New Tool Tackles Uncertainty in Military Logistics Planning
By Matt Shipman, NC State News
Military vehicles and equipment belonging to the 1158th Transportation Company of the Wisconsin National Guard are loaded on railcars at Fort McCoy, Wisconsin. US Army Photo by Scott T. Sturkol/Released.
Military deployments to austere environments—whether humanitarian missions or combat operations—involve extensive logistical planning, which is often complicated by unforeseen events. Researchers at North Carolina State University have created a model to help military leaders better account for logistical risk and uncertainty during operational planning and execution.
Enterprise Resource Planning Systems
“Every branch of the military now uses an enterprise resource planning (ERP) system that handles everything from requisitions to shipment of supplies to inventory tracking,” says Brandon McConnell, corresponding author of a paper on the new model and a Research Assistant Professor in NC State’s Edward P. Fitts Department of Industrial and Systems Engineering. “These ERP systems make it possible to create computational models that can be used to identify the most efficient means of meeting the military’s logistical needs.
“These models would be particularly valuable during expeditionary operations, in which the military is seeking to establish its presence—and its supply chain—in an environment that is subject to a fair amount of uncertainty.
“The model that we’ve developed can not only facilitate the military’s ability to efficiently determine what will be needed where, but can also assess risk in near real time in order to account for uncertainty,” says McConnell, a former infantry captain in the US Army who served two tours in Iraq.
A New Planning Model
The new model, called the Military Logistics Network Planning System (MLNPS), draws on three sources of information. First is logistical data from the ERP system. Second is operational data, such as an operation’s mission, organization and timeline. Third is data on “mission specific demand,” meaning logistical requirements that are particular to a given mission and its environment. For example, a combat operation being conducted in a cold, damp environment would have different requirements than a humanitarian mission being conducted in a hot, dry environment.
The MLNPS also uses two factors to assess risk and determine how risk might affect military capacities. The first factor is the likelihood that an event will happen; the second factor is what the consequences of that event will be. For example, if the likelihood of two events is identical, the model would give more weight to the event that could have a greater adverse impact on military personnel and mission performance.
MLNPS Current Status
“Right now, the MLNPS is a robust proof-of-concept prototype, designed to demonstrate the potential value of powerful computational tools that can make use of ERP systems,” according to McConnell. “Existing logistical tools are both valuable and powerful. However, I’m not aware of any other methods that make use of ERP data and are also fast enough for operational use when time is of the essence.”
Efforts to capture readiness-based performance metrics and model a logistical plan while including the collection of contingency (branch) plans are already underway. Current efforts are also working toward integrating alternative routing and decision points into the modeling framework.
Future efforts should expand this emphasis on readiness and provide prescriptive decision-support capability rather than the predictive outputs this research paper illustrates. Outputs should include identification or visualization of the readiness-versus-cost tradeoffs.
The authors envision these ongoing efforts resulting in a risk mitigation design matrix to provide insights on how to ‘‘optimize’’ various resilient, robust, adaptive, or flexible logistics network mitigation strategies against potentially disruptive conditions or catastrophic events within an operational risk landscape (defined by event probability versus consequence).
“This research lays the mathematical and operational foundation for construction of a network-based model that captures routing alternatives and characterizes solutions for capacity planning and resiliency analysis in near-real time,” says Joseph Myers, Army Research Office Mathematical Sciences Division Chief at the Combat Capabilities Development Command’s Army Research Laboratory. “This project will provide military logistics planners with capabilities that are currently lacking in prevalent logistics planning tools.”
The methods the research outlines apply to other contexts that employ an underlying stochastic queueing network model that exhibits nonstationary and/or non-Markovian (nonexponential) arrival processes. Applications include disaster relief, humanitarian operations, and understanding international commercial global supply chain disruptions.
The paper, “Assessing Uncertainty and Risk in an Expeditionary Military Logistics Network,” is published in the Journal of Defense Modeling & Simulation. The paper was co-authored by Thom Hodgson and James Wilson, professors emeritus in the Fitts Department of Industrial and Systems Engineering; Michael Kay and Yunan Liu, associate professors in the Fitts Department; Russell King, the Henry Armfield Foscue Distinguished Professor in the Fitts Department; Greg H. Parlier, adjunct professor in the Fitts Department; and Kristin Thoney-Barletta, associate professor of textile and apparel, technology and management in NC State’s Wilson College of Textiles. The authors earned the 2019 Barchi Prize for best paper from the Military Operations Research Society. Support for this research was provided by the Army Research Office.
Uncertainty is rampant in military expeditionary operations spanning high intensity combat to humanitarian operations. These missions require rapid planning and decision-support tools to address the logistical challenges involved in providing support in often austere environments. The US Army’s adoption of an enterprise resource planning (ERP) system provides an opportunity to develop automated decision-support tools and other analytical models designed to take advantage of newly available logistical data. This research presents a tool that runs in near-real time to assess risk while conducting capacity planning and performance analysis designed for inclusion in a suite of applications dubbed the Military Logistics Network Planning System (MLNPS) which previously only evaluated the mean sample path. Logistical data from combat operations during Operation Iraqi Freedom (OIF) drives supply requisition forecasts for a contingency scenario in a similar geographic environment. A nonstationary queueing network model is linked with a heuristic logistics scheduling methodology to provide a stochastic framework to account for uncertainty and assess risk.