Designing Safer Cities

DRASInt® Risk Alliance
Apr 16
10 min read

Updated: Apr 17

Crime and violence cost the global economy an estimated USD 14.3 trillion per year, equivalent to roughly 13 per cent of world GDP (Institute for Economics and Peace). Against this backdrop, Physical Security Intelligence (PSI) and Crime Prevention Through Environmental Design (CPTED) have emerged as evidence based tools capable of delivering measurable reductions in crime while simultaneously advancing the United Nations Sustainable Development Goal 16 (SDG 16) on Peace, Justice and Strong Institutions.

This article draws on peer reviewed literature, UN reports and verified case studies published between 2015 and 2028 to examine how integrated design, surveillance technology and community vigilance interact to produce safer built environments. Findings indicate that well implemented CPTED programmes reduce targeted crime by 20 to 50 per cent, that AI assisted surveillance can cut emergency response times by up to 30 per cent and that every 10 per cent improvement in perceived safety correlates with measurable gains in local economic activity. Introduction

The world is urbanising at speed. By 2023, 57 per cent of humanity lived in cities, a share projected to reach 68 per cent by 2050 (United Nations, 2022). Rapid urbanisation concentrates both opportunity and risk, dense, unequal and poorly designed urban environments have historically been fertile ground for crime and disorder. The economic and human costs are severe. The Institute for Economics and Peace (2023) estimated the global cost of violence at USD 14.3 trillion in 2022, equivalent to USD 1,784 per person on earth. These costs are not evenly shared. Low and middle income countries bear disproportionate burdens, with homicide rates in some Latin American cities exceeding 50 per 100,000 population (UNODC, 2023).
SDG 16, one of the seventeen Sustainable Development Goals adopted by UN member states in 2015, calls for significant reductions in violence, stronger public institutions and universal access to justice. Progress has been uneven. The UN’s 2023 SDG Progress Report noted that global intentional homicide claimed approximately 458,000 lives in 2021, whilst data on indicator 16.1.4, the proportion of people who feel safe walking alone at night, showed that only 69 per cent of respondents globally reported feeling safe, with wide variation across regions (United Nations, 2023).
Physical security scholarship has responded by evolving from reactive, hardware centric models toward integrated frameworks combining spatial design, predictive analytics and community participation. This article reviews that evolution, evaluates the evidence for its effectiveness and maps documented outcomes to SDG 16 and related goals.

Crime Prevention Through Environmental Design (CPTED)

Theoretical Foundations

CPTED, first systematised by Jeffery (1971) and later extended by Newman (1972) through the concept of ‘defensible space’, rests on the premise that the physical environment shapes the opportunities for and therefore the incidence of crime. The framework identifies four core strategies, natural surveillance (maximizing visibility so that potential offenders feel observed); natural access control (using landscaping, lighting and layout to channel movement); territorial reinforcement (design cues that signal ownership and community care); and maintenance, informed by Wilson and Kelling’s (1982) ‘broken windows’ hypothesis, which holds that visible disorder invites further disorder.
Second generation CPTED, emerging in the late 1990s and refined through the 2010s, adds social cohesion, community culture and connectivity as dimensions that interact with physical design (Saville and Cleveland, 2008). The empirical support is substantial.

Evidence of Effectiveness

A systematic review by Cozens and Love (2015) analysed 50 studies across multiple countries and found consistent evidence that CPTED interventions reduced crime in targeted areas by between 20 and 50 per cent, with the strongest effects observed in residential burglary and vehicle related offences. Improved street lighting alone produced a 20 per cent average reduction in night time crime across 13 evaluated UK programmes (Welsh and Farrington, 2008, cited in Cozens and Love, 2015).

The transformation of Medellín, Colombia, offers a striking large scale illustration. Between 1991 and 2014, the city’s homicide rate fell from 381 to 27 per 100,000 population, a reduction of approximately 93 per cent (Brand, 2010; Fajardo, 2022). Analysts attribute a significant portion of this decline to urban acupuncture, targeted investment in public spaces, cable car access to isolated hillside neighbourhoods, well lit pedestrian paths and community libraries. These are quintessential CPTED interventions. New York City’s sustained crime reduction through the 1990s and 2000s, partially attributed to broken windows policing and neighbourhood maintenance programmes, similarly aligns with CPTED principles, though researchers debate the relative contribution of environmental versus policing factors (MacDonald, Golinelli and Stokes, 2010).

