Next Generation End-To-End Logistics Decision Support Tools. Evolutionary Logistics Planning

Abstract

Logistics planning and decision support systems have traditionally focused on planning large scale military operations with limited forecasting and execution tools causing many military logistics support tools to fall short of providing a true end-to-end solution. A true end-to-end solution will yield a system that can be used for logistics training, long-term logistics planning operations, real-time logistics planning and execution during an operation, and real-time decision support for immediate replanning and response to ongoing operations for all echelons of a military hierarchy. In this paper we will explore technologies that will provide flexible and accurate plan development leading to better plans, increased decision support, and ultimately better execution of military logistic plans. Advanced logistics planning and forecasting tools built by DARPA projects such as the Advanced Logistics Program (ALP), Ultra*Log, and Network Centric Logistics (NCL) successfully implemented capabilities that provide portions of an end-to-end logistics solution. These systems were built using the cognitive agent architecture (COUGAAR) which provides support for large multi-agent systems that require distributed processing and allow for numerous applications and technologies to be seamlessly integrated into large scale logistics systems. In order to provide the next generation of forecasting and execution utilities that will lead to an end-to-end solution, large multi- agent systems will need to incorporate technologies that provide the following attributes: technologies that isolate and focus on specific areas of a plan, technologies that provide greater flexibility in planning and technologies that will provide a mechanism for human interactions. Under the solutions section of this paper four technical solution areas are discussed: 1) Optimized distribution 2) Evolutionary planning 3)Focused forecasting 4) Execution and simulation. Existing and new techniques in these areas will provide the necessary logistics planning attributes for the next generation of logistics decision support systems