How Do Humans and Computers Perform Differently in Site Selection? Insights from an Ambulance Station Placement Game.

This study explores the integration of human decision-making and computational optimization in emergency response system deployment (ERSD), using a novel interactive game-based approach. Players of the interactive game place configurations of ambulance stations (AS) on a map, attempting to maximize ambulance coverage within the area of the map. We analyze these configurations, contrasting them with an optimized computer-generated configuration. Our analysis includes a heatmap to visualize the spatial distribution of the configurations, Hausdorff distance calculations for configuration similarity, Uniform Manifold Approximation and Projection (UMAP) for dimension reduction, and hierarchical agglomerative clustering. Additionally, medoid analysis identifies representative AS placements within clusters. Results yield intriguing differences between human intuition and the computational outcome, revealing potential implications for enhancing ERSD through hybrid models incorporating knowledge-based and data-driven approaches.