Our primary research areas are robotics and natural language processing, combined to study understanding language in the context of the physical world. Our application areas focus on deploying and using robots in human-centric environments – that is, places where robots have not traditionally been present, such as homes, assistive care facilities, hospitals, schools, and public spaces like shopping malls or restaurants.


  • Human-Robot Interaction: The overarching research area for the IRAL lab is human-robot interaction, or HRI: the study of making robots that can interact gracefully with people, rather than being confined to human-free areas and interacting only with trained specialists.
  • Grounded Language Acquisition: Learning models of language using data from the noisy, probabilistic physical world in which robots and humans both reside. This makes language learning easier (how do you learn the meaning of "green" without a camera?) and makes robots more able to understand instructions and descriptions.
  • Assistive Robotics: Robots that provide assistance with independent living for those who would otherwise be dependent on or otherwise make use of human caretakers. projects include assistance with medication management and cooking, and retrieving objects and information.
  • Mobile and Fixed Manipulation: Understanding instructions in the real world is necessary to perform tasks, but ultimately the difference between robots and computers is the ability to perform tasks in the physical world. We focus on building systems that use manipulation to accomplish tasks as needed by end users.
  • Environmental Monitoring: Another application area of human-friendly robotics is monitoring areas that are difficult or unsafe for people to access, including underwater reefs, inaccessible woodlands and the surface of other planets. Monitoring water quality and vegetation diversity requires the use of flying robots in tight communication with scientific specialists.
  • Engineering Diversity: As robots become more ubiquitous, the demographics of people who need to be able to use them expand. In order to build robotic systems that interact well with a variety of different groups, we actively explore ways of improving diversity in today's engineering world.