Police analytics can be extremely valuable in the fight against terrorism and crime. By identifying which events are most likely to escalate, predictive techniques can both improve prevention capability and control costs by deploying officers before escalation and where they are needed most.
Predictive technologies also allow police services to focus more efficiently by highlighting patterns for further investigation. It can also be a resource tool to explore security threats by studying the people, events, locations and organizations involved in events that require a law enforcement presence, and can also deliver information to the field when needed.
“Forward-thinking police departments are now turning to analytics technologies to find quicker, smarter and more accurate means of analysing data to better assess, predict and prevent criminal activity. With predictive analytics, analysis of historical data trends is used to better inform daily operational decisions, enabling a step-change within policing,” notes a recent Accenture study.
No wonder then that a growing number of public safety and law enforcement agencies are using analytics and mining available intelligence as a way of helping their frontline officers. In fact, in its research of best practices, the Ottawa Police identified more than 300 police agencies in the United States alone along with several in Canada that are using police analytics.
The Future of Policing
In many cases, changes in policing practice arise from crises. For example, in an earlier column I wrote about how police forces are turning to body cameras in response to public concerns about police practice; so too with analytics. For several police forces, the development of analytical programs arose due to a high profile crime (or series of crimes) that was felt ought to have been solved more quickly had there been some kind of analytics capability in place. In some cases it was the recognition that integrating information from various internal databases (or, more broadly, among aligned agencies) would have turned up information to solve a case more quickly or even prevented some crimes. To put it another way, bits and pieces of information dispersed among several forces would have provided enough information to solve the case if taken together. In the U.S., this concept has spawned state-level, multi-agency fusion centers.
Officers are able to evaluate incident patterns and analyze an array of data inside of the Real Time Crime Center, a $3 million state-of-the-art crime monitoring and analysis hub.
In other cases, geographical mapping and other analytical software would have been able to provide an assessment as to where an offender was likely located, leading to quicker capture.
For the Ottawa Police Services (OPS), there was no specific catalyst, only a desire to evolve the crime analysis function and leverage modern technology to assist frontline officers and ensure a common operational picture across its jurisdiction. The lack of a crisis meant there was no time pressure to form the unit, nor was there a predetermined model for the unit – it could be designed based on appropriate research and discussion.
Finding out What Works
The approach taken by the OPS was to thoroughly understand the opportunities, benefits, and costs of standing up an analytics and intelligence-oriented operations centre. A critical part of the research phase was to visit jurisdictions employing police analytics/intelligence programs – from newly deployed to long-standing centers – to see them in practice, and to hear from those running the programs and from frontline officers on how a data-driven and analytics unit was of benefit to them.
Upwards of 300 Real Time Crime/ Operations/Intelligence Centers exist in the United States alone, with each state having DHS-mandated fusion capabilities (fusion refers to integration of policing and intelligence data from beyond the local police force). Units in many cities in both Canada and the United States were visited, including Memphis, Miami-Dade, Chicago, Calgary, Toronto, and York. Some of these units, such as Miami-Dade and Toronto, were relatively new, while others such as Calgary and Chicago had been operational for 10 or more years. Some cities used analysts in their operations while others did not. Some were fusion centers and others had only in-house data. Some utilized comprehensive analytics software purchased from major vendors, while others developed internally or used basic off-the-shelf software. The idea was to conduct broad-based research and then visit selected centers to maximize the learning potential.
Insights gained from the investigation of other analytics and intelligence units. What OPS learned in their North American investigation of police analytics units could fill a book. I asked Randy Mar, one of the principle developers of the centre, and Kevin Mason, the lead on the analytics component, what they considered among the top lessons learned.
- A fusion approach is best. Geographically, Ottawa borders Quebec and the City of Gatineau in particular. With five bridges spanning the Ottawa River, criminal activity on either side of the provincial border impacts the other. An analytics/intelligence system would ideally access data from all areas. As the Nation’s Capital, there are multiple municipal, provincial and federal agencies responsible for public safety and law enforcement, including RCMP, Ontario Provincial Police, Sûreté du Quebec, Border Services, Military Police, and others. Finally, close proximity to other cities along the Highway 401 corridor (like Cornwall and Kingston with their own police services) widens the scope. Research into other analytics initiatives has shown the value of integrated regional data. The OPS design described in this article starts small and sets the foundation and opportunity to establish a National Capital Region fusion center over the longer term.
- A critical component is the analysts themselves. Despite working at the data analytics Centre, they must develop a relationship with the officers they support. It must be remembered that analytics is not about software, it’s about analysts interpreting the data, providing insights, and working with officers to analyze and disseminate relevant intelligence. To be a true support function you need an analytics analyst who knows the frontline officers, has a relationship with and understand their needs. They must be able to recognize what the officers need before they ask for it – computer systems cannot do this. The OPS design reflects this with analysts that, while being centrally located, are assigned to platoons in addition to having geographic responsibilities. As well, the unit has a staff sergeant in command. A mix therefore of trained civilian analysts and uniformed police.
- Leverage your partners. OPS noted that the more successful police units recognized the advantages of teaming up with partner agencies. For example, OPS partnered with the Telfer School of Management at the University of Ottawa on a variety of analytics and big data projects that look at how to best employ analytics and how to use different analytics packages. Other centers have partnered with vendors (software providers). INTERSECT, a national capital region body comprised of all the key public safety partners, will be an important ally.
- It’s about people not just software. This may sound a lot like point #2, but it’s really tied to the technology. Without naming vendors, they found that some products tended to be very expensive (beyond the budgets of many police forces), and when modifications or upgrades were required, these too could be prohibitively expensive. Ottawa instead turned to less complex software that could be modified by the user and upgraded at low cost. For the start-up of the OPSOC, the emphasis was more on geographic systems (GIS), with analysts being trained in how to use and disseminate the results of their efforts. In looking at existing analytics initiatives, OPS consistently heard that analytics was not about technology replacing people but technology enhancing the effectiveness of what people (officers and analysts working together) can do. Again it’s the people, not the software.
