Human-Robot Teaming

The HEAT Lab!

Temporal plans exist to guide robots to accomplish their goals, while also coordinating when these activities should occur. In general, we want temporal plans that are adaptable to events that are beyond the direct control of agents; e.g., a robot may experience slippage or sensor failures. To do this, we must answer questions such as how and when do new or unexpected events arise in practice? And how "good" is a temporal plan at adapting to these unexpected events?

Further, trust and cooperation are fundamental to human interactions. How much we trust other people directly influences the decisions we make and our willingness to cooperate. It thus seems natural that trust be equally important in successful human-robot interaction, since how much a human trusts a robot affects how they might interact with it.

In previous summers, we explored the usefulness of a new metric called robustness, which assesses the likelihood that a multi-robot plan succeeds. We showed that robustness is a better measure of multi-robot plan quality and that we can generate plans that optimize for robustness. We also explored whether humans behave differently in terms of their levels of trust and cooperation depending on if they are interacting with a robot or another human.

With these motivations in mind, this summer's projects include three specific subteams:

  1. Robot Brunch   How do we best disentangle interacting robots’ schedules?   In this project, we will
    • Construct new representations that better capture the uncertainty that interactions with others imposes on a robot’s plans
    • Design new algorithms for optimally decoupling multiple robots’ schedules
    • Evaluate our approaches using mutli-robot navigation tasks on the mock factory floor in our labs
  2. Human-robot teamwork   How do we schedule reliable interactions with human teammates?   In this project, we will
    • Create and implement new human-robot teamwork tasks that utilize our lab’s NAO and possibly other, new robots
    • Evaluate our approaches for measuring and generating robust multi-robot temporal plans on tasks requiring intricate scheduling
  3. Poly-bots   How many scheduling possibilities actually exist?   When robots team up with humans, we want to give human teammates as many possibilities as we can for completing their activities. However, counting how many possibilities exist, is actually very difficult! In this project, we will
    • Develop new approximate techniques for computing the number of schedules afforded by a temporal plans
    • Explore the connections between temporal plans and the geometrical construct called a (n-dimensional) polytope to leverage existing methods for computing the volume of n-dimensional polytopes
    • Design new tools for visualizing and measuring temporal plans

Interested?   Much more about these projects and their background are available from our HEAT Lab homepage

Join us!

Mentor: Professor Jim Boerkoel

Jim Boerkoel is an Assistant Professor in the Computer Science Department at Harvey Mudd College and director of the Human Experience & Agent Teamwork Lab. The goal of the HEATLab is to develop techniques that augment humans' own cognitive and physical abilities to create integrated human-agent teams that are more capable than their individual counterparts. Prior to joining HMC, Jim worked as a Postdoctoral Associate with the Interactive Robotics Group at MIT. He completed his Ph.D. thesis on developing distributed approaches for constraint-based, multi-agent scheduling with Ed Durfee at the University of Michigan.

Required Background

This is not a list of skills needed, but a list of skills that may be used—let us know if you have done some of this before (and how your background has prepared you to pick up the others!) The list: artificial intelligence, algorithm design, interaction design, ROS development, python, behavioral economics, social psychology, probability, information visualization, geometry