Simulating the impact of kanbans in a multiple reentrant manufacturing system

• Main Author: Go, Joan Corazon C.
• Format: text
• Language: English
• Published: Animo Repository 2011
• Subjects: Just-in-time systems; Production engineering; Industrial Engineering
• Online Access: https://animorepository.dlsu.edu.ph/etd_masteral/5882; https://animorepository.dlsu.edu.ph/context/etd_masteral/article/12760/viewcontent/CDTG005028_P.pdf
• Institution: De La Salle University

A reentrant manufacturing system (RMS) is a system where this sequence of operations is repeated several times before the jobs are completed. This made the system difficult to manage as jobs entering a particular resource could be at different job states. Hence, in managing the WIP in the system, the decision is not only to determine which among the jobs queued to feed next as consideration of job states is also important. When these jobs are not prioritized appropriately, the system produces large amount of work-in-process (WIP) which leads to longer cycle time and poor throughput performance. This study aimed to explore the use of pull system through kanbans in managing of WIP and throughput of reentrant manufacturing systems and compared typical scheduling and control mechanisms that are associated to push and push-pull strategies. This study deals with a reentrant manufacturing system for a single class job which considers stochastic processing times. A simulation model was developed to examine the impact of kanban in the throughput and WIP performance of RMS. Comparing Kanban system to the traditional push and the push-pull strategies, the proposed system was able to show improvement in both throughput and WIP performance of an RMS. In determining the impact of the changes in kanban size to total WIP and throughput, this study has shown that generally increasing the number of kanbans does not necessarily translate to a better throughput. It is also important to consider the job state where the kanban should be allocated. This study highlights the importance of assigning bigger kanban size on the latter job state of the system and smaller at the newer state.