Improving workforce scheduling using artificial neural networks model

This paper demonstrates a decision support tool for workforce planning and scheduling. The research conducted in this study is oriented on batch type production typical for smaller production systems, workshops and service systems. The derived model in the research is based on historical data from Public utility service billing company. Model uses Artificial Neural Networks (ANN) fitting techniques. A set of eight input indicators is designed and two variants were tested in the model with two different outputs. Several comprehensive parameter setting experiments were performed to improve prediction performances. Real case studies using historic data from public weather database and communal consolidated billing service show that it is difficult to predict the required number of servers-workers in front office. In a similar way, this model is adequate for complex production systems with unpredictable and volatile demand. Therefore, manufacturing systems which create short cycle products, typical for food processing industry, or production for inventory, may benefit of the research presented in this paper. ANN simulation model with its unique set of features and chosen set of training parameters illustrate that presented model may serve as a valuable decision support system in workforce scheduling for service and production systems.