Abstract
<jats:p>Order picking in shared storage. Simulations with the use of statistical methods Order picking is the most time- and work-consuming process in a warehouse. Therefore, rationalisation of the process of order picking is crucial in ensuring the effectiveness of warehouse operations. Although many companies utilise modern, automated warehouse systems, still majority of companies (about 60%) use the classical, manual picker-to-parts systems. There are two main storage systems that can be applied in the warehouse: dedicated storage, meaning that every item has its dedicated location in the warehouse and shared storage, where every item can be stored in many, sometimes very distant locations. Most companies use the shared storage systems, as they allow much better utilisation of storage space. However, its main drawback is that when picking items from such storage, we need to select locations, from which items need to be picked. So far, this problem has only been addressed to a very limited extent. With the use of this monograph, we try to fill this gap. The main aim of the monograph is to develop a methodology for selection of locations in shared storage using multivariate statistical analysis methods. It has been realised by means of describing every location be the set of variables and by means of selected linear ordering methods the most attractive locations (from the point of view of specific order) were selected. For these locations the picker&rsquo;s route and order picking times were designated. The system was tested for various storage orders (random and class-based), various strategies of selection of locations and selected linear ordering methods (COPRAS, TOPSIS, TMAL(+) and TMAL(&ndash;)). For given storage order and given strategy of selection of locations, the most effective methods were selected. The empirical part of the analysis was conducted by means of the simulation methods. The presented approach was also applied in the case study, where the untypical layout of the warehouse was presented. Obtained results were extensively analysed by means of the descriptive statistics methods and by using the statistical inference. For such layout a new routing heuristic was developed. Conducted research has also many limitations and indicated many new directions for future analyses.</jats:p>