Physical AI is artificial intelligence built into machines that sense their surroundings and act on them, rather than software that works only with text, images or data. Examples include factory and warehouse robots, humanoid robots, autonomous vehicles and drones. Such systems combine sensors such as cameras and lidar, onboard computing chips that run AI models locally, and an actuator or motor that turns decisions into movement.
The term became common in investment marketing as AI models improved at understanding the physical world and as chip makers promoted computing platforms for robotics. Thematic funds and analysts use it to group companies across robotics, sensors, edge chips and industrial automation. The commercial outlook is less certain than for software AI: hardware is costly, safety rules are strict, and many products remain at the pilot stage.
Ref 1180001546