Abstract: “Sizing and redesign of new and existing supply chain nodes, like warehouses or logistic hubs are highly data-dependent tasks, incorporating lots of transaction information which are mostly available through ERP systems. Therefore, item-based forecasting and lot sizing models, approximations or both are used widely when planning new supply chain nodes. This paper introduces a new approach of data-aggregation based on probability functions of each item (e.g. stock keeping units (SKU)) incorporating compensating behavior and time-dependent aspects.”

Keywords: data mining, big data aggregation, supply chain analysis, material flow analysis, warehouse planning