Embodied Intelligence

Robots and other physical agents that sense, move, and learn through interaction with people and environments.

Embodied intelligence treats the body and the environment as part of the computation. This track includes locomotion, manipulation, tactile sensing, multi-agent physical interaction, and human–robot collaboration.

We welcome work on skill learning, safety-aware control, benchmarking of real robots, and methods that reduce the gap between simulation and hardware. Studies that report compute, energy, and failure cases in real deployments are strongly encouraged.

Authors should describe the embodiment (sensors, actuators, workspace), the task, and the risks of physical contact with people or infrastructure.

All papers in this track must follow the PIJ Ethics Code.