MaintenanceMate™: An Intelligent Troubleshooting Agent for Advanced Maintenance

As weapon systems continue to grow in numbers and sophistication, they pose an ever-increasing challenge to the maintenance community. Compounding the problem is the fact that seasoned maintainers are continuing to retire in large numbers. This leaves junior maintainers having to fend for themselves when troubleshooting systems. What junior maintainers need today is computer-based assistance in the form of an expert troubleshooting aid that embodies the knowledge of expert maintainers. In response to this need, this project is developing an intelligent troubleshooting agent (ITA), that is capable of guiding/instructing junior maintainers in complex equipment troubleshooting tasks that require the use of maintenance test equipment. Phase I of this effort resulted in a prototype system that demonstrated the interactions of the ITA with junior maintainers in complex troubleshooting tasks. Phase I also presented the advantages of the proposed solution over existing methods. Phase II is concerned with implementing the ITA and demonstrating its troubleshooting, active learning, and failure prediction capabilities on a selected subsystem of the LAV-25 weapon platform.