Collaborating Human - Robot Swarms to Achieve Robust Space Exploration Capabilities 2006-01-2022

NASA has determined that human and robotic systems will cooperate to enable the space exploration enterprise. This entails many possible forms of interaction from essentially separate sequential activities linked by data exchange (e.g. robotic precursor missions enabling subsequent human exploration) to intensely coupled simultaneous collaboration in human - robotic teams. These interactions as well as the individual robotic and human exploration system components must be shaped to make the total system robust and flexible in the face of exploration challenges that cannot be fully defined or anticipated. One powerful concept for this purpose is found in recent research on robotic “swarms”. Interacting robotic swarms have been studied in numerous research efforts as a potential means of achieving flexible and robust capabilities with comparatively simple robots. The availability of multiple robots and their ability to interact allows complex performance from the ensemble beyond the capacity of any individual, and the inherent redundancy of the swarm leads to robust capability in the face of hardware failures and unforeseen specific challenges.

This paper presents the results of a preliminary collaborative Hamilton Sundstrand - MDA study of this concept as applied to human - robotic space exploration by “swarms” comprised of more than one human and more than one robot working in cooperation. Our study is investigating this unique type of team which is typical of anticipated NASA exploration scenarios. We are placing a special focus on the “mini-swarm” comprising 4 human EVA astronauts and at least two robotically enabled planetary rovers implicit for early lunar missions in the recent ESAS report. We are seeking to identify opportunities for new interaction paradigms and emergent capabilities that result as well the unique challenges they create and the physical, data, and command and control interactions that are required to enable them. Concepts we are exploring, their implications for exploration systems development, and future plans for our research effort are discussed.