Search results

With increasing demands for competitiveness, demand fulfilment and cost efficiency, the need to optimise workforce planning in logistics has become crucial. This applies not only to external customer demands, but also to internal customers, i.e. production. For this reason, the purpose of this paper is to develop a simulative, data-driven model that predicts the internal shopfloor material logistics demands.

It is a hybrid approach that includes both deterministic and probabilistic components and is an alternative to advanced but data and knowledge-dependent machine learning algorithms. Inductive, self-developed procedures, heuristic calculation rules and consideration of real-world factors form the basis of the prediction of the number of picks. The number of picks predicted in the first step forms the basis for deriving the number of employees required in the second step, and thus the basis for optimised workforce planning. The developed approach was then validated in a case study in a real company.

The results show that the model significantly optimises not only the planning efficiency, but also the forecasting effectiveness through better decision making in demand prediction and workforce planning in internal shopfloor material logistics compared to the status quo on a weekly basis (95.5% accuracy in the case study). This improved decision making leads to increased efficiency throughout the intralogistics/production system.

A structured approach is described for systematically predicting the number of internal picks, which is highly relevant in practice and cannot be found in the existing literature (from the data model to the calculation rules, including statistical influencing factors, to the prediction). In terms of future research, the model has the potential to be used and validated in additional companies.