MRP•NEXUS: Intelligent Demand Forecasting and Inventory Optimization Using Machine Learning

Abstract

The accurate demand forecasting and inventory optimization cannot be overstated when it comes to having efficient and smooth supply chain processes and reducing costs associated with inventory and risks connected with it. Inventory control systems that use traditional methods of operation depend on manual planning and separate forecasting process, which makes them ineffective in responding to dynamically changing demand patterns. In this paper, an intelligent decision support system is proposed by combining an ensemble machine learning algorithm with material requirements planning to enable forecasting and inventory optimization and detect conflicts in this process. The forecasted demand is then used to calculate Economic Order Quantity (EOQ), Safety Stock, and Reorder Point, whereas a conflict detection mechanism automatically detects important inventory states. The proposed framework has been developed as a web application with the help of Flask and SQLite, which supports forecasting, inventory tracking, and procurement planning in an interactive dashboard. The experimental analysis shows the accuracy of 91.3% with MAPE of 8.7% and MAE of 35.6, reflecting the efficiency of the framework for intelligent inventory management.