An Intelligent Agent-based Self-evolving Maintenance and Operations Reasoning System

Abstract

Joint Strike Fighter (JSF) autonomic logistics seeks to reduce development, production, and ownership costs for the next generation fighter aircraft by increasing system reliability, while reducing maintenance requirements to essential levels. Prognostics and health management (PHM), which enables maintenance to be planned on the basis of actual component or system health state, represents a key component within the autonomic logistics system architecture. The challenge is to develop advanced technology to integrate PHM information from a variety of different sources into a dynamically evolving knowledge base. Prototype software described herein and referred to as the self evolving maintenance and operations reasoning system (SEMOR), utilizes intelligent software agents in JADE, both model and case-based reasoners and reinforcement learning modules. The fundamental approach enables PHM reasoning to be effective in the absence of field experience through the model-based reasoning module as well as realize the benefits of case based reasoning as a PHM knowledge base grows. A reinforcement learning (RL) module is employed to evolve a maintenance integrated model (MIM), a database containing PHM and maintenance relationships and attributes. Intelligent software agents are used in their true capacity to negotiate decisions regarding database adaptation, maintenance, and logistics actions prior to human review. This paper presents the software system design, describes key technical components, provides a demonstration scenario and concludes with remarks on the technical challenges and future developments.