Bringing the end user into play
A second aspect of the approach taken by the OPS to design their analytics and intelligence program was to involve frontline officers, like Constable Ian Matyas, who visited four North American analytics centres on behalf of OPS. Also, Cst Lana Cameron was part of a group that determined what the analytics unit would focus on (referred to as trigger calls), what intelligence is best conveyed, and to which responding officers – enroute to the call, on scene, or beginning an investigation. (For interviews with these two constables, read “Police Officers’ View of Analytics” in this edition, following this article.) By integrating police officers into the research and later the design of the unit, and visiting many different kinds of analytics units, OPS was looking at best practices and ways that analytics can best support the frontline officer.
Designing the OPSOC
The eventual design of the unit, which went operational (in start-up mode) in October 2016, was based on the research and input of these and other officers.
Armed with these and other findings, the OPSOC team set out to develop the concept of operations and design the new centre. An evolutionary approach was taken to start smaller, prove the concept in operation (termed virtual backup), and set the foundation to potentially evolve into a regional fusion center.
The evolution approach would give the organization time to learn, tie into budget and resourcing realities, as well as partnership development. Implementation is to be done in four phases:
- Intelligence dissemination (virtual backup enroute, on scene, and during early investigative stages);
- Directed preventative/targeted patrol;
- Strategic operational oversight, redistributing resources and ensuring operational deconfliction; and,
- National Capital fusion center.
The overarching objective and organization of the unit is described in OPS documents: “The Ottawa Police Strategic Operations Centre (OPSOC) has been developed to be agile to the needs of the organization supporting frontline and investigative policing, managing resources and in the future leveraging multi-agency information.”
How will the frontline be supported?
By bringing the expertise of a Watch Commander (Staff Sergeant), Crime Intelligence Analysts (Civilian), and experienced Investigative and Intelligence Coordinators (Constables) into a common room, the OPSOC will support an evidence-based policing approach that incorporates the tenets of intelligence-led and problem oriented policing, and academic findings to enable frontline and investigative efficiency.
The Crime Intelligence Analysts have dual assignment roles within the OPSOC; each analyst is assigned a platoon as well as a specified geographic area. The platoon assignment allows for relationship building which includes the analyst with the operational team and provides clear support lines for the platoon to access an analyst. The geographic assignment provides equal workload for analysts to identify patterns, trends and problems throughout the city to support sergeant-directed patrol (agile, measured) and/or directed pro-active patrol (scheduled, measured).
The analysts are a professional group trained to support open source intelligence (OSINT), geospatial intelligence (GEOINT), information searches, and correlate data to information, then into relevant intelligence on a per case basis. The final product to the frontline is geared to be timely, relevant, and answer to the needs of the officer.
The analysts and sworn members will be bringing into the OPSOC specialty software and training to enable for geospatial awareness of intelligence, resource allocation, and existing logistics across the city. In addition, the analysts have multiple software applications to support both internal and external situational awareness.
The OPSOC is designed to provide “virtual” support frontline officers in three broad types of situations:
- On emergency calls (referred to as trigger calls);
- For everyday policing by identifying patterns, trends and potential crime areas; and
- Officers can contact their analyst or the OPSOC from the road for immediate support/ answer questions, even if it is not a trigger call.
Fifty types of trigger calls (such as kidnapping, missing persons, active shooting) were identified that would initiate involvement of the OPSOC. When one of these calls occurs, the watch commander initiates OPSOC trigger procedures, which results in all members of the OPSOC working on the one call. Analysts would look at social media, conduct scrapping, look for patterns, trends and problems. Once a pattern is identified and geospatial awareness is developed, the analysts then provide an awareness and a tactical report is passed on to the officer.
Let’s take the case of an active shooting scene. While enroute to the call, the OPSOC provides responding officers with relevant information on the shooter to help with situational awareness. Additional information mined from internal data sources and scrapped from social media would then be provided by the OPSOC while on scene and after the scene has been secured. The focus being on analytical and information support with appropriate information being provided as needed rather than all information coming across the officers screen while enroute to the call (a common complaint from the frontline officer perspective).
In addition to the trigger calls, two additional kinds of involvement were planned for the crime intelligence analysts and the OPSOC. Call it a problem-oriented policing approach for every day policing. One example of this is ongoing analysis looking at crime patterns and trends. The Crime Intelligence Analysts will work on identifying patterns, trends and problems of crime within the city using a geospatial approach.
The geospatial approach allows for sergeant-directed patrol and/or directed pro-active patrol in areas requiring focused attention to address crime issues. These geofenced areas can be measured for patrol effort, efficiency to address the issue, and effectiveness to resolve the issue.
A second type of involvement in every day policing lies in the way the unit is organized. Having OPSOC staff assigned to platoons creates a situation in which each platoon can contact the Strategic Operations Centre (OPSOC) from the road and speak with their analyst or coordinator for immediate support to an ongoing call, even if it is not a trigger call.
In either case (trigger call or every day policing) the OPSOC provides timely analysis and relevant information to officers. This both frees up officers’ time and allows for directed pro-active patrol. The idea therefore will be to develop and deliver timely and relevant information for frontline officers and others working in investigative positions based on integration multiple information sources and analysis using geospatial systems.
Jonathan Calof is the Executive Editor of FrontLine Safety and Security magazine.
Subsequent Frontline Safety and Security articles will track the ongoing evolution, functionality, and development of the Ottawa Police Strategic Operations Centre.