一項測試工作得以完成,涉及到人員、單位、產品、測試設備、流程等各項因素,本研究針對個案廠商的新型筆記型電腦的測試作業,探討其測試作業所引發的排程問題做一通盤完整的瞭解,然後建立一般化數學規劃模式,並運用最佳化軟體ILOG求解.。 從研究結果發現在數學規劃模式二中針對11個機種的測試作業,求出綜合測試排程的總完工時間為1122小時,相當於1.55個月,而個案廠商僅對單一個機種的預估排程通常為1-2個月的時間,本文研究的測試排程結果可提供該部門管理階層在安排測試作業時之參考。在模式三中,針對瓶頸的測試設備C,在多機種的測試情境下,最小化總測試時間,而最佳化結果決定了每個測試任務的使用設備,其中測試設備C1的產能使用率為89%;產能之所以沒有充分使用,其中人力的因素較具其影響性。另外,分析每一大類測試項目的重測作業造成整體作業延遲的影響度,求出不同測試項目重測下的最佳化排程。研究結果顯示,若是測試作業使用到設備C之下,該測試作業若是進行重測,其造成排程拉長及延遅的影響度大於其它測試項目。 本研究成果主要在於求出最小化的完工時間(Makespan),供個案廠商的管理階層參考,以期能有效利用測試部門的資源,提高設備利用率及人員工作效率,同時使得相關的測試人員、行政人員在作業進行中能夠滿意,並達到各級人員的共識且減少不必要的溝通協調。
A testing job can be done smoothly due to many factors such as operators’ ability, cooperation of working units, products, testing equipments and process. In this study , we aim at knowing and discussing the scheduling problems of the notebooks testing jobs and then build up the general mathematical programming models and solve the problem by optimal software ILOG. It is discovered in the study result that in the model 2 ,which is focused on the scheduling of 11 new NB models , the optimal result shows that the makespan of total 11 models is 1122 hours , which equals to 1.55 month and this result is shorter than the predicting makespan of the company’s singe model scheduling, which is about 1 to 2 months. We hope this study may provide reference for scheduling the testing jobs. In the model 3 , which is focused on the scheduling of the bottleneck testing equipment C, the result assigns the optimal equipment to each testing job and the utility rate of equipment C is 89% .That the utility rate does not reach 100% is due to the manpower. Besides, we analyze the influence of the re-testing of the each item. The result shows that if the testing job uses equipment C , then the influence which makes lateness is greater than others. In this study, the major result is to find out the minimum testing makespan. We expect to provide it as reference for the company''s management. We hope that the company can use the testing resource effectively and improve the utility rate of testing equipments and working efficiency. Meanwhile, we want to let the testing staff and the administrative personnel be satisfied with their jobs. And each person have a common consensus and reduce unnecessary communication and argument.