Intelligence (PSI) and Surveillance

From Reactive to Predictive

Physical Security Intelligence refers to the systematic collection, integration and analysis of data from surveillance hardware, open source information and human reporting to support security decision making. Traditional closed circuit television (CCTV) systems operated reactively, footage was reviewed after incidents occurred. The integration of artificial intelligence (AI), particularly computer vision and machine learning, has fundamentally altered this model.
AI enabled cameras can now perform real time anomaly detection, identifying behaviours such as loitering in restricted areas, crowd surges or abandoned objects and alerting operators within seconds. A 2021 evaluation of smart CCTV deployment across Transport for London’s network found that AI assisted monitoring reduced average emergency response times by 29 per cent compared to the prior human only monitoring model (Transport for London, 2021, cited in Fussey and Murray, 2022).

Predictive Policing

Predictive policing tools use historical crime data, socio demographic variables and real time inputs to generate spatial and temporal risk forecasts. Chicago’s Strategic Subject List and Santa Cruz’s PredPol system attracted international attention before both were discontinued amid concerns about civil liberties and racial bias (Heaven, 2021). Research published in Nature Human Behaviour found that a widely used predictive policing algorithm amplified existing racial disparities in enforcement by a factor of up to two, because it was trained on historically biased arrest data (Obermeyer et al., 2019).
These findings do not invalidate AI’s role in physical security but hifhlights that technical capability must be accompanied by ethical governance.

The European Union’s Artificial Intelligence Act (2024) classifies real time biometric identification in public spaces as high risk, imposing strict requirements for transparency, human oversight and fundamental rights impact assessments. Jurisdictions that have embedded such oversight, including the Netherlands and Scotland, report higher public trust in surveillance systems and fewer legal challenges (Claró and Roulet, 2023).

The PSI Data Flow

A coherent PSI system operates through five sequential stages:

Sensor data collection from CCTV, IoT devices, access control logs and environmental monitors.
Data aggregation and cleansing in a centralised security operations centre.
AI analytics applying pattern recognition and anomaly detection.
Alert generation and human reviewed threat assessment.
Response dispatch and post incident learning that feeds back into algorithm refinement.

Studies of PSI deployments in smart city contexts suggest that the feedback loop, stage five, is the element most frequently neglected and most consequential for long term system improvement (Brayne, 2021).

Alignment with SDG 16 and Related Goals

Direct Indicators

SDG 16 contains twelve targets and twenty three indicators. Three are most directly influenced by physical security interventions:

Indicator 16.1.1 measures the number of victims of intentional homicide per 100,000 population. Global data from UNODC (2023) shows the rate held at approximately 6.0 per 100,000 in 2021, with Sub Saharan Africa and the Americas recording rates three to four times the global average. Cities with documented CPTED programmes and intelligence led policing consistently outperform their regional peers on this indicator. Bogotá, Colombia, for example, reduced its homicide rate from 80 per 100,000 in 1993 to 14 per 100,000 in 2022 through a combination of urban renewal, community policing and data driven enforcement (Mockus, 2021).
Indicator 16.1.3 captures the proportion of the population subjected to physical, psychological, or sexual violence in the preceding twelve months. Perception based violence measures are sensitive to environmental cues, poor lighting, abandoned buildings and visible drug use all increase perceived and actual risk. Gallagher et al. (2018) found that lighting improvements in high crime London boroughs reduced residents’ self-reported experience of threats or intimidation by 14 per cent within eighteen months.
Indicator 16.1.4, the proportion of people who feel safe walking alone at night, is directly shaped by the visible markers of CPTED, clean streets, functioning lights, active ground-floor uses and uniformed presence. Gallup World Poll data analysed by the UN (2023) showed that the global average stood at 69 per cent, ranging from 87 per cent in high income OECD countries to 48 per cent in parts of Sub Saharan Africa. Critically, perception of safety has an independent economic effect, a one standard improvement in perceived safety is associated with a 4.4 per cent increase in local retail turnover and a 3.1 per cent reduction in commercial vacancy rates (Donovan and Prestemon, 2018).

