New entrants – ranging from AI-first start-ups (e.g. Sereact, Covariant) to tech giants (e.g. Nvidia, Tesla, Apple, Google) – are joining traditional players in reshaping the ecosystem. A vital innovation lies in the simulation/training layer. High-fidelity world models, kinematic, physics-based photorealistic simulations (depending on the application) and synthetic data generation enable the development and deployment of robust AI skills. Some software- only start-ups have built upon existing hardware to upgrade their capabilities (e.g. Covariant, Intrinsic). At the same time, vertically integrated companies have emerged to provide a solution across the entire technology stack (e.g. Neura, Figure, Tesla, Boston Dynamics). In most cases, their innovations belong in the space of context-aware robotics using the humanoid form factor.Independent of whether the approach entails a modular vs. vertically integrated solution within the layered technology stack, scalable impact requires the seamless integration of the AI technology stack into the existing manufacturing environment. Today, manufacturers rely on a complex industrial software tool chain from which product and process data from diverse sources needs to be ingested and integrated with adjacent automation equipment to ensure system-level intelligence. Along with traditional integrators, new service providers have entered the market, including those offering robots-as-a-service (RaaS) models for deployment, operation and maintenance. This has broadened accessibility and lowered the barrier for companies that lack internal capabilities. 3.2 Strategic partnerships are essential In this fast-evolving landscape, strategic partnerships offer one effective pathway for manufacturers who want to harness the latest robotics innovations. No single company can realistically develop all of the advanced capabilities alone at the pace at which technology is progressing. The obstacles are especially high for smaller enterprises that lack the resources to build up the required capabilities independently. The most successful manufacturers identify the right partners with whom to collaborate. By forging strong partnerships with technology providers, research institutions and peers within and across the industry, manufacturers can stay at the forefront of change and employ collective expertise. For example, a car manufacturer might collaborate with an AI start-up to co-develop a robotic-equipped assembly line, while also working with a university robotics lab on new manipulation techniques. Such collaboration helps companies keep up with rapid advances – tapping into partners’ specialized knowledge in AI, sensors or software updates – rather than falling behind the state of the art. It also reduces integration hurdles: when robot makers, AI developers and factory engineers plan solutions together, they can anticipate and solve compatibility issues early, ensuring smoother deployment. Similarly, partnerships invite co-creation of solutions tailored to real operational needs. Manufacturers can guide integrators and equipment suppliers on what is needed on the factory floor, and in turn receive highly customized, robotic systems that no party could have built in isolation. Beyond technical innovation, collaboration can also unlock new markets and capabilities, risk and investment sharing, and early alignment with regulators and standards bodies. In summary, the rise of the physical AI technology stack is not just about new technology stacks, but also about building new relationships and ecosystems. It is a world in which vertically integrated robot companies work alongside component specialists, and end users collaborate closely with innovators. Embracing this ethos of building partnerships and ecosystems is how manufacturers will navigate the new age of intelligent robotics – staying agile, sharing investments, benefits and risks, and co-creating the next generation of industrial automation. Scalable impact requires the seamless integration of the AI technology stack into the existing manufacturing environment. Physical AI: Powering the New Age of Industrial Operations 17

Physical AI Powering the New Age of Industrial Operations 2025