An Integrated Lean and Machine Learning Approach to Inventory Management in Automotive Accessories SMES

In the Peruvian automotive accessories sector, low inventory turnover is a critical problem that hinders operational efficiency, increases storage costs, and reduces liquidity. This study aims to design and validate an integrated inventory management model to improve stock turnover in an automotive accessories warehouse. The model combines ABC classification, Economic Order Quantity (EOQ), slotting, 5S methodology, and machine learning-based demand forecasting. The methodology includes a systematic literature review, problem diagnosis, tool integration, simulation in Arena software, and pilot implementation. Key results revealed a 33.43% improvement in inventory turnover, a 30% reduction in average inventory, a 60% decrease in picking time, and a 19.95% reduction in total logistics costs. Additionally, the study achieved an 18.5% improvement in Inventory Record Accuracy (IRA) and a 22.69% increase in Location Record Accuracy (LRA). These findings highlight the effectiveness of integrating lean and technological tools in enhancing inventory management in automotive accessories warehouses. The study addresses a research gap in Latin America and offers a replicable model that contributes to improving operational efficiency and sustainability in similar industrial contexts.