Initial Development of Cooperative Control and Localization of Multiple Spacecraft Using a Multi-Agent Mission Operations System

Multi-satellite swarms are becoming very popular due to their low costs and short development time. Instead of large and costly monolithic satellites, small satellite swarms can be flown as distributed sensing applications for atmospheric sampling, distributed antennas, synthetic apertures among other exciting applications, delivering an even greater mission capability. This paper reports on the results of a NASA STTR Phase-I project that contributes to the development and demonstration of a mission operations system for robust, coordinated operation of mobile agent swarms in dynamic environments with a high degree of autonomy. Through a collaboration with the University of Hawai’i at Mānoa, Interstel Technologies’ Comprehensive Open-architecture Solution for Mission Operations Systems (iCOSMOS™) is being enhanced to coordinate and control swarms of space vehicles and other assets. The proposed iCOSMOS-Swarm™ will enable motion planning for large numbers of agents in densely crowded areas and robust position estimation with built-in cooperative localization. The major tasks for Phase-I included (1) the development of a scalable multi-agent coordination module to coordinate large agent swarms, a multi-nodal software architecture for diverse (heterogeneous) assets, and hierarchical cooperative localization module for robust inter-agent positioning, (2) enhanced system performance with improved data handling and nodal message passing and dynamic system configuration for node addition and removal, and (3) significantly enhanced simulation capabilities to eventually support up to at least 100 simultaneous nodes and end-to-end simulation of five satellite nodes in real time or up to at least 1000x real time. The results for Phase-I include the design of the iCOSMOS-Swarm™ product and the verification of the methodology using simulation results from a baseline benchmark mission with one microsat and four CubeSats to collect dynamic, multi-dimensional data sets over a wildfire outbreak or similar event through the use of multiple detectors, spread out in time, space, and spectrum.