Software agents and robots are becoming increasingly part of human lives. For example, mobile sensors are being deployed for a variety of military and civilian applications. These robots are for the most part teleoperated. We envision a future where large numbers of networked humans and robots will be working together for numerous applications ranging from manufacturing to environment exploration, crisis response and search and rescue. These multi-robot systems will be coordinating autonomously, and in many cases will be interacting with humans in a variety of interactions. I am interested in both (a) Developing algorithms and techniques for Autonomous Multi-Robot Coordination, and (b) developing a variety of techniques for Human Single- and Multi-Robot Interactions.
In the area of Autonomous Multi-Robot Coordination, I have particular interest and considerable amount of research in developing algorithms for decentralized task allocation with formal performance guarantees, negotiation algorithms for resolution of conflicts with multiple issues and algorithms for adversarial interactions. The algorithms take into consideration the dynamics of the environment, uncertainty and risk. These algorithms have wide applicability ranging from manufacturing, search and rescue, to package handling and transportation.
In the area of Human-Robot Interaction (HRI), I am working in developing a formal framework that models HRI from the perspective of computational complexity (constant, linear, exponential) of the human’s control task and needed interaction with the robots. This allows a systematic study of human interaction, adjustable autonomy, human-robot teaming and control with multiple robots. This framework has enabled us to develop effective algorithms for scheduling operator attention, and also discover a new concept, Neglect Benevolence, namely that it may be beneficial for a robotic swarm to be neglected for some amount of time before the human operator provides the next input, as mission goals change. Additionally, I am working towards the development of cognitively-based analytical models of the human operator so that the overall human-robot system can be formally verified.