Physical AI is Making its Way into Factory Floors

  • Post category:PI NEWS
  • Reading time:2 mins read

Industrial automation is on the verge of a paradigm shift: machines, robots, and mobile systems are learning to “understand” their surroundings. Physical AI—that is, AI that actively perceives the physical space and acts contextually—requires a unified understanding of positions, movements, and identities within that space.


As robot density and heterogeneity increase, a critical challenge becomes apparent—secure and scalable spatial coordination is needed. Image-based perception and integrated sensors alone are not sufficient to coordinate the various robots, people, and systems—often from multiple vendors—within the same dynamic environment.

Consequently, humanoid robots, autonomous mobile robots (AMRs), and AI-controlled vehicles in factories, warehouses, and logistics centers are increasingly demonstrating advanced perception, planning, and cooperation capabilities. After all, for coordinated interaction between stationary equipment and mobile robots in the flexible production of tomorrow, machines require a unified understanding of their position in space. Here, embodied AI has rapidly evolved from experimental robotics to real-world industrial applications. Driven by foundational models, simulations, and AI in robots with a physical body (embodied AI), these systems promise a new level of flexibility and productivity in industrial automation.

This is where omlox plays a central role. omlox defines an open and interoperable location-tracking standard that provides a common spatial reference across technologies and manufacturers. The omlox Hub harmonizes location data from multiple sources, enabling software-defined safety zones, traffic rules, and AI-driven coordination. In addition to cameras, omlox provides precise and fast localization of objects in space based on Ultra-Wideband (UWB) technology. In particular, the combination of radio-based and image-based location-tracking enables seamless tracking of objects in space.

Crucially, the omlox standard is operated by a non-profit industry association and jointly developed by many organizations in the fields of automation, logistics, robotics, and IT. This multi-stakeholder governance model strengthens independence, avoids vendor lock-in, and supports digital sovereignty —key requirements for industrial users seeking long-term reliability and freedom of choice.

In the age of physical AI, omlox®functions like the playing field and shared rules of a sports team: individual robots and AI agents may be highly capable, but true performance, safety, and trust only emerge when everyone operates based on the same spatial understanding. By providing this neutral, open spatial layer, omlox enables the secure and sustainable scaling of physical AI and transforms increasingly crowded manufacturing facilities into coordinated, intelligent environments.

Dr. Matthias Jöst
omlox Committee Chair at PI