Application of LSTM model for forecasting production orders

Abstract

Production planning is critical in modern industry, especially in custom machine manufacturing, where efficiency and meeting deadlines are essential. Time series analysis has become pivotal in optimizing production systems in recent years. This study presents the application of LSTM neural networks for production scheduling predictions, modeling temporal patterns and seasonal fluctuations based on historical data. The goal is to enable more efficient planning and accurate delivery times, thereby improving overall production performance. Preliminary results suggest that LSTM can outperform traditional statistical models, such as linear regression. It is crucial to tailor the model to the company's specific needs and relevant data.