USULAN METODE PENJADWALAN JOB SHOP DI PT X MENGGUNAKAN ALGORITMA MATRIX HEURISTIC DENGAN MENGAKOMODASI PENJADWALAN ULANG

• Main Author: Christian Julian Tanaka, Angel
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/84138
• Institution: Institut Teknologi Bandung

PT X is a company that produces specialized machines for the automotive industry using a Make-To-Order (MTO) production system. Based on primary data obtained from the company's ERP, it was found that 37.68% of operations experienced delays. Using a fishbone diagram, the researcher determined that the focus of the problem to be solved from this research was the absence of a scheduling system that covers the assembly process and meets rescheduling needs. The job shop scheduling case in PT X is an NP-hard problem that needs to be solved using a heuristic algorithm. In this research, researchers will use a heuristic algorithm to develop a proposed scheduling method that includes the assembly stage and rescheduling requirements. After considering several alternative heuristic algorithms for the job shop case, this research proposes using the matrix heuristic algorithm as the method used for scheduling at PT X. This algorithm can be used for job shop scheduling and can process large amounts of data. Although this algorithm does not yet cover the assembly stage and meets rescheduling needs, it is flexible and can be developed to meet both the assembly and rescheduling stage needs. Researchers also propose using relative remaining processing time (RRP) as a priority dispatching rule and the earliest due date (EDD) as a tie-breaking rule for the matrix heuristic algorithm. Based on the results, the developed algorithm can produce a better solution for scheduling the production process of 5 products with a makespan reduction of 1292,2 hours or 12,37%. The reduction in makespan occurs because the RRP rule is used as a priority dispatching rule considering pacing and structural dependence factors. Apart from that, this algorithm can also accommodate rescheduling needs.