Waiting time estimation in a manufacturing system using the number of machine idle periods

To manage a manufacturing system, it is important to know the product flow time. Often a product's waiting time for machine service is a large portion of its flow time, and many mathematical formulae have been developed to estimate the waiting time. In a manufacturing system, these queueing formulae are important because they show the relationship between product flow time and some important production system control variables. However, most queueing formulae need the variance of inter-arrival time distribution, which is difficult to estimate in the manufacturing industry. This paper presents a new queueing formula which uses variables easier to estimate than the variance. The key variable among these is the number of machine idle periods. The formula can also quantify some modeling assumption errors (such as the Poisson arrival assumption error) and reflect the effect of scheduling on queueing estimation. Accuracy of the estimator is also discussed.