Human-swarm interactions based on managing attractors

摘要: Leveraging the abilities of multiple affordable robots as a swarm is enticing because resulting robustness and emergent behaviors swarm. However, swarms are composed many different agents, it difficult for human to influence by managing individual agents. Instead, we propose that should focus on (a) higher level attractors system (b) trade-offs appear in mission-relevant performance. We claim theoretically allows abstract details agents collective whole. Using model with two attractors, demonstrate this concept showing how limited can cause switch between attractors. further using quorum sensing manage scalability interactions mitigating vulnerability agent failures.