Human management of a hierarchical system for the control of multiple mobile robots

We have developed a hierarchy of human supervisory mediation levels into a robotic system. The purpose of this work was to create a semi-autonomous system which should enable completion of all feasible tasks. We propose the mediation hierarchy permits increased interaction between the human supervisor and all levels of a robotic system. This interaction permits the supervisor to maintain the system in a stable state. We have incorporated these hierarchical levels into MASC, a Multiple Agent Supervisory Control system. MASC provides the human supervisor with a three dimensional graphical human-machine interface. Our current test bed is the University of Pennsylvania's General Robotics and Active Sensory Perception (GRASP) Laboratory's multiagent system. This system includes two mobile observation robots and two mobile manipulatory agents. We designed human factors experiments to test the MASC system's usability as well as the mediation hierarchy theory. Novice users were recruited to participate in the experiments which included three different tasks: the single agent, the two agent and the four agent tasks. This thesis reviews the design of the experiments, analysis of the experimental results and a discussion of the results. The results are discussed in three perspectives, the MASC interface, the mediation hierarchy and the multiagent's system. In general, we found novice users were capable of executing the assigned tasks. We also substantiated the higher level interactions as well as the need for the lower level interactions of the mediation hierarchy. This thesis also presents a detailed review of all related research, as well as the contributions and future directions of this research.