Figure 4 - uploaded by Huanfa Chen
Content may be subject to copyright.
Source publication
In this paper, we address the problem of planning police patrol routes to regularly cover street segments of high crime density (hotspots) with limited police forces. A good patrolling strategy is required to minimise the average time lag between two consecutive visits to hotspots, as well as coordinating multiple patrollers and imparting unpredict...
Citations
... The most focused policing strategy will have a hard time delivering results if officer are just not available to patrol the intended places. Last, deploying routing and spatial algorithms can optimize the deployment of call response units across relevant police sectors Chen et al., 2015). As fast response remains an important performance indicator in policing, real-time tracking and analyses of police officer's geolocation can support dispatch agents to assign the unit that is most likely to deliver the fastest response time to an emergency call (Przeszlowski et al., 2022). ...
Police resources are scarce public goods and have to be used with maximum efficiency. One dimension of efficiency is the spatial and temporal alignment of police presence to calls-for-service. The research question of interest in this paper is whether the police are where and when they are needed most. We employ a supply and demand model based on police presence and the occurrence of calls for police service. We use GPS data from 100 tracked police vehicles and data on calls-for-service from a mid-sized European police agency. We find that most segments receive relatively more police presence than calls-for-service alone would warrant and that calls-for-service as well as emergency calls show little spatiotemporal variation. Further, police presence is found to be the main driver in changes regarding police provision across streets. We discuss the need to develop more evidence-based frameworks to investigate the effectiveness of police patrol and response deployment. The insights can assist police chiefs to identify streets with high demand for police services and to understand the predictability of police demand within their jurisdiction.
... Although the concept of representing the benefit of patrolling a specific area of the graph is interesting, this dataset is not available for general use. Chen et al. [2015] propose an approach for defining patrol routes using ant colony optimization technique. Although the authors also model the street as a graph, they consider just a subset of types of crimes (mobile phone and car theft) since the focus is on on-foot patrolling. ...
It is a well-known fact that criminality is an open, yet important, issue in most urban centers worldwide. Especially in Brazil, creating solutions to reduce crime rates is a top priority. To reduce crime rates, many cities are adopting predictive policing techniques. Predictive policing techniques are highly based on extracting valuable knowledge from a massive dataset that contains information about times, locations, and types of past crimes. The extracted knowledge is then used to provide insights to police departments to define where the police must maintain its presence. These datasets may also be used for a critical predictive policing task: defining where police patrols should patrol. Such patrols are commonly defined to cover a series of crime hot spots (areas that present high criminality levels) and have some restrictions to be considered (number of available police officers, cars, etc). Thus, defining the route for each police vehicle is a complex optimization problem, since in most cases, there are many hot spots and the existing resources are scarce, i.e., the amount of vehicles and police available is much smaller than necessary. Unfortunately, high-quality crime rates data are not easy to obtain. Aiming to tackle this problem, this article proposes the PolRoute-DS dataset, a dataset designed to foster the development and evaluation of police routing approaches in large urban centers. The PolRoute-DS combines the spatial structure of the city of interest (in the context of this article, the city of São Paulo) represented as a connected and directed graph of street segments with criminal data obtained from public sources. PolRoute-DS is available for public use under the Creative Commons By Attribution 4.0 International license (CSV and PostgreSQL versions) and can be downloaded at https://osf.io/mxrgu/.
... improve ← True; STATIC ← STATIC*; Continue 16: [Chen et al. 2015] propõem uma abordagem para definir rotas de patrulhamento usando a técnica de otimização de colônia de formigas. Embora também modelem as ruas da cidade como um grafo, eles consideram apenas um subconjunto de tipos de crimes (furto de celular e carro), já que o foco é o patrulhamento realizado a pé. ...
Neste artigo uma heurística híbrida, baseada na metaheurística Iterated Local Search, com busca local Variable Neighborhood Descent, é proposta para resolver, de maneira aproximada, um problema de otimização relacionado ao Roteamento de Viaturas Policiais em Grandes Centros Urbanos (RVP-Urb), onde o principal objetivo é diminuir o risco de áreas com alta taxa de criminalidade, reduzindo a violência nas cidades.
