Akademik Veri Yönetim Sistemi
JOURNAL OF POLYTECHNIC-POLITEKNIK DERGISI, cilt.28, sa.5, 2025 (ESCI, TRDizin)
Businesses need to use their resources, such as vehicles, machinery, etc., efficiently and effectively not to interrupt their customers' services. Maintenance is critical for business resources to be always ready and operational. In addition, spare parts and consumables used in maintenance-repair processes have a significant share in operating expenses. For this reason, successful maintenance and inventory management in maintenance departments provides benefits to businesses in many dimensions. This study proposes a dynamic model integrating inventory management and predictive maintenance plans in a bus fleet. Machine learning and deep learning algorithms were used to predict the average mileage of breakdowns on the company's data. By using machine learning algorithms, high prediction success was achieved with 95% R2 value, and the inventory management study was carried out according to the maintenance plans obtained from the predictive maintenance study. With the proposed inventory management model, the amount of spare parts needed per year, safety stocks, reorder points and order quantities were determined. The study was tested for the ten most critical spare parts and achieved a 2.11% reduction in inventory costs.