APIS II: Extension and Flight Testing of the Autonomous Persistent Intelligent Swarm Algorithm

Multi-agent Cyber-Physical-Human (CPH) systems in realistic environments operate under uncertain conditions. Task inaccuracies, both in terms of exact location and exact priority, are to be expected. Furthermore, real-world deployment can encounter communication range limitations, leading to disconnected swarm components over short and medium periods. The APIS algorithm, first presented in [1], addresses this issue through a bee-foraging heuristic approach to task completion, balancing direct task completion (harvesting), mapping (dancing), and informational interchange (dancing). We propose to implement the Autonomous Persistent Intelligent Swarm (APIS) system in a real-world heterogeneous swarm setting, evaluating its efficacy in maintaining lines of communication, reducing performance uncertainty, and maintaining swarm cohesion. We will undertake two parallel efforts: First, improving the fidelity of the APIS simulation and evaluating its mission efficacy in various system configurations. Second, creating a physical APIS-compatible system capable of completing various tasks and mapping the environment to validate the simulation results and serve as a proof-of-concept technology demonstrator.