The power of collaborative systems
The power of collaborative systems
As advances in technology enable the introduction of multiple unmanned vehicles (UVs) into the battle space, those vehicles must be managed by distributed human operators as an integrated total resource that supports higher-level objectives while operating at a rapid operational tempo. The critical problem in shared use of the multiple unmanned vehicle capabilities by a number of operators is finding an optimal method for resource allocation and scheduling. This R&D effort is directed toward developing an Automated Mission Scheduler (AMS) that uses the most advanced multiagent constraint optimization technology to facilitate users’ collaboration in asset management.
For this ONR-supported SBIR Phase I effort, Perceptronics Solutions has teamed with the USC Computer Science Department TEAMCORE Research Group and Charles River Analytics, Inc., and is further supported by Professor Mary (Missy) Cummings of the MIT Aeronautics and Astronautics Department.
Our system concept is based on innovative technology products contributed by our team members. At the heart of our approach is the TEAMCORE Group’s ground-breaking work on solving the Distributed Constraint Optimization Problem (DCOP) to optimally allocate multi-vehicle tasks in collaborative control of UV systems. Charles River Analytics Inc. has developed the Mixed Initiative Interaction Module (MIIM) to provide an advanced human/computer interface (HCI) supporting an operator’s command and control of a group of heterogeneous unmanned assets in coordinated littoral operations. Perceptronics Solutions’ Tactical Group Decision Analysis System (TGDAS) directly supports the project by providing an intelligent workflow mechanism which drives the team collaborative decision process; and our Mixed Initiative Team Performance Assessment System (MITPAS) provides a OneSAF Objective System (OOS) based UV simulation environment including proven scenarios and metrics.
Our innovative approach is: Adapt the formalism of the DCOP to optimize the shared use of multiple UV resources by multiple operators. Integrate the AMS into the MIIM as a tasking and multi-vehicle mission planning component that allows a variety of domain-specific mission objectives to be added at a higher-level without changing the core lower-level autonomy software for each vehicle class. And use the OOS- based MITPAS as a simulation environment for the Phase I proof-of-concept demonstration and as an actual robotic control interface for potential Phase II field evaluation.
Home | About Us | Projects | News & Events | Papers | Contact Us
© 2007, Perceptronics Solutions, Inc. All rights reserved. | Web Site Design by Capital Idea Ventures, Inc.
