A Data-driven Approach for the Production and Flow Shop Planning Model

Accurately predicting demand is foundational for efficient scheduling and sequencing in production workflows, and integrating machine learning algorithms plays a crucial role in achieving this goal. Effective demand forecasting optimizes resource allocation and enhances overall operational efficiency, showcasing its significance in modern production planning paradigms. Therefore, this article investigates a classic production scheduling and sequencing problem. To solve the developed model, demand parameters were predicted using a machine learning method rather than relying on simple and error-prone historical data. Hence, the support vector machine algorithm was employed for demand prediction across various periods. Subsequently, the estimated demand values were incorporated as inputs into the mixed-integer nonlinear programming (MINLP) model, which was then solved. Eventually, sensitivity analyses are carried out on the model to assure efficiency and present the improvements in production system costs, reduced waiting times, minimized machine downtimes, and the assured performance with this approach.