Three-Dimensional Agent-Based Model and Simulation of a Burglar's Target Selection

Residential burglary has been one of the most active research topics in the study of crime. It is also one of the most common crime types and sometimes has a long-lasting impact on victims not only financially but also psychologically. In order to understand residential burglary, many methods have been used such as analyses of crime data and national crime survey data, interviews with incarcerated burglars (and active ones), and developing models and simulations. Recently agent-based modeling and simulation (ABMS) has widely become utilized to understand burglars' behaviours at an individual level. However, many studies have developed and simulated their burglar models and environments in two dimensions (2D). This paper presents realistic, three-dimensional (3D) models and simulations of a residential burglar and its virtual environment. By incorporating 3D aspects of the models and simulations, accurate micro-interactions between the burglar and its environment can be simulated. The presented burglar model shows its target selection process based on wealth and layout cues. The benefits and limitations of the 3D ABMS approach are discussed. A few case studies show the validity of this approach and the possibility of its predictive and preventive use.