Manuel D. Rossetti
University of Arkansas
The goal of this CELDi Center Designated Project is to enhance previously developed multi-echelon supply chain simulation software so that it can be used by CELDi organizations through a SaaS architecture that leverages the resources available within cloud computing (CC)
At the end of this project, CELDi member organizations will have available to them a SaaS/Cloud computing application capable of evaluating the performance of multi-echelon supply networks through simulation.
This software architecture will facilitate software benefits based on a usage model and develop expertise within the center for this new computing architecture.
Realistic large-scale multi-echelon supply chain networks consist of thousands of stock keep units (SKUs), which make the analysis by mathematical models problematic. The complexity of modeling and analyzing multi-echelon supply networks necessitates the need for simulation methods. Simulation is an excellent tool to assist in supply chain analysis since it can (1) predict supply chain performance, (2) allow users to track the system performance at each time period, (3) perform what-if analysis for specific scenarios, and (4) compare alternative system configurations. However, simulations of the size required to model realistic supply networks need large amount of computational resources and may be very time consuming to get statistically accurate results. The purpose of this project is to design a cloud computing architecture that facilitates the computational performance of large-scale multi-echelon supply chain network simulations. The simulation results generated by the SCNS were tested by two test cases, and the results are close to the estimated values from a journal article. Also, the DA generates valid input Excel files that can be simulated by the SCNS, and the output data generated by the SCNS can be imported into and analyzed by the DA. In addition, a test case with 120 SKUs was developed and run for 30 replications to test the simulation time. The results show that total time spent on the cloud computing solution is 35% less than the time spent on the traditional solution. The time spent on the cloud computing solution includes the startup time of virtual machines (VMs), the scheduling time, the execution time and the file transfer time. Without the startup time of VMs, the cloud computing solution can save 70% of the simulation time.