Penerapan Algoritma Logika Fuzzy Mamdani Untuk Optimalisasi Stok Dari Berbagai Jenis Spareparts Handphone

Abstract

Effective inventory management is crucial for businesses in the mobile phone spare parts industry, where accurate stock levels directly impact operational efficiency and customer satisfaction. Traditional stock management methods are often inadequate in dealing with the uncertainty and complexity of demand forecasting. In the mobile phone repair and distribution industry, demand for spare parts varies significantly depending on factors such as new model launches, device failure rates, and fluctuating market trends. Conventional stock management systems often fail to handle this variability, leading to issues like stock-outs, which harm customer satisfaction, or overstocking, which increases storage costs and the risk of obsolescence. This research addresses these challenges by applying the Fuzzy Mamdani method to optimize spare parts inventory. The study focuses on transforming crisp stock data into fuzzy inputs to improve prediction accuracy. Data from UD.WSP for the year 2023 is fuzzified by incorporating variables such as initial stock, sales volume, and additional stock, with the output variable being the final stock level. The spare parts analyzed include LCDs, batteries, chargers, screen protectors, speakers, RAM, flex cables, rear cameras, front cameras, and SIM trays. The application of the Fuzzy Mamdani method resulted in a Mean Absolute Percentage Error (MAPE) of 11%, indicating a high level of prediction accuracy. These findings demonstrate that the Fuzzy Mamdani method is a viable solution for optimizing spare parts inventory management, offering an advanced approach to managing inventory amidst uncertainty.