| Summary: | 碩士 === 元智大學 === 工業工程與管理學系 === 99 === The case company is one of the domestic companies that produce pellicle. In the front-end process of the pellicle, because the number of fixtures is limited, lots of products share the same type of fixture, and the time of changing fixtures is tedious, production planners usually arrange the processing sequence of orders based on the change of fixtures, and then the due date of the orders. However, the current dispatching rule often leads to earliness or lateness conditions which may cause the extra inventory or lateness penalty respectively. To remedy this problem, this study employs AutoMod simulation software to build three scenarios and evaluates their performance based on the number of fixture changes and the number of tardy orders.
The three main themes of this research consist of production flow improvement, multifunctional labors, and bottleneck machines. In the production flow improvement research, three models are considered: the first one represents the current status of the company''s production process. Six different dispatching rules are investigated here including EDD (Earliest Due Date), AT-RPT (Arrival Time-total Remaining Processing Time), SPT (Shortest Processing Time), MST (Minimum Slack Time), MDD (Modified Due Date), and CR (Critical Ratio).; the second model sets up a checkpoint to determine the priority of orders based on the urgency of the due date; the last model sends all orders in the same queue, but determines the processing sequence based on the weighted criterion of both due date and the fixture change. The second main theme of this study analyzes the utilization of the current production line, and then assigns the staffs in the low utilization workstation extra task – so called multi-functional labor while in the last theme the bottleneck drift of the last two themes is observed and analyzed.
The simulation results show that in the flow improvement study, model III – the weighted criterion provides better results in both number of fixture change and the number of tardy orders. In the theme of multi-functional labors, three major improvements are identified from the Pareto front and the percentage improvement on the number of fixture changes is between 42.86% and 53.57% while the improvement on the number of tardy orders may fall between 1.31% and 25.33%. As for the last theme, no bottleneck drifting is identified, but the WIP and the utilization of machines and staffs are improved through the models above and particularly in the multi-functional labor.
|
|---|