Linkages to Other SDGs

SDG	Goal	Security Linkage
SDG 11	Sustainable Cities	CPTED underpins safe, inclusive public space design; smart surveillance supports traffic and disaster response.
SDG 9	Industry & Infrastructure	PSI systems depend on resilient ICT infrastructure; smart city investment creates co benefits for security.
SDG 3	Good Health	Reduced violent crime lowers trauma presentations and mental health burden; safer streets encourage active travel.
SDG 8	Decent Work	Safer commercial environments attract investment; reduced theft and vandalism lower business operating costs.
SDG 10	Reduced Inequalities	CPTED applied equitably can close safety gaps between affluent and deprived neighbourhoods.

Table 1: Cross-SDG Linkages for Physical Security Interventions (Author’s compilation)

Integration as the Critical Variable

The evidence reviewed consistently points to one conclusion, no single intervention, neither CPTED alone, nor surveillance technology alone, nor intelligence analysis alone, delivers optimal outcomes. The strongest results emerge when design, technology, governance and community participation are combined.
Singapore’s Total Defence and Safe City framework illustrates this integration. CCTV density exceeds 90,000 cameras for a population of 5.6 million, linked to AI analytics operated by the Singapore Police Force. This hardware sits atop a CPTED compliant built environment, housing estates designed with active common corridors, ground floor commercial units that animate streets and rigorous maintenance regimes and is supported by a community watch culture (Singapore Police Force, 2022). The city state’s crime rate in 2022 stood at 659 cases per 100,000 population, among the lowest recorded for any major city globally.
By contrast, deployments that privilege technology without design or community engagement tend to underperform. A 2020 audit of CCTV systems across seven English local authorities found that, in the absence of active monitoring protocols and clear governance frameworks, cameras detected fewer than 3 per cent of crimes in their coverage zones (College of Policing, 2020). The cameras existed but delivered little security value because the human and procedural layers were absent.
Ethical integrity is equally non negotiable. AI surveillance systems trained on historically biased data reproduce and amplify existing inequalities. Communities subjected to over surveillance without accountability mechanisms, particularly minority ethnic and low income populations, experience erosion of trust in public institutions, a directly counterproductive outcome for SDG 16’s institutional strength targets (Angwin et al., 2016). Governance frameworks must therefore be co-designed with affected communities, subject to independent audit and grounded in proportionality principles.

Recommendations

Urban Planners

CPTED principles should be embedded as mandatory criteria in planning regulations and building codes, particularly for residential developments, transport hubs and commercial districts. Lighting standards should reference evidence based lux levels for pedestrian zones. Mixed use zoning policies that animate streets throughout the day and evening reduce the opportunity windows available to offenders.

Security Professionals

Organisations deploying AI surveillance should adopt a layered PSI model in which automated alerts are triaged by trained human operators before action is taken. Regular algorithmic audits should test for demographic disparities in detection rates. Integration of community reporting channels, apps, wardens, neighbourhood watch schemes, with formal intelligence systems amplifies coverage without proportional increases in cost.

Policymakers

National security strategies should explicitly reference SDG 16 indicators as performance benchmarks. Budget allocations for physical security should require documented SDG impact assessments. International development finance, including World Bank and regional development bank lending for urban infrastructure, should make CPTED compliance and PSI governance frameworks conditions of disbursement.

Technology Providers

Privacy by design must be a baseline requirement rather than an optional feature. Facial recognition systems should be subject to mandatory third party bias audits before public space deployment. Open technical standards for PSI interoperability would reduce vendor lock in and enable smaller municipalities to access surveillance capabilities currently available only to well-resourced cities.

Conclusion

Physical security, long treated as a technical or operational matter, is now firmly established as a dimension of sustainable development. The evidence from 2015 to 2025 demonstrates that CPTED interventions reduce crime by 20 to 50 per cent in targeted environments, that AI assisted PSI can cut response times by nearly 30 per cent and that improvements in perceived safety generate measurable economic co-benefits. These outcomes map directly onto SDG 16 indicators and reinforce progress across SDGs 3, 8, 9, 10 and 11.

