An Integrated BWM–GIS–DEA Approach for the Site Selection of Pallet Pooling Service Centers

The scientific site selection for pallet pooling systems is pivotal to enhancing logistics efficiency and environmental performance. However, previous studies mainly adopt single-objective optimization approaches, which fail to simultaneously account for economic, environmental, and operational performance factors. The contribution of this paper lies in proposing an integrated decision-making method based on BWM-GIS-DEA to address the site selection problem for pallet pooling service centers. First, the Best-Worst Method (BWM) determines the weights of 13 criteria across 5 dimensions: economic, transportation, geographical location, technological, and service coverage. These criteria include factors such as the distribution density of pallet manufacturers and potential customers. Then, suitability maps are generated using Geographic Information System (GIS) spatial overlay technology to identify 6 alternative cities. Finally, a two-layer Data Envelopment Analysis (DEA) model is applied to measure the efficiency of the alternative sites. This method is applied in Inner Mongolia, China, and Ejin Horo Banner is identified as the optimal site with an efficiency score of 1.156, demonstrating superior resource allocation characterized by lower land costs and higher pallet turnover rates. The proposed framework not only fills a methodological gap in sustainable facility location research but also provides a replicable and policy-ready tool to guide practical decision-making.