... In case of stochastic model of route design problem in Table 11, mostly minimization of idle time Rocha. 2012, 2013a, b;Chen et al.. 2015) and addition of unpredictability in patrol route design (Chen 2013;Lin et al. 2013;Chen et al. 2015) are considered. Few papers consider minimization of average response time (Birge and Pollock. ...
... In case of stochastic model of route design problem in Table 11, mostly minimization of idle time Rocha. 2012, 2013a, b;Chen et al.. 2015) and addition of unpredictability in patrol route design (Chen 2013;Lin et al. 2013;Chen et al. 2015) are considered. Few papers consider minimization of average response time (Birge and Pollock. ...
... In this category, there is a clear dominance of Bayesian decision models Rocha, 2012, 2013a, b;Chen et al., 2015), although we observe some applications of M/M/N queuing model (Larson and Mcknew, 1982), continuous time Markov process (Birge and Pollock, 1989), Markov decision process (Chen, 2013;Lin et al., 2013) and stochastic vehicle routing approach (Azimi and Bashiri, 2016). ...
Police patrol is an effective crime prevention tool and boosts public confidence in urban security. Many interesting decision making problems appear in route design, resource allocation and jurisdiction planning. Many cities across the world have adopted a structured and intelligent method of police patrol due to the presence of a variety of operational and resource constraints. In this paper, we present a comprehensive review of the state-of-the-art in this domain, especially from the practice of operations research (OR) point of view. This is the first-of-its-kind review on police patrol presenting a classification scheme based on the type of problem, objective and modelling approach. In this novel scheme, one can track any paper almost readily to find the specific contribution. The applicability of OR in this domain is set to grow significantly as the governments formulate policies related to smart city planning and urban security. This study reveals many practical challenges in police patrolling for future research.
... The most focused policing strategy will have a hard time delivering results if officer are just not available to patrol the intended places. Last, deploying routing and spatial algorithms can optimize the deployment of call response units across relevant police sectors Chen et al., 2015). As fast response remains an important performance indicator in policing, real-time tracking and analyses of police officer's geolocation can support dispatch agents to assign the unit that is most likely to deliver the fastest response time to an emergency call (Przeszlowski et al., 2022). ...
Purpose Police patrol has undergone an evidence-based and data driven transition in the beginning of the 21st century. While crime patterns are well researched, patterns of police presence are not. Despite the abundance of available GPS data, little is known about the spatiotemporal patterns of police forces. Given the paucity of evidence on where everyday policing takes place, we ask: what spatiotemporal patterns of police exist, how do these patterns change over time, and how do these correspond to local crime patterns? Methods Therefore, we analysed more than 77 million GPS signals from 130 police patrol cars and more than 50,000 recorded crimes from 2019.to investigate where and when police patrols are present. All data were geocoded and map matched using high performance computing.Results We found that police, much like crime, concentrates on a small proportion of street segments and that the spatial concentration experiences temporal instability at the micro level. Further, spatiotemporal police presence and its concentration appear to be unrelated to local levels of crime and crime concentration. Conclusions These findings inform police chiefs and researchers alike and enable alterations of patrol deployment in order to refine the spatiotemporal focus of police on local crime. Future considerations are required to research optimal spatiotemporal alignment of police presence to effectively prevent crime.
... The authors considered the spatial pattern of crime hotspots to suggest patrol routes and the effectiveness of collective patrol activities. In [7], a Baysian ant colony algorithm was proposed to minimize the average idle time between two consecutive visits of crime hotspots by police officers. It used probabilistic bayesian model with ant colony algorithm which made the patrol route selections less predictable by offenders. ...
