A Novel Adaptive Intelligent Agent Framework for Agile Supply Chain Management supporting Naval Weapons Planning, Plan Repair, and Scheduling

This document presents a computational model for problems pertaining to the planning, plan repair, and scheduling of weapons handling in aircraft carriers. This class of complex problems can be represented by a supply chain of a manufacturing enterprise with networked weapon component suppliers, factories, warehouses, quality assurance sites, and distribution centers. Through the supply chain, weapons components are acquired, transformed into weapon products, and delivered-to or returned-from customers. In the proposed framework, supply chain elements such as resources, functions, and decision-points, are represented by autonomous intelligent agents using a holonic manufacturing system (HMS) framework. This project will research HMS architectures based on a loose aggregation of intelligent agents that collectively represent an agile weapon supply chain management system. We propose the concept of groups of agents based on the biological cell paradigm. Agents form virtual organizations able to adapt constantly to changing missions. Agents will be able to learn from their past experience as well as from observing changes and corrections made by human weapon handling experts. The proposed framework is generic and will be applied to multiple planning problems related to weapon handling aboard aircraft carriers: weapon arrangement in magazines, weapon on-load and off-load, and plan repairs during weapon on-load execution.