OPTIMASI PENJADWALAN PREVENTIVE MAINTENANCE PADA MESIN PRODUKSI OBH COMBI DI PT COMBIPHAR MENGGUNAKAN ALGORITMA GENETIKA

• Main Author: Thessalonika, Samantha
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/83692
• Institution: Institut Teknologi Bandung

PT Combiphar is a pharmaceutical company, and one of its flagship products is OBH Combi. The production process of OBH Combi involves six types of machines: unscramble machine, filling machine, capping machine, labeling machine, cartoning machine, and packing machine. To measure the performance of the OBH Combi production process, PT Combiphar uses the Overall Equipment Effectiveness (OEE). In 2023, the OBH Combi production line only achieved OEE of 36.347%, which has not yet reached the 41% target. One of the reasons the OEE target hasn’t been achieved is due to high equipment breakdown loss which caused by the current maintenance activities haven’t considered the actual condition of each machine. The high equipment breakdown also led maintenance costs escalating to IDR 580,471,735. Therefore, a maintenance scheduling is needed to minimize total maintenance costs while also considering the availability and reliability of the system. The maintenance scheduling are made by developing a genetic algorithm, starts by generating an initial population of 20, followed by selection, crossover, mutation, and elitism stages. The selection stage is carried out using the roulette wheel method. The crossover stage is carried out with single-point crossover with a crossover rate of 0.95. The mutation stage is carried out with bit flip mutation with a mutation rate of 0.01. These stages are performed iteratively according to the number of generations, which is 300. The proposed maintenance schedule can maintain the average machine availability and reliability at 93.2392% and 94.5204%, respectively. Total maintenance costs were reduced by 10.04% from the existing total costs. The obtained maintenance schedule successfully reduced unplanned downtime by 21.543% which reduced the total expected machine failures from 228 times to 203 times. The production runtime increased, which successfully raised the average OEE value from 36.347% to 37.865%.