Integer Linear Programming on Seasonal Demand Models to Optimize Production Capacity

Make-to-order companies often face challenges in meeting seasonal de-mand throughout the year. This causes financial losses. This research aims to optimize production capacity planning through standard production time analysis, demand forecasting based on previous year's demand data, and production capacity optimization using Integer Linear Programming and Branch and Bound with the help of LINGO and Machine Learning software. The research results show that the Decomposition method produces fore-casts with a MAPE value of 29.5%, predicting monthly demand for the next year. Observations at the three main work stations provide a standard time of 0.0818 minutes for Mixing, 0.2027 minutes for Filling, and 0.7413 minutes for Packing, with the production capacity per person per hour for each work station being 733, 296, and 81 units. Optimizing production ca-pacity produces two alternative solutions with the same total cost. These two alternatives are able to reduce additional working hours by 16% or around 971 hours, and reduce labor costs by 7% or IDR 35,231,250.