Embodied AI Assists with Complex Physical Tasks

University of Chicago researchers develop AI-powered muscle stimulation system to guide users through unfamiliar movements.

Apr. 11, 2026 at 1:13pm

Got story updates? Submit your updates here. ›

A highly structured abstract painting in soft, flat colors featuring sweeping geometric arcs, concentric circles, and precise botanical or physical spirals, conceptually representing the interplay of technology and the human body that powers the embodied AI system.The embodied AI system combines artificial intelligence and electrical muscle stimulation to create a unique and intuitive form of physical assistance, guiding users through complex movements and tasks.Chicago Today

Researchers at the University of Chicago have created a groundbreaking 'embodied AI' system that combines artificial intelligence and electrical muscle stimulation to provide personalized physical guidance and training. This innovative technology can teach users complex movements and skills by directly transmitting 'know-how' to their muscles, revolutionizing applications from physical therapy to industrial training.

Why it matters

The embodied AI system represents a major leap forward from traditional electrical muscle stimulation (EMS) technologies, which have been limited to pre-programmed tasks. This new system can dynamically adapt to a user's context and seamlessly guide them through unfamiliar physical actions, opening up a wide range of possibilities for enhancing accessibility, learning, and everyday task completion.

The details

The embodied AI system, developed by Yun Ho, Romain Nith, and Pedro Lopes, is designed to transmit 'know-how' directly to the user's muscles, imparting an embodied sense of how to perform a task rather than just providing factual information. This procedural knowledge is essential for mastering complex physical actions but difficult to articulate through traditional instruction alone.

  • The embodied AI system was recognized with the Best Paper Award at the ACM CHI 2026 conference.

The players

Yun Ho

A researcher at the University of Chicago who co-developed the embodied AI system.

Romain Nith

A researcher at the University of Chicago who co-developed the embodied AI system.

Pedro Lopes

A researcher at the University of Chicago who co-developed the embodied AI system.

University of Chicago

The institution where the embodied AI system was developed.

Got photos? Submit your photos here. ›

What they’re saying

“Imagine a world where an invisible guide assists you with unfamiliar tasks, teaching you complex movements through your own muscles.”

— Yun Ho, Researcher, University of Chicago

“This technology is not meant to replace audiovisual guidance but to enhance it, offering a unique and personalized learning experience.”

— Pedro Lopes, Researcher, University of Chicago

What’s next

The researchers are working to further refine the embodied AI system, addressing challenges such as electrode calibration, the tingling sensation of EMS, and the difficulty of hardwiring finesse and muscle memory. As AI and EMS hardware continue to advance, the team is optimistic that on-body AI guides will become more commonplace in the near future.

The takeaway

The embodied AI system developed at the University of Chicago represents a significant breakthrough in physical assistance technology, combining artificial intelligence and electrical muscle stimulation to provide users with an intuitive, personalized, and adaptive learning experience. This innovative approach has the potential to revolutionize fields ranging from physical therapy to industrial training, empowering individuals to master complex tasks through direct bodily guidance.