time AI inference and sensor fusion, enabling autonomous decisions with minimal latency and without cloud reliance Robotic hardware Mechanical base, including actuators, controllers and sensors/vision systems, that enables the robot to act, sense and perceive Source: World Economic Forum, BCG. Physical AI: Powering the New Age of Industrial Operations 16