A Study on the Development of a Flexible Operator Assignment System from Active Pharmaceutical Ingredients Manufacturing

• Main Authors: Hsing-KuanWang, 王行寬
• Other Authors: Taho Yang
• Format: Others
• Language: zh-TW
• Published: 2019
• Online Access: http://ndltd.ncl.edu.tw/handle/hp883u

碩士 === 國立成功大學 === 製造資訊與系統研究所 === 107 === The three major areas of the biotechnology industry are pharmaceutical, medical equipment and applied biotechnology industries. The export value of pharmaceutical industry has grown by 5.5% compared with last year, and API(Active Pharmaceutical Ingredients Manufacturing) plays a pivotal role in the pharmaceutical industry. Researcher selects a pharmaceutical company located in the Tainan Science Park in Taiwan for research. The case company are facing problems such as unsmooth flow of production operations and the lack of flexibility in the manufacturing system. The process of pharmaceutical industry is relatively fixed, so the operator arrangement will be the key element. The first part of this paper is to eliminate the waste with lean tools. First, we clarify the current problems by a site tour and solve the problem by using the lean thinking method, including reducing the columns of BPR, implementing the automation module, and modularizing the tasks of each floor. The waste in the factory was improved including cycle time, manpower and resources, the use of operator decreased from 5 people per shift to 4 people per shift; and the process cycle time of each batch decreased from 3.4 days to 2.2 days; The second part is builded with the foundation of previous improvement. We discuss the characteristics of the pharmaceutical industry, the way of operator dispatching and the feature of multi-floor building, shows the particularity of this flexible system. There are three factors are set including floor arrangement, the number of manpower and staff assignment,. We use a simulation software with full factorial experiment design to analyze the current situation and find out the optimal solution for each experiment. Although the lead time of final choice is not the shortest, it is optimal from the perspective of cost, and the production lead time is shorten to about 21%, and the use of operator is reduced by 1. We provide an efficient way to reduce the frequency of floor transitions. This method has the potential to provide an important reference for continuous process operator arrangements.