Optimal allocation of goods in a warehouse: Minimizing the order picking costs under real-life constraints

We optimize the average order picking time in warehouses, that are used in the food industry. However, the logistic process itself is not subject to change. We may influence the allocation of goods to the individual storage spaces only and thus minimize the average traveled distance. Moreover, we have to consider all constraints that occur in a real-life setting. For instance, it is crucial to avoid placing heavy goods on top of fragile items. First, we provide a mixed integer programming formulation of the problem. We then proceed developing new constructive and improvement heuristics. The latter are based on a k-opt approach combined with a statistical test. Overall, we achieve a saving of up to 20% of the total traveled distance, and hence a significant cost reduction. This was confirmed by our customer, who applies our method in daily use.