Scott J. Mason, Edward A. Pohl, Heather L. Nachtmann, Manuel D. Rossetti
University of Arkansas
This project has the objective of working toward information superiority through enhanced cognitive decision support, integration of course of action analysis, knowledge management and mining, and pattern recognition and learning into the network and improve the logistician’s ability to convert data and information into relevant knowledge and understanding. Also, through network and modeling analysis. this project attempts on creating a framework that can be used to construct mathematical and logical models of sense and respond (S&R) military logistics networks.
The modeling and analysis of traditional military logistics networks is based on the paradigm of nodes and arcs. Nodes in this network typically represent manufacturers, warehouses, depots, and bases that exist at fixed locations. The relationship of each node with respect to the other nodes in the network collectively defines the echelon structure of the overall supply chain. Each base may use several depots, and each base usually acts independently of one another. Further, each depot, as well as each warehouse, often acts independently of their counterparts.
The arcs connecting the nodes represent transportation links between locations. Network flow occurs when an item is transported from one echelon to an adjacent (typically downstream) echelon. These items (entities) are often considered to be either non-repairable (i.e., consumable) or repairable. The driving force behind entity flow is base-level demand. This demand is often assumed to be static and either deterministic or probabilistic with a known underlying stationary distribution. In addition, item lead times are considered to be static in most traditional military logistics network analyses. Demand fill rate and inventory levels (position) are the two most frequently considered performance metrics of interest.
With the evolving notion of S&RL, the structure of the traditional military logistics network, while still being based primarily on nodes and arcs, will become both dynamic and adaptable. While some network nodes will continue to follow the traditional role of a fixed supply chain location, other nodes representing infantry units, maritime vessels, and so on, may “move.” Other nodes may be “temporary,” such as those relating to contingencies that are only active/probable for some finite amount of time. While some adaptable network nodes can be thought of as “potential nodes,” other nodes can be “lost” with some non-zero probability due to opportunities/missions that are no longer viable.
Recent and future developments in sensing technologies will provide better/improved access to real-time node locations and their status. The military’s vision of total asset visibility creates the potential for lateral supply between nodes in the adaptable network. This, in turn, leads to increased network connectivity and hopefully reduced item lead times. In fact, arcs in an adaptable network can be permanent or temporal, added or removed, created or lost. While current bases will be characterized by dynamic, but less variable demand, potential bases/demand locations are sources of uncertainty in an adaptable network. Although traditional performance metrics such as fill rate and inventory levels still apply, newer, more appropriate performance metrics, some of which may be evaluated simultaneously within the framework of a multi-objective problem, must be developed for adaptable networks. For example, performance metrics such as network responsiveness, vulnerability, and position may help to properly evaluate possible decision options in adaptable networks.
*New modeling paradigms for supply chain simulation
*New simulation methodologies, mathematical models and techniques for the optimization, performance evaluation, and improvement of future military logistics support networks
*New methods of modeling and incorporating human performance issues and collaborative decision making within an S&RL environment