Analysis and improvement of batch-based production lines with Bernoulli machines

Steady state analysis of production systems have been extensively studied during the past several decades. This type of analysis is usually applicable in large-volume manufacturing as the transient can be ignored compared to the overall production run-time. However, when the volume of the production is relatively small, transient may play a major role during the production process and steady state analysis may become inapplicable. Such systems are commonly seen in practice, where production orders are handled as batches and a system works with one batch at a time. In this paper, under the framework of serial production lines with finite buffers and with machines having the Bernoulli reliability model, we study the problem of performance analysis and improvement of such systems. Specifically, we derive the mathematical model for the two-machine case and derive analytical formulas to evaluate production rate, consumption rate, and average completion time of the system. In addition, workforce allocation and bottleneck identification problems are discussed. For longer lines, preliminary results based on simulations are presented.