Operational Research Methodology on Master Schedule Planning for Surface-Mount Technology Line

Lee, Chew Aun (2020) Operational Research Methodology on Master Schedule Planning for Surface-Mount Technology Line. Final Year Project (Bachelor), Tunku Abdul Rahman University College.

Text LeeChewAun Full Text.pdf Restricted to Registered users only Download (2MB)

Abstract

The title of the project is “Operational research methodology on master schedule planning for Surface-Mount Technology line” which is focused in develop a smart master planning schedule to help in solving the problem of scheduling that faced by Hotayi Electronic (M) Sdn. Bhd. Schedule planning is a crucial component for manufacturing, especially in manufacturing of Surface Mount Technology (SMT). It was needed to be planned well so that the efficiency of production can be higher. However, planning schedule is a very big challenge in SMT production because of vary types of constraints are needed to take in consideration. In this project, Genetic algorithm was applied to create the schedule by using Python programming language. Before creating the result, the input data must be analyzed in details so that when go into the algorithm, it can classify which constraint was needed to take as priority consideration. After Genetic Algorithm, the results were coming out as line utilization and number of tardy jobs. To do the verification, the inputs were modified from original which are random distribution and 100% workload by three different distribution and two different workloads. All these distributions are normal distribution, exponential distribution and uniform distribution. The workloads also had been testes with 80% and 120% workload to the original workload. In addition, no matter what kind of distribution and workload used, the pattern of result are same, only the overall utilization and number of tardy jobs will vary. After that, the result also compared with other researcher’s algorithm, shortest processing time and trial and error method. In short, the result of Genetic Algorithm is better than shortest processing time method but not good as trial and error method.