Data-driven policing: how geospatial analysis is fighting violent crime

Summary:

• West Midlands Police's cutting-edge 'Project Guardian' aims to tackle violent crime in the area
• The project, which uses geospatial analysis to target crime hotspots, has helped achieve a 16% reduction in knife crime
• A combination of temporal and geospatial analysis allows precisely targeted police presence and data can be updated frequently
• Collecting data from partners such as hospitals and business improvement districts has been vital to the success of the project

West Midlands Police has transformed its approach to tackling violent crime through cutting-edge geospatial analysis, enabling it to turn broad, imprecise crime hotspots into highly targeted, actionable locations. By leveraging technology from Esri UK, the force achieved a 16% reduction in knife crime under Project Guardian. This award-winning initiative combines advanced mapping, temporal analysis and data-driven problem solving to deliver precision policing and strengthen partnerships. Here DCI John Askew and Tori Starmer, Intelligence Analyst, speak to Grant Powell MBCS about the project and the benefits of intelligence-led crime prevention.

Before introducing geospatial analysis, what was the biggest challenge in tackling knife crime in the West Midlands?

John: The biggest challenge was precision. Our hotspots were huge areas covering four or five square miles. Our old mapping tools gave us rough density maps, but they didn’t tell us which roads were criminogenic or when crimes were most likely to occur. So patrols weren’t focused, and problem solving was difficult. We were essentially working blind compared to what we can do now.

How did Esri UK’s technology change the way you identify and respond to crime hotspots compared to previous methods?

Tori: It’s been transformative. We can now drill down from broad areas to specific street segments — even down to 10 metres within a 100-metre road. We can break hotspots into grids, calculate statistical significance and identify the most criminogenic streets.

John: We also overlay temporal analysis, so we can tell officers, ‘be on this road between 8:00 and 10:00.’ That level of precision means patrols are in the right place at the right time. The speed is incredible, and where refreshing hotspots used to take months, we can now do it in a week. And it’s not just about finding hotspots, it’s also about gaining a deeper understanding of the micro-locations and the factors behind them; schools, shops selling illegal vapes, risky premises… that’s where the real problem solving starts.

Project Guardian achieved a 16% reduction in violent crime. What specific strategies or insights from the data made that possible?

John: Guardian was built on three pillars. First, the Knife Crime Task Force, which saw 50 officers deployed into data-driven hotspots for high-visibility patrols, knife arches, school engagement, weapon sweeps and plainclothes tactics.

Second, the ‘intelligence cell’, which involves scanning daily for actionable intel linked to hotspots, identifying people believed to have carried knives on multiple occasions and managing them through visits, diversion or enforcement.

Third, problem solving using the SARA model, which stands for Scanning Analysis Response and Assessment. We look at offenders, repeat victims, high-risk facilities and environmental factors. Then we build tailored plans, often with partners, to tackle root causes.
The data allows us to [both] randomise patrols for maximum deterrent effect and focus on the right streets at the right times. That’s what has driven the reduction.

What were the biggest challenges in collecting, cleaning and utilising crime data for this project, and how did you overcome them?

Tori: Partner data was the most difficult, such as health, local authority, HMRC, DWP… everyone agrees sharing data is the right thing, but, as mentioned previously, often the practical ability isn’t there. Location accuracy was another issue. If someone turns up at hospital with a knife injury and won’t talk, we might record the crime at the hospital or their home address. That skews the data. We’ve had to cleanse it, excluding hospitals, custody blocks and police stations from hotspot generation. We also avoid merging datasets to prevent double counting. Instead, we overlay and then interrogate anomalies through scanning and analysis. And with the software, we can refresh hotspots in a week, validate them with neighbourhood teams, and iterate quickly.

What role has collaboration with external partners played in the success of this project?

John: Partnerships are vital, but they’re challenging. We’ve tried to integrate local authority and health data into our hotspot analysis, but data sharing and system incompatibility make that difficult. The prime focus of hospitals, for example, is on treating patients, not generating intelligence for us. Where we’ve had success is in using our analytical outputs to guide partner action. Business improvement districts (BIDs) have also been key because retailers often report incidents to them more readily than to us. We overlay their data with ours rather than merging it, to avoid double counting. When all the hotspots converge, we know that’s where we need to focus.

How did you measure the financial impact of reducing crime, and why is that figure significant for the community?

John: We use cost-of-crime metrics, which estimate the societal cost of each offence — lost productivity, healthcare, ambulance use, criminal justice costs and wider social impacts. Harm scores are useful internally, but partners understand pounds and pence.

For you

Be part of something bigger, join BCS, The Chartered Institute for IT.

When you tell a council that a hotspot of crime costs them £20 million a year, it changes the conversation. It helps unlock resources and focus investment where it will have the greatest impact. It’s not about valuing victims in monetary terms; it’s about showing the scale of harm and the benefits of prevention.

What is Operation Fearless and how has it been shaped by lessons learned from Project Guardian?

John: Operation Fearless is a major new initiative designed to tackle crime and make areas safer. It was launched in Erdington earlier this year, backed by £880,000 funding from the Police and Crime Commissioner. The funding, sourced from money seized from criminals, has been instrumental in transforming the area. With increased police presence and enforcement activity, residents in Erdington have already reported feeling much safer. 

The biggest lesson learned has been that concentrated, ring-fenced deployment works. We thought we’d need six months to turn things around. With focused problem solving and dedicated officers, we did it in three. We also learned that static communities respond better than transient ones. Erdington was slightly less challenging than Birmingham city centre, where people come and go from all over the country. So, for our next deployment — Soho Road — we’re choosing a location that suits the model.

Another lesson is the importance of early engagement with partners. Their budgets are set annually, so if we want resources in April, we need to start talking [months earlier]. And finally, we proved that throwing quality effort into one area can make a huge difference, not just in crime reduction but in public confidence.

Beyond knife crime, what other areas of policing do you believe can benefit most from geospatial data analysis?

John: Pretty much everything. Burglary, vehicle crime, robbery, exploitation… they all have victim, offender and location elements. We’re already adapting tactics for violence against women and girls, including offences involving sex workers. We’ve set up a problem-solving hub and a prevention hub, and we’re training analysts and officers across the force. The SARA model works for any crime type. If it works for something as complex as knife crime, it will work elsewhere. And that’s exciting.

Finally, congratulations on winning the Data Science Project of the Year award. What are your thoughts on receiving that recognition?

It’s always great to be recognised for the work we’re doing, especially when it involves something as sensitive as knife crime. We’re proud of the journey we’ve been on. It’s been a big shift for us. Senior leaders now talk about it every day and it’s well and truly embedded as part of our operating model. It’s also been great for the team. People like Tori, who started as a young analyst, are now subject matter experts in hotspot policing. That recognition shows we’re doing something unique and impactful. However, we’re also mindful that one stabbing is one too many. So, while the award is absolutely fantastic, there’s still a lot more work to do to keep our streets, towns and cities safe.

For more information please visit the West Midlands Police and Esri UK.