The Impact of Lot Sizing on Queuing Delays an Educational Software Tool

It is now widely accepted that both large and small lot sizes can cause long lead times and consequently bad customer service in terms of late deliveries. The impact of the lot size on the lead time consists of a convex relationship, implying an optimal lot size minimising average lead time. In order to educate people with this undoubtedly correct, but yet controversial issue, we developed an educational software tool which helps to clarify the important determinants and allows for self-tuition and experimentation. Therefore, in the first section we explain why this phenomenon is so important and how it can be applied in practice. In the second section, we introduce a small example and explain the theoretical foundations of the model. The third section describes the graphical user interface. ACLIPS: A CAPACITY AND LEAD TIME INTEGRATED PROCEDURE FOR SCHEDULING Aclips is a general hierarchical procedure for job shop planning and scheduling developed by Lambrecht, Ivens and Vandaele. [Lambrecht, Ivens and Vandaele, 1998] The procedure consists of four phases. The first phase includes a lot sizing decision and a lead time approximation phase. In the second phase, the tuning phase, management can intervene to obtain acceptable product lead times. The third phase deals with scheduling and therefore includes manufacturing order grouping, release date setting and detailed scheduling. Finally in the execution phase, the schedules are transferred to the shop floor and are executed. This four-phase hierarchical approach is illustrated in figure 1. Lot Sizing Phase There is a convex relationship between lot size and lead time. Three main effects influence lead times: batching, congestion and stochasticity. Large lot sizes lead to long lead times because it ties up resources for long periods of time. This effect is called the batching effect. On the other hand, due to frequent setups, small manufacturing lot sizes cause a lot of lost machine time. As a consequence, both utilisation and total lead time increase. [Karmarkar, 1987] In addition, variability of both arrival, production and setup times has a substantial influence on lead times. In steady state, increasing variability will always increase average lead times and thus work-inprocess levels. [Hopp and Spearman, 1996] 1 This research was supported by the Flemish Science Foundation (FW0-project G.00 63.98), the BOF fund of the University of Antwerp and WTCM, Belgium. The author acknowledges Michaël Dumon and Frank Hoflack who did an excellent job on the software development. Lot Sizing & Lead T ime Approximation Phase