Dependency Between Performance of Production Processes and Variability - an Analysis Based on Empirical Data

It is commonly accepted that variability is one of the main challenges in designing and managing manufacturing processes. Many process improvement concepts that focus primarily on communication and information exchange, flow time reduction, etc., finally influence variability. In particular, they reduce variability or mitigate the operational effects of variability. Vendor managed inventory or collaborative planning, forecasting, and replenishment are just two examples of such concepts. The effect on selected performance measures is mostly shown by idealized quantitative models, but there are only few results from real-world processes. In our study we want to illustrate and quantify the impact of variability on the performance of production processes by means of two real manufacturing processes. Case process one is an assembly process of frequency inverters and case process two is the assembly process of sliding glass top systems. For the inverter assembly process we want to show the operational impact of reduced demand variability (reduced forecast error), achieved by implementing Vendor managed inventory as well as collaborative planning. In the glass top assembly process internal variability is addressed by assessing the impact of the production lot size. Both processes are mainly evaluated by using WIP and flow time as key performance measures. The analysis is conducted with rapid modeling software based on open queuing networks. Our results show that a reasonable decrease in inventory and flow time can be achieved without any decline of customer service.