Ssci and university of washington (prof. Behcet acikmese) propose to develop, integrate and test an innovative distributed intelligent swarm control & utilization system (DISCUS). The DISCUS will be…

This STTR Phase I contract was awarded by the National Aeronautics and Space Administration Shared Services Center to Scientific Systems Co Inc., a small business, to develop an innovative Distributed Intelligent Swarm Control & Utilization System (DISCUS). The $124,962 effort will focus on requirements, algorithm development, and initial integration of guidance, navigation, and control algorithms for swarm applications, with plans for further maturation and demonstration under realistic conditions in a Phase II effort. DISCUS will leverage advances in distributed state estimation, probabilistic guidance and control, collision avoidance, and fault-tolerant control to enable effective and resilient swarm behavior. This research supports the broader advancement of autonomous systems, artificial intelligence, and unmanned vehicle capabilities, which are high priorities for agencies like NASA, DARPA, and the Department of Defense.

Ssci and university of washington (prof. Behcet acikmese) propose to develop, integrate and test an innovative distributed intelligent swarm control & utilization system (DISCUS). The DISCUS will be based on advanced distributed state estimation techniques, probabilistic guidance and control under collision avoidance and other relevant mission constraints, real-time contingency management including reactive collision avoidance with UN-responsive team members, and low-level fault-tolerant control robust to subsystem and component failures. Decentralized estimation is based on using RSS (received signal strength) and TOA (time of arrival) sensors, and fusion of information from EO (electro- optical) sensors. Guidance and control (G & C) is based on extensions of an innovative approach to swarm density control using a markov chain monte carlo (MCMC) approach with guaranteed satisfaction of the ergodicity, motion, and safety constraints. Reactive collision avoidance will be based on extensions of a suite of SAA algorithms previously developed or under development by ssci, while fault tolerance will be achieved by combining ssci's approach to fault detection, identification and accommodation (FDIA) with low-level baseline control. Focus on phase I will be on the requirements and algorithm development, initial integration of a diverse suite of GNC algorithms, and feasibility demonstration on a simplified swarm simulation. Phase II will involve further maturation and full integration of DISCUS algorithms, and their demonstration under realistic conditions through hardware experiments.