A Modular Swarm Optimization Framework Enabling Multi-Vehicle Coordinated Path Planning

Period of Performance: 05/17/2016 - 12/16/2017

$750K

Phase 2 SBIR

Recipient Firm

Heron Systems, Inc.
22685 Three Notch Road Array
California, MD 20619
Firm POC
Principal Investigator

Abstract

The advancement of Unmanned Aerial Systems (UAS) with computing power and communications hardware has enabled an increased capability set for multi-vehicle collaborative operations. By cooperatively allocating unmanned resources, vehicle tasking, and planning the subsequent vehicle paths, the efficiency of UAS operations can be maximized. Heron Systems proposes to develop the Multi-Agent Cooperative Engagement (MACE) framework into a mature prototype that enables collaborative resource allocation, task allocation, and path planning for unmanned systems operating in dynamic environments subject to diverse communication conditions. This Phase 2 work will focus on refining the path planning portion of MACE as well as maturing the resource and task allocation library developed during Phase 1. The path planning architecture will define key modules to plan paths to a global objective, assess potential obstacles, and avoid collisions while maintaining progress towards the global objective. The framework will be constructed in a modular fashion to allow a plug-and-play capability for the resource/task allocation as well as the various components of the path planning pipeline, giving end users the flexibility to explore other methods for UAS collaboration. At the conclusion of Phase 2, the MACE framework will be demonstrated using Heron Systems� HWIL simulation/stimulation environment. Once verified via the HWIL environment, the MACE framework will be deployed onboard several aerial assets and tested against scenarios specifically tailored towards precision agriculture applications.