Abstract
Thin Film Transistor - Liquid Crystal Display (TFT-LCD) is one of Taiwan’s most important industries, and Color Filter (CF) is a key component of TFT-LCD. Chunghwa Picture Tubes (CPT) has several CF fabs, including Y1 (two lines of 4.5-th Generation) with capacity of 240K, Y2 (two lines of 6-th Generation), H1 (one line of 3.5-th Generation) with capacity of 65K, H2 (one line of 4.5-th Generation) with capacity 75K. The equipment of different CF fabs can support each other with certain constraints. Furthermore, the production cost at different CF fabs is different. Order assignment to fabs need to take into account each order’s quantity, type, specification, as well as each fab’s and its downstream fab’s capacity and capability. Currently, orders are assigned and released to CF fabs on a manual basis with less flexibility of dynamic support among CF fabs. This three-year project will develop Advanced Planning and Scheduling (APS) for CPT’s multiple CF fabs. The first year (from September 2009 to August 2010, in progress) develops order release and capacity planning policies for Y1 fab. An AutoMod simulation model has been built for performance evaluation and what-if analysis. Originally, the second year will develop order release and capacity planning policies for H1 and H2 fab, and an AutoMod simulation model will be built for performance evaluation and what-if analysis. However, as there will be a strategic change of H1 in July 2010 and Y2 is a newly built 6-th Generation plant ramped up in early 2010, this project will use Y2 to replace H1 in the second year project. The third year will develop APS (including integrated order assignment policy and order release/dispatching policy), considering the production cost and supporting mechanism among CF fabs. This project plans to develop APS for CPT color filter fabs. With APS, it is expected to reduce planning and scheduling time by 50% (from manual to automated operation), increase average equipment utilization by 5%, increase on-time delivery rate by 5%, reduce average WIP level by 5%, and reduce production cycle time by 5%.