Our goal is to develop algorithms for allowing robots to adapt intelligently to humans when they are working together as a team.
There has been much prior work on getting robots to adapt to disturbances in their environment, and also much work on recognizing human intent. This work aims to fuse the two together, for the case in which we have flexible plans for the human-robot team. By analyzing the structure of the plan -- including the actions and their requirements, timing constraints, and allowable choices for the humans and robots -- we can provide a single framework that allows a robot to adapt to its human teammate's intent.
