Proceedings of the 2003 Winter Simulation Conference S. Chick, P. J. Sánchez, D. Ferrin, and D. J. Morrice, eds. T.LOADS TREATMENT OF ASSIGNING AND FILLING ORDERS Bob Hamber Tactical Logistics Modeling & Simulation Team Expeditionary Systems Division Naval Facilities Engineering Service Center Port Hueneme, CA 93043-4370, U.S.A. 1.2 Description ABSTRACT At its core, TLoaDS is a supply chain throughput model. It simulates how distribution resources like pallets, forklifts and trucks work together to move supply commodities from supply points forward to where they are consumed. The core module is a stochastic discrete event simulation, like the commercial supply chain, warehousing and factory industries now turn to for critical problems. It uses a cur- rent commercially available supply chain modeling envi- ronment, with custom government owned TLoaDS code and user interface components. When it runs, the simula- tion engine keeps track of how all the individual commodi- ties, orders, shipments, material handlers, and transporters interact with each other and their current environment ac- cording to the current process rules. The Tactical Logistics Distribution System (T.LoaDS or TLoaDS) is a powerful and flexible simulation application for assessing current or future tactical distribution systems. It is a discrete event analytical model for assessing the pros and cons of new doctrine, distribution techniques, organ- izational structures, and equipment concepts. It can also be used for mission planning allocating available resources to sustain a military force in a wide variety of scenarios. This paper covers how TLoaDS models the order assigning and filling process. While order assigning is simple, the order filling process dynamically selects the combination of con- tainer, handler, and transporter appropriate for the current order, conditions and rule set. A scoring system with user adjustable weighting factors allows biasing the shipment plan towards more efficient use of transporters, or more re- sponsive filling of orders. As TLoaDS runs, it generates a wide variety of time- based, quantified, and descriptive, text, plotted, and ani- mated outputs that shed far more light into what is going on in the system, than spreadsheet, knowledge base, neural network, or linear programming models do. 1 INTRODUCTION TO T.LOADS Numerous non-core TLoaDS modules described in a previous WSC paper on the TLoaDS architecture (Hamber 2001) aid the analyst in managing and processing the vast amount of input and output data involved in a study. These include modules to 1) manage the different simulation lev- els involved in a study; 2) help prepare and advise the ana- lyst in setting up the model inputs; 3) understand the de- tailed and overall performance of the distribution system. There is even on-line documentation in the familiar Micro- soft Help environment. With animation on, the nominal run speed for a 50 node model without atrocious bottle- necks is 1 scenario-day per run-minute on a 1GHz PC with 512M of RAM. With animation off, it runs four to ten times faster. 1.1 Need New doctrine, procedures, techniques, organizations, and equipment impose challenges to Navy and Marine Corps expeditionary logistics. These subsystems interact and warrant a model that simulates these dependencies. The military supply chain is substantially different from commercial supply chains, so commercial supply chain models are inadequate. Three of the key differences are: 1) tactical supply points move, whether ships, combat service support detachments, or combat trains, while ware- houses, distribution centers and stores, rarely do; 2) tactical nodes can have chronic shortages of resources, while in the commercial world, more resources are usually acquired as needed; and 3) the tactical distribution system is subject to enemy attack; in the commercial world this is still too rare to model. For more information on TLoaDS, see our website: <http://www.nfesc.navy.mil/amphib/tload s/index.html>. 1026