Analytical forecasts show police force where, how to allocate officers
Western Australia Police forecasts workforce demand and crime within 1.5%
As director of strategy and performance for Western Australia Police, Shaun Hodges knows the crucial role of analytics in public safety.
"Intelligence-driven policing means looking at the analytical side of things," explains Hodges. "We not only look at where crime is happening, how much of it is happening, and who is likely to be involved in it, but also how many police officers are going to be available to deal with it."
The agency has used SAS for 20 years in operational areas like financial management for budgeting and forecasting as well as for business intelligence and data integration.
Using advanced analytics is certainly helping us to deliver better policing outcomes.
Director of Strategy and Performance
Forecast accuracy within 1.5 percent
Analytical forecasts show WA Police how to most effectively allocate their workforce where and when the demand is most needed.
"If we are going to spend $1 million on a breath-testing campaign, we can analyze historical data to find out where to stage it," Hodges says. "We can forecast workforce demand and crime within 1.5 percent, such as look at the number of break-ins that have occurred this month and forecast how many are likely to occur next month – within 1.5 percent. This helps us plan where our workforce needs to be to counteract crime."
He wants information in each police officer's hands to improve how they do their job. "We want to link data and analysis to thermal maps so that each regional manager can see if they need more police resources," Hodges says. "They will produce reports where crime has occurred and overlay it with data from police vehicles. This means we can really start to look on the ground at where the crime is happening to plan strategies."
Seeing the impact on crime
Data integrity is essential to forecast accuracy, Hodges emphasizes.
"Initially, our accuracy margin was 10-15 percent," he explains. "We discovered that this was due to a delay in the officers submitting their reports. We improved accuracy dramatically once they could see the impact that data-driven policing can have on reducing crime."
Eventually, Hodges says, he wants the department to go beyond predicting how many crimes will occur to uncovering the lead indicators for different types of crime. "What I want to do now is understand what drives crime," he explains.
His overall objective is to make sure the department makes the right decisions about where to increase police presence to serve the community when it's at its most vulnerable.
"With my background as a police officer and detective, I understand what information police officers need when they are dealing with situations and need to get on the front foot," says Hodges, a former officer. "If they're better informed when they go into a situation, they achieve a better outcome – and that's a better outcome for the community."
"Using advanced analytics is certainly helping us to deliver better policing outcomes."
The world's largest police area
One of eight jurisdictions in Australia, WA Police covers 2.5 million square kilometers, including two regions, 14 districts, 163 police stations and nearly 4,000 uniformed officers.
Western Australia Police needed a solution that would drive better policing outcomes by being able to more accurately forecast where resources should be allocated.
Able to forecast, within 1.5%, the likelihood of particular crimes occurring in areas and allocate the most appropriate policing resources to counteract it.