An Integrated ANN-Aggregate Planning-RCCP Framework for Make-to-Stock Bag Production: A Case Study of Keyna Products

Abstract

Company XYZ applies a make-to-stock (MTS) system for bag production, where inaccurate demand estimation can cause overstock, stockouts, and infeasible shop floor plans. This study proposes an integrated planning framework for the Keyna product family by combining demand forecasting using an Artificial Neural Network (ANN), aggregate production planning, master production scheduling, and Rough-Cut Capacity Planning (RCCP) using the Bill of Labour Approach (BOLA). Historical weekly demand data from 7 January 2023 to 24 February 2024 were used to develop the ANN model and generate six-week forecasts of 2601, 4635, 1542, 5773, 2394, and 5097 units. Forecasts were allocated into Keyna Mini Satchel (80%) and Keyna Round Satchel (20%), followed by aggregate planning and disaggregation using the Britan and Hax method to produce the Master Production Schedule (MPS). Capacity feasibility was evaluated across ten work centers using RCCP-BOLA. The level strategy produced a lower total cost (IDR 1,109,575,000) than the mixed strategy (IDR 1,975,350,000). RCCP results indicate recurring capacity shortages at work center 5 (painting), work center 6 (drying), and work center 7 (eyelets), for which targeted improvement actions were proposed.