Location Optimization of Emergency Bell Based on Coverage Analysis for Crime Prevention

Typically, emergency bells are security facilities that, when activated, trigger an alarm and immediately dispatch a police car to prevent crime. However, there currently exists an ambiguity in the criteria for emergency bell installation. Consequently, this study aims to find an optimal location for emergency bells whilst considering several factors like cumulative crime incidents. In particular, we exploited emergency bell location data, data on five major crimes, and the geographic information of administrative dongs (primary division of districts) in this study. Specifically, we performed correlation analysis, principal component analysis, and K-means clustering for exploratory data analysis. To effectively cover all 17,437 crimes, which are not covered by the existing emergency bells in Gwangju metropolitan city from 2018 to 2021, the results from the implementation of the emergency bell location set-covering problem revealed the need for about 6228 emergency bells. More precisely, the emergency bell maximal covering location problem was employed to derive the coverage percentage for 250, 500, 800, 1000, and 1500 emergency bells. The results showed that 2850 emergency bells were required to cover over 80% of crime occurrence coordinates, saving over half of the budget compared with covering them all. Overall, this study is noteworthy in its potential role as a roadmap for the optimal placement of emergency bells for future crime prevention.