Abstract
In the Thin Film Transistor-Liquid Crystal Display (TFT-LCD) industry, the module process is labor intensive, as it is more difficult to apply automation to this process, compared with Array, Color filter, and Cell processes. In addition, module process is considered a multi-model assembly line, which means several models from a basic product family are manufactured simultaneously. Therefore, a module process with integrated arrangement and line-balancing can reduce labor requirement and increase production efficiency. This research considered several practical characteristics of the TFT-LCD module process, including multi-skilled operators and operator efficiency, to solve a multi-model assembly line balancing (MuMALB) problem. A new mathematical model was proposed to arrange labors, tasks, workstations, and machines. Furthermore, data from the TFT-LCD module factories were used to evaluate the performance of heuristic two-phase adaptive genetic algorithm (AGA) based on experimental design and response surface method. This study integrated theoretical research and practical application to develop a heuristic two-phase AGA for assembly line balancing problem for the TFT-LCD module process. Hence, the efficiency could be improved, and the cost could be reduced, which would ultimately improve the global competitiveness of TFT-LCD manufacturers of our country.