Manuel D. Rossetti (PI); Kevin Taaffe (PI); Keith Webb (RA); Kyle Lassiter (RA)
Department of Industrial Engineering – University of Arkansas
The main goal of this research is to review methods and software to deal with less than perfect information in the supply chain and to develop a software tool to enable analysts to better understand how data uncertainty can lead to risk in the supply chain. Current inventory models assume perfect knowledge about costs and the distribution of future demands and lead time. However, it is usually the case that the level of knowledge is less than perfect due to estimation errors or imperfect data. Failing to account for data uncertainties can lead to failing to meet customer service standards or to excessive supply chain costs.
Technical Approach: The project developed a web application that runs Monte Carlo simulations of (r,Q) inventory models. The application uses random distributions to represent uncertain parameters in the model. Algorithms developed in a previous CELDi project for optimizing and reporting the performance of (r,Q) models were used throughout the application. The Vaadin framework was used to develop the graphical user interface in Java.
Results: A web based software tool implemented in Java that allows the user to easily set up and run Monte Carlo simulations of inventory models with uncertain cost and demand parameters.
Broader Value to CELDi Members: This proposed research has the broader impact of allowing analysts to evaluate if an item poses a significant supply chain risk, estimate the value of improved information in the supply chain, and determine which inventory model parameter is most contributing to performance uncertainties.
Future Research and Potential Extensions: Adding file input and output. Adding modules to allow simulations of periodic review and other types of continuous review inventory models. Improving the visual presentation of results.
Login and click link below to view Full Report
You must be logged in to view this content - login