NAVSUP Weapon Systems Support Reforming Wholesale Predictive Model for Improved Readiness

Jan. 24, 2019 | By kgabel
By Kelly Luster, Director, Office of Corporate Communications, NAVSUP Weapon Systems Support The U.S. Navy has a global reach with nearly 300 deployable battle ready ships, 92 submarines, and more than 3,700 operational aircraft operating in oceans and waterways around the world. Each year there are more than 500,000 equipment and repair part demands from a $34 billion inventory of more than 30 million individual parts in support of Navy, Marine Corps, Joint and Allied Forces customers worldwide. Ensuring the Navy has the right materiel, when they need it, and where they need it, falls to the Navy’s only Program Support Inventory Control Point (PSICP), NAVSUP Weapon Systems Support (WSS). In recent years, the Department of Defense (DoD) and the Navy have looked for operating efficiencies while increasing readiness, leading to increased focus by the Office of the Secretary of Defense (OSD) on metrics like wholesale demand forecast accuracy. Through the DoD Comprehensive Inventory Management Improvement Plan (CIMIP), a set of DoD-wide metrics were created to measure forecasting performance. Through its reform process, NAVSUP WSS has targeted wholesale predictive modeling as a focus area for improved planning and forecasting ability. Refining the ability to more accurately plan what customers need will lead to a more agile and lethal Navy. By refining its wholesale predictive model, NAVSUP WSS may be part of the solution to meet Secretary of Defense James Mattis’ requirement recently outlined in a memo to several undersecretaries. In the memo, Mattis wrote that in order to see effective and efficient change, the departments must focus on meeting the most critical priorities first. He wants to see reduced operating and maintenance costs and an 80 percent mission capability on the Pentagon’s F-35, F-22, F-16 and F-18 inventory platforms. According to Lt. Cmdr. Duncan Ellis, an operations research analyst at NAVSUP WSS, trying to determine what to buy, how much to buy and where to put it, is very challenging. NAVSUP WSS can have a greater success by leveraging more of its available data as a springboard to build better predictive models; enabling greater accuracy and efficiency and helping commanders determine risks when making inventory policy decisions. Working with the Navy Post Graduate School and Penn State University, Ellis is leading a team developing a new model for NAVSUP WSS to replace the current Wholesale Inventory Optimization Model (WIOM 4.0). “Our goal is to be ahead of the problem,” said Ellis. “The new model will help us better look at risks to materiel availability caused by uncertainty in both fleet demand patterns and our repair and procurement pipelines, with the goal of helping NAVSUP WSS make more sophisticated investment decisions. We need to ensure we make efficient use of the Navy Working Capital Fund (NWCF) to invest in the right materiel to support future warfighter requirements.” Understanding the hundreds of thousands of National Item Identification Numbers (NIINs), and the factors affecting the supply chain, is a formidable task. “At any given time, the Navy’s inventory is distributed across the entire global operating environment. Some is in warehouses, some is with the customer on ships or forward deployed, and some is at organic and commercial repair sites,” said Ellis. Managing this rotational pipeline of stock is a challenging task. “Additionally, we support materiel over a diverse demand portfolio. While some materiel has predictable regular and steady customer demand, a large proportion of our items have irregular, episodic, or very limited demand. Essentially, there is no easy way to go about creating a model to account for everything.” Ellis quotes George Box, a man often referred to as “one of the great statistical minds of the 20th century.” Box stated that “all models are wrong, but some are useful.” The point Box was making is that models cannot account for every variable or piece of data affecting the outcome they’re modeling, but that doesn’t mean they aren’t useful. Model design is an exercise in determining the required complexity to inform a particular task. “Models are a bit like maps,” said Ellis. “Maps contain a level of complexity required to solve a particular problem, for example, locating a position or navigating between points. A map with insufficient detail to perform the required task is useless. However, a map with the same level of detail as reality is too complex, and equally as useless as the map with insufficient detail. Models, like maps, are purposeful simplifications designed to provide insights into the behavior of complex systems and aid decision making.” The new wholesale model, WIOM 5.0 will enable NAVSUP WSS to directly leverage the millions of rows of transactional data the ERP system accumulates. “This broad data set will enable us to better understand the range of effects across multiple metrics of interest for potential inventory investment decisions,” said Ellis. Previous models like WIOM 4.0 rely on fairly simple input data sets, and often make suspect assumptions about forecast accuracy or variation. A big change in moving to WIOM 5.0, is rather than relying on forecasting data, it looks directly at historical transactional records to provide a robust solution that relies on a minimal number of assumptions. WIOM 5.0 combines a sophisticated inventory simulation tool with a resource optimization model. The simulation will leverage current asset posture with historical transactional data to include demand, procurement, and repair times, to assess the performance of inventory level decisions. The optimization will use the performance metrics from the simulation to help NAVSUP WSS make efficient and effective inventory policy tradeoffs. The NAVSUP WSS WIOM development team has an aggressive timeline. The current plan is to have a prototype version of the new model completed by the end of 2018. Following initial release, technical testing, fine-tuning, and integration is planned for the spring with production implementation scheduled for late summer of 2019. As the critical link between the supply chain and the warfighter, NAVSUP WSS’s launch of WIOM 5.0 will translate directly to warfighter readiness and Navy lethality. Winter 2019