A well-crafted police patrol route design is vital in providing community safety and security in the society. Previous works have largely focused on predicting crime events with historical crime data. The usage of large-scale mobility data collected from Location-Based Social Network, or check-ins, and Point of Interests (POI) data for designing an effective police patrol is largely understudied. Given that there are multiple police officers being on duty in a real-life situation, this makes the problem more complex to solve. In this paper, we formulate the dynamic crime patrol planning problem for multiple police officers using check-ins, crime, incident response data, and POI information. We propose a joint learning and non-random optimisation method for the representation of possible solutions where multiple police officers patrol the high crime risk areas simultaneously first rather than the low crime risk areas. Later, meta-heuristic Genetic Algorithm (GA) and Cuckoo Search (CS) are implemented to find the optimal routes. The performance of the proposed solution is verified and compared with several state-of-art methods using real-world datasets.
... Currently, there is a poor understanding of how routine day-to-day patrol intervenes with criminal opportunities. Moreover, police patrol routing strategies which incorporate the responsibilities of police officers on patrol are still missing [7]. To address these deficiencies, this paper examines which algorithms can regulate effective police patrol routing, with a focus on overt or visible crime, i.e., crimes occurring in public places, which can be seen and heard by other people and can draw police attention. ...
... The major difference occurs when the vehicles are idle, i.e., not responding to an emergency call. While police vehicles are assigned to a patrol beat and have to patrol streets [6][7][8], the emergency medical services (EMS) and the fire brigade are redeployed to a base station. The redeployment models can be static (the allocation is fixed and a vehicle is sent back to its home base whenever it becomes idle), or dynamic (at the moment of relocation, the state of the system is taken into account), but they neither have to be visible nor have a preventative task [9][10][11][12]. ...
... Not only the limitations affect the results in the Kansas City preventative patrol experiment, some articles even prove the opposite: random preventive patrol is one of the most important and time-consuming tasks employed on a daily basis by the police [90]. Moreover, high randomness can create a perceived omnipresence of the police and in developing routing strategies it is often mentioned as a prominent characteristic of preventative police patrol [7,8]. Therefore, an efficient and effective allocation of the available resources is crucial. ...
Police patrol is a complex process. While on patrol, police officers must balance many intersecting responsibilities. Most notably, police must proactively patrol and prevent offenders from committing crimes but must also reactively respond to real-time incidents. Efficient patrol strategies are crucial to manage scarce police resources and minimize emergency response times. The objective of this review paper is to discuss solution methods that can be used to solve the so-called police patrol routing problem (PPRP). The starting point of the review is the existing literature on the dynamic vehicle routing problem (DVRP). A keyword search resulted in 30 articles that focus on the DVRP with a link to police. Although the articles refer to policing, there is no specific focus on the PPRP; hence, there is a knowledge gap. A diversity of approaches is put forward ranging from more convenient solution methods such as a (hybrid) Genetic Algorithm (GA), linear programming and routing policies, to more complex Markov Decision Processes and Online Stochastic Combinatorial Optimization. Given the objectives, characteristics, advantages and limitations, the (hybrid) GA, routing policies and local search seem the most valuable solution methods for solving the PPRP.
... A study by Chen et al. (2015) looked at how ant colony algorithms can be used to plan patrol routes. Their study developed a patrolling strategy using Bayesian methods and ant colony algorithms. ...
Police forces are constantly competing to provide adequate service whilst faced with major funding cuts. The funding cuts result in limited resources hence methods of improving resource efficiency are vital to public safety. One area where improving the efficiency could drastically improve service is the planning of patrol routes for incident response officers. Current methods of patrolling lack direction and do not consider response demand. Police patrols have the potential to deter crime when directed to the right areas. Patrols also have the ability to position officers with access to high demand areas by pre-empting where response demand will arise. The algorithm developed in this work directs patrol routes in real-time by targeting high crime areas whilst maximising demand coverage. Methods used include kernel density estimation for hotspot identification and maximum coverage location problems for positioning. These methods result in more effective daily patrolling which reduces response times and accurately targets problem areas. Though applied in this instance to daily patrol operations, the methodology could help to reduce the need for disaster relief operations whilst also positioning proactively to allow quick response when disaster relief operations are required.
... Intuitively, the more important hotspots should have higher visiting frequency or lower average idleness. The flexibility can be measured using weighted global average idleness (WGAI) (Chen, Cheng and Wise, 2015), with (ℎ ) representing the weight of the hotspot ℎ and the larger value representing the higher priority: ...
