Automated Mission Scheduler (AMS): Agent Facilitated Collaboration for Optimal Utilization of Multiple Unmanned Vehicle Systems

This proposal is for Phase II of an SBIR project to develop an Automated Mission Scheduler (AMS) system that employs agent facilitated collaboration for optimal utilization of multiple unmanned vehicles (UVs). A major requirement is therefore to create a collaboration management system and interface that will allow Clients to negotiate rapidly among themselves and with the UV Operators to select a mutually acceptable UV allocation and tasking plan. Our innovative technical approach has three main components: (1) Adapt the formalism of the Distributed Constrained Optimization Problem (DCOP) to the situation of optimizing the shared use of multiple UV resources; (2) Integrate the AMS into the Mixed-Initiative Interface Module (MIIM) as a tasking and multi-vehicle mission planning component; and (3) Design and develop an effective Collaboration Management System (CMS) with a Collaboration Management Interface (CMI) based on the application of relevant psychological and human factors principles. Our successful Phase I effort resulted in a complete end-to-end, man-in-the-loop demonstration of all the key AMS capabilities. Our Phase II plan includes development of a complete AMS-STEE product and its transition to several prestigious Naval and SOCOM activities for field evaluation, continued research and development, and introduction to already identified and potential operations users.