TL;DR

Physical AI is revolutionising robotics, enabling machines to perform varied tasks and respond to changing circumstances. With over 4.7 million industrial robots now operational and installations growing at twice the rate of a decade ago, businesses across sectors are adopting intelligent automation.

Beyond the Production Line

The latest frontier in artificial intelligence is where physical devices meet AI capabilities. More than 4.7 million industrial robots were in operation in 2024, according to the International Federation of Robotics, with this figure rising by over 500,000 annually.

Physical AI takes robots to a new level. They can now combine autonomy with hardware that moves objects using sensors to perceive their surroundings. This marks the evolution from deterministic machines performing identical tasks to systems that complete varied, complex tasks and respond to changing circumstances.

The World Economic Forum notes that robots are moving from “high-volume, low-variation” environments to “high-variation, low-volume” settings, making deployment viable in small-scale and dynamic operations.

Real-World Impact

Siemens reports that AI-enabled robots performing pick-and-place operations reduce automation costs by 90%. Amazon uses over one million robots across its fulfilment centres, with AI capabilities enabling far greater flexibility. Their Vulcan robot uses feedback sensors to handle items from socks to electronics with appropriate force.

Healthcare has shown particularly strong growth, with nearly twice as many new robots installed in 2024 compared to the previous year. Applications range from rehabilitation exoskeletons to medicine delivery and surgical assistance systems.

The Data Challenge

Training physical AI remains expensive and time-consuming. Hundreds of hours of demonstrations can be required for a robot to learn a task. Unlike large language models that can ingest internet data, real-world environment data is harder to collect. Simulated environments help but require diverse original datasets to avoid producing unreliable results.

Looking Forward

For UK businesses, physical AI presents opportunities to improve efficiency, fill workforce gaps and enhance worker safety. However, successful adoption requires strategic thinking about workflow redesign rather than simple cost-cutting. Companies that involve their people in envisaging new ways of working will achieve the most successful AI adoption.