The Effect of Product Structure Complexity and Process Complexity on Optimum Lot Size in Multilevel Product Scheduling

Suharyanti, Yosephine and ARIYONO, VINCENTIUS (2010) The Effect of Product Structure Complexity and Process Complexity on Optimum Lot Size in Multilevel Product Scheduling. In: The 11th Asia Pacific Industrial Engineering & Management systems Conference, 7-10 Desember 2010, Melaka, Malaysia.

Preview

Text C13_The effect of product structure.pdf Download (2MB) | Preview

Abstract

It is difficult to justify the scheduling result before creating the schedule itself. In the other hand, for practical purposes it is important to make scheduling decision immediately to respond the orders. This research concerns to find a short-cut justification of lot size in order to minimizing makespan for scheduling of multi-level products in job shop environment. It was hypothesized that as the complexity of scheduling problem, the lot size decision is a function of operations complexity. In this research, operation complexity is presented by product structure complexity and process complexity. Number of level, number of part in each level, and total number of part in each product are the parameters representing product structure complexity. Setup time, run time, time variations, number of operation applied on each part, and routing complexity are the parameters corresponding to process complexity. Up to now the research still has not covered yet the evaluation of routing complexity. The analysis was based on a set of hypothetical data, those are limited up to 5 levels of product structure, up to 5 items in each level, up to 21 items in each product, setup time-unit run time ratio up to 10, and number of operations up to 5 for each item. Several lot size decisions are applied in the scheduling of some replicated cases, and the lot size giving the minimum makespan is pointed as optimum lot size. To get a general parameter for lot size, the term “setup time -total run time” ratio (R) was formulated, and optimum lot size is represented by R*. The result showed that number of levels, average number of item in each level, setup time-unit run time ratio, and variation on number of operations are the factors affecting R*, as proved by these respective correlation coefficients: 0.26; -0.21; 0.43; 0.30. The value of most R* (91.09%) is varied from 0.1 to 0.4.

Actions (login required)

View Item