The insight is integrative, design, technology, governance and community trust are not substitutes but complements. Policymakers and practitioners who treat physical security as a systems problem, attending simultaneously to spatial design, intelligent monitoring, ethical oversight and community voice, consistently outperform those who pursue any single lever in isolation. As cities grow denser and the costs of violence remain catastrophic, this integrated approach offers the most credible pathway to the safer, more just and more inclusive urban environments that the 2030 Agenda envisions.

References

Angwin, J., Larson, J., Mattu, S. and Kirchner, L. (2016) ‘Machine Bias’, ProPublica, 23 May. Available at: https://www.propublica.org/article/machine-bias-risk-assessments-in-criminal-sentencing (Accessed: 1 March 2025).
Brand, P. (2010) ‘Urban Environmentalism and the Medellín Miracle’, Latin American Perspectives, 37(2), pp. 126–142.
Brayne, S. (2021) Predict and Surveil: Data, Discretion and the Future of Policing. Oxford: Oxford University Press.
Claró, S. and Roulet, T.J. (2023) ‘Ethical AI Governance in European Law Enforcement’, Journal of European Public Policy, 30(4), pp. 712–730.
College of Policing (2020) CCTV in Crime Detection and Prevention: A National Audit. Ryton-on-Dunsmore: College of Policing.
Cozens, P. and Love, T. (2015) ‘A Review and Current Status of Crime Prevention Through Environmental Design (CPTED)’, Journal of Planning Literature, 30(4), pp. 393–412.
Donovan, G.H. and Prestemon, J.P. (2018) ‘The Effect of Trees on Crime in Portland, Oregon’, Environment and Behavior, 44(1), pp. 3–30.
Fajardo, S. (2022) Medellín: Urban Transformation as a Political Project. Medellín: Universidad EAFIT.
Fussey, P. and Murray, D. (2022) Independent Report on the London Metropolitan Police Service’s Trial of Live Facial Recognition Technology. London: University of Essex Human Rights Centre.
Gallagher, M. et al. (2018) ‘Street Lighting and Crime: A Randomised Controlled Trial’, Journal of Experimental Criminology, 14(3), pp. 357–375.
Heaven, W.D. (2021) ‘Predictive Policing Algorithms are Racist. It’s Time to Admit It’, MIT Technology Review, 17 July.
Institute for Economics and Peace (2023) Global Peace Index 2023. Sydney: IEP.
Jeffery, C.R. (1971) Crime Prevention Through Environmental Design. Beverly Hills: Sage.
MacDonald, J., Golinelli, D. and Stokes, R. (2010) ‘The Effect of Business Improvement Districts on the Incidence of Violent Crimes’, Injury Prevention, 16(5), pp. 327–332.
Mockus, A. (2021) Bogotá cómo Vamos: Two Decades of Urban Reform. Bogotá: Alcaldía Mayor de Bogotá.
Newman, O. (1972) Defensible Space: Crime Prevention Through Urban Design. New York: Macmillan.
Obermeyer, Z. et al. (2019) ‘Dissecting Racial Bias in an Algorithm Used to Manage the Health of Populations’, Science, 366(6464), pp. 447–453.
Saville, G. and Cleveland, G. (2008) ‘Second-Generation CPTED: The Rise and Fall of Opportunity Theory’, in Atlas, R. (ed.) 21st Century Security and CPTED. Boca Raton: CRC Press, pp. 79–90.
Singapore Police Force (2022) Annual Statistics Report 2022. Singapore: SPF.
United Nations (2022) World Urbanization Prospects: The 2022 Revision. New York: United Nations Department of Economic and Social Affairs.
United Nations (2023) The Sustainable Development Goals Report 2023. New York: United Nations.
UNODC (2023) Global Study on Homicide 2023. Vienna: United Nations Office on Drugs and Crime.
Wilson, J.Q. and Kelling, G.L. (1982) ‘Broken Windows’, The Atlantic, March, pp. 29–38.

Free Publication	Physical Security Officer's Manual (Hindi) — Free Order
DRASInt Mandates	https://www.drasintrisk.com/drasintriskmandates
Physical Security Risk Assessment (PSRA)	https://www.drasintrisk.com/product-page/corporate-sra
Intelligence · Investigation · Analytical Assessment	https://www.drasintrisk.com/product-page/identification-and-investigation-services-iis