By Toby Keeping, Director of Sales—Law Enforcement, G2 Research Ltd.
ocation data is ubiquitous, and nowhere is its use more important than in law enforcement. From criminal investigations to operational planning and officer safety, few tools provide value across a police department the way location data does. Gleaning intelligence from suspect-focused, warrant-approved raw data delivered by GPS trackers, cellular phone records, License Plate Recognition, Automatic Vehicle Location (AVL), financial transactions, and other similar sources continues to have significant impacts on both investigations and internal efficiencies. However, many agencies struggle to obtain the full value of the data they already collect on the individuals approved for police surveillance by the courts.
Often, the investigative value of location data relies on data that is less than 24 hours old. Investigators are routinely interested in the “most recent” location of a suspect in order to set up surveillance. If weeks of tracker and cellular data on the suspect have been collected, reams of valuable intelligence are simply ignored with such a narrow focus. For example, is it more important to know a target visited a warehouse yesterday, or to know that he visits the warehouse every second day at 3:00 p.m. and comes from and goes to the same location before and after each visit?
The reason for this shortcoming is simple—time. For most crime analysts and investigators, the analytical process to gain intelligence from thousands of location fixes is manual and time-consuming, but this is not true for all. Here are five ways agencies can use fully analyzed location data to improve investigative success while reducing operational effort.
Application 1—Intelligence Discovery
The collection of location data for investigative purposes is not new. Agencies that can leverage the body of data for an investigation and process it in its entirety are able to see beyond the “right now” element of an investigation. As a result, they are able to find answers to key investigative concerns regarding a target’s patterns and destinations of choice. Understanding the holistic value of the data that is captured, not simply the most recent location of the target, provides the greatest return on surveillance investments. In the event there are connections with other cases, the ability to rapidly examine the information to uncover key intelligence can also lead to improved communication between agencies or between investigative units within a police department.
For example, by running recent data through its system, a major Canadian metropolitan police department proved that a suspect who had originally been cleared based on an interrogation had, in fact, been in the area of the break-ins. Upon further review, the suspect was again apprehended.
Application 2—Predictive Policing
Understanding crime mapping and where and when to apply valuable resources is the common use of location data for predictive policing, but there are other applications as well. By understanding the recurring patterns in a suspect’s data, agencies have the ability to not only predict what a suspect will do, but where and when they will do it. The ability to strategically deploy physical surveillance units to watch or apprehend a suspect provides significant resource management benefits.
The predictive value of this data for a law enforcement agency also increases operational effectiveness and officer safety. Using predictions developed from analysis of location data to direct investigative and technical teams reduces the amount of effort it takes to find the suspect and expedites the replacement of surveillance equipment at the least risky time. Officer safety is paramount at all levels of law enforcement.
|Figure 1: icuWorkbench Prediction|
|Prediction from icuWorkbench. Within one minute, the icuWorkbench can analyze 70, 000 location points and produce a target’s Pattern-of-Life. With another mouse click, the target’s future movements can be predicted. Here, the call-out is showing the predictable activity of a target on Saturdays at 11:00 a.m., ranked by the most predictable activity above a 70 percent confidence rating.|
Application 3—Organized Crime
Going beyond the holistic understanding of a single target, more and more agencies seek to compare the data from multiple targets to better understand the relationships between them. Using multiple sets of location data from two or more suspects to show that they interact at a specific location is powerful for any investigator. Further, understanding patterns in how those entities interact deepens the investigative value of the data. Identification of potential meetings, or dead drops, across many targets not only is attainable with the right tools, but also allows investigators to make intelligence-based decisions throughout long and complex investigations by more fully understanding the relationships at play.
One agency shared that they had identified meetings between two groups of criminal elements who had not been previously thought to be working together. Once the investigators could identify and visualize the targets’ interactions, and with the knowledge that a physical surveillance team was located between the gangs’ vehicles, it was determined that the surveillance teams may have been observed, and, as a result, operational changes were made to ensure officer safety. Another narcotics task force in the United States summarized their results by saying “By far the most beneficial feature has been the multi-track report that has allowed us to identify common residences and targets previously unknown. These have directly led to the creation of high-priority targets and initiation of new and larger conspiracy cases.”
|Figure 2: Travel and Interaction Analysis|
|Multiple Target Correlation Analysis. Within three minutes, the icuWorkbench can correlate the travel activity of these six targets, each with 92 days of data, in order to identify possible meeting locations between targets (red) and locations that are shared (gray), with no overlapping visitation times. The number in the cells represents the number of locations that are shared between the two targets. With another mouse click, detailed reports are available to visualize the actual interaction of the targets.|
Often, when the request to understand a target’s behavior occurs with critical urgency, there is little data available. When actual GPS tracking data is unavailable, it can be a challenge to identify the patterns and intelligence from stand-alone sources of data such as phone records, financial transactions, and other similar sources. Finding ways to fuse those disparate sources of data into a single, cohesive, data set allows the data to collaborate in telling a story about the target in the absence of pre-existing surveillance data.
|Figure 3: Target Comparison|
|Comparing the location and travel of different individuals can lead to significant investigative findings. In this image, two suspects are being compared—one with GPS data (pink) and the other with GPS data and cellular data (red). The comparison clearly shows that the suspect in red was making calls in conjunction with the route he drove in his GPS-tracked vehicle. Further, the data supports that the two suspects were indeed at or near the same address at a point in time that was relative to a specific call for service.|
Application 5—Professional Standards
GPS data is generated from an increasing number of police cars where officer safety is the key objective. Of course, when serving the public interest, there is a duty to answer citizen or stakeholder questions and to sometimes review the actions and whereabouts of the drivers of those vehicles. Having a rapid ability to use data to validate or refute a complaint helps address those situations directly and promptly. In cases where follow-up investigation is required, the ability to analyze the data effectively and efficiently assists agencies’ executives in making the right decisions in a timely manner.
Technology is at the heart of these capabilities, and agencies that desire these enhanced capabilities have gone beyond merely visualizing the data and have, instead, sought technology that provides thorough analysis of the data. Allowing technology to analyze the data removes countless hours of effort for analysts. Even smaller agencies can gain significant returns. For example, by adopting data analysis technology, the Waterloo Regional Police in Ontario, Canada, were able to re-assign an investigator from behind a desk into the field. More importantly, the speed by which the analysis was completed provided additional support to officers in the field. They were provided additional intelligence on those they were monitoring and were able to quickly identify travel patterns and suspicious behavior. This has had a direct impact on public safety, as well as guiding officers into situations with a better understanding of the target.1
In an era of Intelligence Led Policing, the public expects heightened security, while simultaneously demanding more efficient police budgets. By employing analytical technology against location-based surveillance data, law enforcement agencies around the world are achieving just that. They are both employing technology to reduce the effort to obtain intelligence from the data and exploiting new facets of intelligence from the data they already collect, leading to better decisions across the organization. These decisions help determine the next steps in an investigation, to the best place to apply resources, and the optimal time and place to put an officer into a potentially risky situation—all with the goal of ensuring a safer public. ♦
1“From Desk to Duty—How icuWorkbench Re-Assigned an Investigator,” G2 Research Ltd. (via Slideshare), 2013, http://www.slideshare.net/G2ResearchLtd/case-study-waterloo-regional-police-27295806 (accessed February 11, 2014).
Please cite as:
Toby Keeping, “Five Ways to Leverage Location Data,” Technology Talk, The Police Chief 81 (March 2014): 62–63.