Bilge Mutlu
University of Wisconsin–Madison
Date/Time: Tuesday, 6 January 2026, 11:00 am
Place: Aysel Sabuncu Brain Research Center Seminar Room (SC-106)
Abstract:
Human-robot interaction is often framed in terms of interfaces, autonomy, or task performance. In this talk, I argue that HRI is more productively understood as the study of distributed cognitive systems where cognition emerges across humans, machines, and environments over time. I present three cognitive principles that shape interaction with artificial agents: joint action as a coupled dynamical process, embodiment as priors for inference, and cognition as situated in the physical environment. Drawing on examples from collaborative, embodied, and assistive robotics, I show how these principles jointly explain when interaction succeeds and where it breaks down. I conclude by arguing that assistive human-robot interaction serves as a critical stress test for cognitive theories of interaction.
Bio:
Bilge Mutlu is the Sheldon B. & Marianne S. Lubar Professor at the University of Wisconsin–Madison, with appointments in Computer Sciences, Psychology, and Industrial & Systems Engineering. He directs the People & Robots Laboratory and is the founding director of the Institute for Design & Embodied AI (IDEA), a campus-wide initiative focused on human-centered robotics and embodied AI. His research develops principles and methods for designing robotic and AI systems that integrate effectively into everyday environments. Dr. Mutlu teaches courses in human-computer interaction, interaction design, and user experience (UX) design. He directs the NSF-funded INTEGRATE Graduate Research Traineeship on robotics and the future of work and has held leadership roles in the ACM/IEEE Human-Robot Interaction community and the Human-Computer Interaction Consortium. He received his PhD in Human-Computer Interaction from Carnegie Mellon University.