Mechanical Engineering / Human-centered robotics

MOSAIC Robotics Lab

MultisensOry Systems for Adaptive and Interactive Cognition. We design, model, and study interactive robotic systems that help people sense, act, and learn in richer ways.

Research Directions

MOSAIC develops multisensory interactive systems that adapt to how people perceive, learn, and act. The lab combines mechanical and wearable device design, psychophysics, computational modeling, and closed-loop control to build technologies that respond to human perception.

Multisensory virtual environment with visual, sound, and haptic contact cues

Sensory Congruence & Conflict

In virtual and augmented environments, what people see and what they feel often disagree. We study how users adapt to these mismatches and build models that can predict perception from sensory interaction traces.

Human-robot collaboration with multisensory feedback

Multisensory Human-Robot Interaction

Physical collaboration depends on a shared sense of the task. We design robots and interfaces that infer how people weigh sensory information, then adapt motion and feedback so the human-robot pair stays perceptually aligned.

Surgical or health training system with adaptive visual, audio, and haptic feedback

Interactive Tools for Training & Health

Training and rehabilitation systems need feedback that transfers beyond the simulator. We pair real-time sensing with adaptive visual, audio, and haptic cues to support surgical training, rehabilitation, and other high-stakes motor learning settings.

News

  1. MOSAIC Robotics Lab presents a conference paper on trimodal alignment of multisensory contact data and workshop poster on learning user-specific sensory strategies from interaction data at IEEE IROS.
  2. MOSAIC Robotics Lab opens in the Department of Mechanical & Aerospace Engineering at the University of Houston.
  3. New work on active touch and referred haptics appears at IEEE World Haptics Conference.
  4. Research on vision and referred haptic feedback in VR environments is published in Springer proceedings.