PERANCANGAN SISTEM PENJADWALAN LANTAI PRODUKSI PT ANUGRAH YADEN UTAMA MENGGUNAKAN GENETIC ALGORITHM

• Main Author: Khairandi, Farhan
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/50857
• Institution: Institut Teknologi Bandung

SMEs (small and medium-sized enterprises) specializing in medical equipment nowadays have been facing challenge, profoundly at the moment of COVID-19 pandemic that happened through the year of 2020. Pandemic makes the demand for medical equipment rising in many health facilities. The opportunity become a real challenge for many SMEs that still developing, because they have limited resources, marketing strategy, and quality compared to global key player for the similar industry. SMEs are challenged to evaluate their production system in order to be more responsive to demand fluctuation by maximizing productivity while keeping the quality of their products to fulfill the demand shifts. PT AYU as SMEs that produce hospital furniture has problem in projecting demand fulfillment. Since they have no scheduling system for their shopfloor level. The scheduling process terminates at job order level, which it should be further decomposed to operational level. Incomplete scheduling will makes the production schedule inaccurate, because they overlook the production capacity. Production system implemented at the shop floor is job shop and new job continually arrive at anytime (dynamic scheduling). In order to adjust the scheduling methodology with system environment, new approach has been developed. In this thesis, modified metaheuristic method for scheduling is proposed to formulate the optimal schedule. The proposed metaheuristics is further tested with job shop mathematical model which is modifed from Ozguven, et al. (2008) to measure the effectiveness of the metahuristic in producing near optimal solution. Metaheuristic method is used to catch computation efficiency over analytics approach, especially for complex problem size of NP Hard combinatorial problem such in job shop scheduling. Dynamic scheduling with genetic algorithm studied by Lin, et al., (1997) is modified in this thesis to optimize the shop floor scheduling in PT AYU with objective function of mean turnaround time. Then the performance of the designed scheduling methodology is tested with historical data retrieved from the shop floor. Based on the result, the proposed methodology successfully decrease the mean turnaround time by 27,2% which valued Rp 1,14 M of potential loss.