Optimising fugitive interception

Abstract

In the Netherlands, fewer than half of violent crimes lead to convictions. A key method for increasing convictions is red-handed arrests, which can be enhanced by developing a decision-support system for optimally positioning police units. Existing models often assume an empty city, using maximum road speeds for both fugitives and police, which ignores real-world traffic conditions.
This thesis aims to improve the realism of these models by incorporating traffic effects, such as congestion and delays from traffic lights, into the optimisation of police strategies. The research examines how these traffic conditions impact the escape and interception processes, with the goal of increasing the accuracy of interception strategies and reducing the number of unpunished violent crimes.
The research question addressed is: 'What is the added value of considering realistic traffic conditions in the optimal positioning of police units for fugitive interception?' To answer this, a discrete event simulation model was developed using insights from a literature review and interviews. Simulation modelling was chosen to test complex scenarios without the biases and costs of real-world experiments.
The literature review identified that traffic includes static, semi-static, and dynamic components, such as traffic lights, open bridges, and congestion. It also revealed factors affecting criminal behaviour and route choices under stress, supplemented by interviews with stressed parcel delivery drivers. This knowledge, combined with understanding police behaviour during interceptions, informed the development of the simulation model.
The results showed that incorporating traffic conditions into the simulation model increases the probability of interception. Specifically, when fugitives face delays before reaching highways, the likelihood of interception improves because police units, moving faster due to their priority status, benefit more from traffic delays.
Optimisation of police positions was tested with and without traffic delays, and it was found that positions optimised with realistic traffic conditions were more robust. This was particularly evident in city centre scenarios but not in port dock areas, likely due to the high interception rates in the docks which made traffic impacts harder to assess.
In conclusion, models excluding traffic conditions are less effective in intercepting escape routes compared to those incorporating traffic. Therefore, integrating traffic into optimisation models is crucial for maximising interception probabilities. To mitigate the negative effects of omitting traffic, deploying additional police units is recommended.
The study also found that the impact of traffic on interception timing is more significant than accounting for the suspect's mental state. Traffic affects the timing of interceptions: accurate traffic estimates provide more time for police, while incorrect ones reduce it, affecting interception success. Future research should explore the effects of dynamic traffic conditions, such as varying green times and more differentiated fugitive and police behaviours, to optimise model accuracy while managing computational demands. Recommendations for police include integrating realistic traffic conditions and prioritising traffic lights over congestion, as well as experimenting with different escape speeds to maintain effectiveness in varying scenarios.