Effect of Variability on the Optimal Flow of Goods in Supply Chains Using the Factory Physics Approach

Abstract

In a supply chain, parameters such as demand, service times and the flow between nodes are commonly assumed to be constant quantities for the time horizon or window, while transformation is also assumed to be immediate. However, this is not the way it is. Demand is random; operations require a processing time that is also random, and this causes variations in the links of a chain that generate queues. Cycle time and the amount of work in process are properties that quantify the performance of the flow within a network subject to the effects of the system’s innate variability. This document provides a series of steps for analyzing optimal flow in a supply chain by using analytical methods. First, we use the minimum cost flow model to find the optimal solution for the distribution of the product in the chain. Then, assuming that each link in the chain is a G/G/c queueing system, the cycle time and quantity of work in process are quantified for the optimal solution in three scenarios of service time and demand variability. We used for our example the data corresponding to the three-level supply chain, in the context of the vehicle industry. Taking as a reference a scenario where variability is low (C²_s <= 0.5), the cycle time in the network increases its value to six times its value when there is high variability throughout the system.