A Dynamic Stochastic Approach for Predicting the Production Capacity of an Industrial System

Time loss in the manufacturing process of a production line reduces productivity and harms the system’s brand image and long-term performance. The present research focuses on the use of scientific analysis and diagnostic techniques to correct current errors and to anticipate the future behaviour of a system. This paper proposes a stochastic approach, FORCAST-FBM, which is based on a hybridization of three methods: Failure Mode and Effects Analysis (FMEA), Bayesian Networks, and Monte Carlo Simulation. Our objective is to address a crucial issue in industrial production systems—namely, the forecasting of the quantity to be produced within a probable time frame during an upcoming production period. This approach plays a key role in production planning and management development. The proposed solution applies to any system in which production follows a chronological sequence across parallel production facilities, where components move along an automated path with no backward flow.