“My first year with the Richmond police, I went to every crime scene where the victim remained in order to try to understand data behind the crime,” Dr. McCue says. “I’ve stood next to parents who’ve lost young people. I wanted to be part of a solution that stops this from happening.”
Along the way, Dr. McCue has applied her training in behavioral analysis and data analytics to predicting where U.S. military will encounter IEDs (Improvised Explosive Devices) in Iraq and Afghanistan and forecasting likely locations where the Lord’s Resistance Army, a terrorist group in Central Africa, will attack.
“The key is to look beyond the emotionally-charged, heinous aspects of deviant and violent behaviors and see the underlying patterns,” Dr. McCue explains. “Even the most serious patterns of violent crime are relatively predictable if you look at them in the right light. That’s why computer algorithms are so helpful – they’re dispassionate all the time.”
Dr. McCue acknowledges, wryly, the skepticism she encountered when first introducing data analytics to Richmond’s finest soon after she joined the city’s police department in 2000. “Cops are from Mars, analysts are from Venus – we get this all the time,” Dr. McCue says with a laugh. “I’d be working with veteran detectives who were still using note cards to manage cases and they’d come in at the end of the day and say, ‘Hey, that crystal ball thing you’ve got going, I think I could use some help from it.’
“I always remembered my place on the force. The analyst’s role is to support the operational personnel,” she adds. “I’m never going to tell a cop what to do – it’s important to understand where I fit in that ecosystem.”
While shootings by the Northern Virginia sniper were captivating the nation, Dr. McCue led a team of DigitalGlobe analysts to integrate existing computer models in order to create a unique data-based behavioral profile specific to the sniper. The team started with Signature Analyst, the software program used to predict the locations of IEDs, and added, among other features, a powerful capability called ViewShed, which incorporates line of sight analysis.
“We started with a huge area where this guy was operating and were able to reduce it by 90 percent, based on predictable patterns from the first series of shootings,” Dr. McCue says. “Then we factored in how the sniper was setting up his shots and identified an even smaller area to search.”
The Department of Homeland Security, which had retained DigitalGlobe, distributed the company’s predictive model to a range of federal, state and local law enforcement agencies throughout Northern Virginia, Maryland and Washington, DC. Security was intensified in the areas most likely to be targeted – and the sniper went underground.
“Once we released our model and police started using it, the shootings stopped – the sniper had been very active, and then for six or seven months – nothing,” Dr. McCue says. “We’d like to think that putting officers in the kinds of locations the sniper favored was a major deterrent.”
The sniper was finally apprehended on June 17, 2011 in Arlington National Cemetery in Virginia. A search of his possessions and home revealed instructions for making IEDs and other explosives, along with extremist literature. The sniper subsequently pled guilty to the shootings.
“Our goal is to prevent bad things from happening by pre-positioning resources and responding more rapidly,” Dr. McCue affirms. “It’s a very exciting time, with our predictive technologies reaching new levels of effectiveness, and becoming part of day-to-day police work.
“I’m very much an evangelist for using advanced analytics; I personally believe this is the right path and a significant contribution to the field of law enforcement.”