... The agent-based framework is used to model two patrolling strategies, BAPS and a benchmark strategy Christofides Cyclic Patrolling Strategy (CCPS), which is a deterministic and cyclic patrolling strategy based on graph theory (Chen et al. 2015). The reason for using CCPS as the benchmark is that the real-world patrol strategy is confidential and difficult to obtain and that the family of cyclic strategies are classical algorithms for the patrolling problem and perform well in different situations (Chevaleyre 2004). ...
In urban areas, crime and disorder have been long-lasting problems that spoil the economic and emotional well-being of residents. A significant way to deter crime, and maintain public safety is through police patrolling. So far, the deployment of police forces in patrolling has relied mainly on expert knowledge, and is usually based on two-dimensional spatial units, giving insufficient consideration to the underlying urban structure and collaboration among patrol officers. This approach has led to impractical and inefficient police patrol strategies, as well as a workload imbalance among officers. Therefore, it is of essential importance to devise advanced police patrol strategies that incorporate urban structure, the collaboration of the patrol officers, and a workload balance. This study aims to develop police patrol strategies that would make intelligent use of the street network layout in urban areas. The street network is a key component in urban structure and is the domain in which crime and policing take place. By explicitly considering street network configurations in their operations, police forces are enabled to provide timely responses to emergency calls and essential coverage to crime hotspots. Although some models have considered street networks in patrolling to some extent, challenges remain. First, most existing methods for the design of police districts use two-dimensional units, such as grid cells, as basic units, but using streets as basic units would lead to districts that are more accessible and usable. Second, the routing problem in police patrolling has several unique characteristics, such as patrollers potentially starting from different stations, but most existing routing strategies have failed to consider these. Third, police patrolling strategies should be validated using real-world scenarios, whilst most existing strategies in the literature have only been tested in small hypothetical instances without realistic settings. In this thesis, a framework for developing police patrol strategies based on the urban street network is proposed, to effectively cover crime hotspots, as well as the rest of the territory. This framework consists of three strategies, including a districting model, a patrol routing strategy for repeated coverage, and a patrol routing strategy for infrequent coverage. Various relevant factors have been considered in the strategy design, including the underlying structure of the street network and the collaboration among patrollers belonging to different stations. Moreover, these strategies have been validated by the patrolling scenarios in London. The results demonstrate that these strategies outperform the current corresponding benchmark strategies, which indicates that they may have considerable potential in future police operations.
... This paper is a further development of, and substantial improvement on, a previous work (Chen, Cheng, & Wise, 2015). In addition to the broad background introduced above, the current paper is substantially improved in five aspects. ...
... In addition to the broad background introduced above, the current paper is substantially improved in five aspects. First, only three guidelines were discussed in Chen et al. (2015), namely, efficiency, flexibility, and unpredictability. ...
... Here two more guidelines-scalability and robustness-are developed, which measure the general applicability of the routing strategy in different situations including different team size, hotspot areas, and emergencies, as this has not be discussed in any previous literature. Furthermore, the guideline of unpredictability is further quantified here, which was only conceptually discussed in Chen et al. (2015). Second, the Bayesian Ant-based Patrol Strategy (BAPS) is further developed in accommodating these guidelines. ...
A cooperative routing strategy for daily operations is necessary to maintain the effects of hotspot policing and to reduce crime and disorder. Existing robot patrol routing strategies are not suitable, as they omit the peculiarities and challenges of daily police patrol including minimising the average time lag between two consecutive visits to hotspots, as well as coordinating multiple patrollers and imparting unpredictability to patrol routes. In this research, we propose a set of guidelines for patrol routing strategies to meet the challenges of police patrol. Following these guidelines, we develop an innovative heuristic-based and Bayesian-inspired real-time strategy for cooperative routing police patrols. Using two real-world cases and a benchmark patrol strategy, an online agent-based simulation has been implemented to testify the efficiency, flexibility, scalability, unpredictability, and robustness of the proposed strategy and the usability of the proposed guidelines.
