Browse Content by Topic:
Data Analysis & Public Safety: How Predictive Policing Can Benefit 9-1-1 Personnel
Author: Sean Bair & Susan C. Smith, Bair Analytics
Copyright: 9-1-1 Magazine, Feature Content
Often, discussions about 9-1-1 technology are geared toward live computer aided dispatch screens, automated vehicle locator information and applications and communications systems between dispatch and officers in the field. While all critical to emergency services, there lies a growing interest in analytics as they relate to data that is gathered in computer aided dispatch and other systems. Whether analyzing response times on various call types or determining the optimal positioning of emergency medical services through the temporal and spatial analysis of call data, administrators are realizing the real need for, and appropriate use of, analytics.
BAIR Analytics, a Colorado-based law enforcement, military and emergency response analytics company, has met the needs of hundreds of agencies across the United States and Canada with ATACRAIDS – Automated Tactical Analysis of Crime Regional Analysis and Information Data Sharing. The web-based information and intelligence sharing and analysis program is optimized for all levels of an agency, from communications/dispatch to analysts to officers to fire and emergency medical responders. Separate applications for law enforcement and other first responders maintain the necessary confidentiality of law enforcement data.
In law enforcement, decision-makers at all levels seek to better understand how, where, when, what and why calls for service are occurring across the jurisdiction. Certainly educated guesses about certain areas of the jurisdiction generating certain call types – such as intoxicated subjects in/around bars, traffic congestion and accidents during and following major entertainment events, and shoplifters apprehended inside shopping malls – have been assumed without the use of sophisticated technology. Additionally, temporal (time) analysis may have been conducted informally such as scheduling additional traffic units during rush hours and providing more DUI enforcement teams on weekend nights.
But, lacking in law enforcement has been on-going, in-depth analysis of calls for service by type, location, time and other factors. Using the data routinely collected in existing systems, a system like ATACRAIDS allows decision-makers to conduct this level of analysis with the results being both specific and practical. Customized dashboard analytics allow each user to build their own custom analytics or to select from nearly 200 predefined analytics that help the user visualize calls for service and self-initiated activity both within their own jurisdiction and across jurisdictional lines. With dozens of variables on call data easily accessed, users can search across data related to individual officers, districts, addresses, call type and many other fields to answer questions to help guide future deployment and efforts.
For instance, many jurisdictions regularly re-visit the boundaries of their districts, beats and zones. Using information gleaned from calls for service data, decision-makers can identify average response times to various call types in different areas of the jurisdiction and adjust boundaries to achieve optimal response times. This is a very traditional use of analytics in law enforcement. With an analysis system, date ranges can be just as quickly altered to compare those response times during different times of the year, such as seasonally or for places like college towns, “school in session” versus “school not in session”. Similarly, jurisdictions subject to wide swings in population on- versus off-season can analyze response times, call volumes and call types and adjust staffing accordingly based on the results of the analysis. This may result in significantly different district or beat boundaries during different times of the year. And, post implementation, decision-makers can continue to analyze response times and tweak the boundaries as the face of the jurisdiction changes.
It is important, for instance that data from calls for service data can be sorted and queried with the results of the searches available for download or further analysis. Often times, downloading the data for further analysis is useful for administrative and/or reporting purposes. Supervisors of all levels can leverage the power of the analytics developed to understand their team, shift or assignment in much more significant detail than previously available. Literally any question concerning the data included can be broken down, analyzed, and summarized for further use and decision-making purposes if you have the data and software used for analysis. And, with a regional system, data from surrounding jurisdictions can be included in the analysis as well. No longer can law enforcement afford to stop the analysis at the jurisdictional borders. Rather, true regional call and crime analysis can be conducted. Collaboration on shared concerns is much more likely when both, or all, jurisdictions involved are able to recognize and respond to the problem.
Call data can also be analyzed in comparison to other types of data. Comparing top crash locations with volumes and types of citations delivered is an excellent approach to gauge the effectiveness of citations issued in comparison to crash types experienced. If an analysis reveals that tickets are not being written for the reasons leading up to crashes at certain locations, enforcement methods can be altered to more effectively address the actual problems. And, with the new enforcement methods in place, additional analysis can be completed that support or do not support the changes made. For instance, if it were determined that crashes at 1st and Main Street were most often the result of red light violations during the morning and evening rush hours on weekdays, traffic cars could be assigned to observe and heavily enforce through citations those running red lights. With the traffic cars specifically assigned during rush hours on weekdays, analysis could be conducted to determine if these efforts reduced this specific type of accident during rush hours on weekdays. Targeted enforcement is the key to efficiency and effectiveness.
Decision-makers also crave timely and useful reports. The ability to receive timely, useful and accurate dynamic reports that can be delivered on-demand instantly or through scheduled emails is critical. These custom reports can be sent to an end user through a .pdf attachment which could include a pin map, density map, selected analytics and even contain links to live maps of incidents or links back to the raw data from the report. These reports can be shared amongst specifically those interested in the information, thus preventing “everyone from getting everything” resulting in information overload. But for very specific activities and groups, these targeted reports are very useful. For instance, a supervisor may wish to hold his/her staff accountable for call volumes in a particular area of the jurisdiction. By automatically delivering the same report to others involved in the effort on a certain date/time just prior to, for instance, a weekly meeting, the supervisor assures all participants have the same information to utilize in assessment and decision-making. This also gives supervisors an opportunity to hold their staff accountable for activity or call reductions.
Also valuable are sophisticated analytical functions able to compare activity over two time periods, such as last summer compared to this summer, or the past three months compared to the most recent month. This functionality is particularly valuable when police have completed a particular response or initiative in the jurisdiction, such as directed patrols during certain hours to a particular area of town. Decision-makers can visually see areas that increased versus areas that decreased in activity. Forecasting can also be conducted through the selection of a certain type of call and/or certain areas of the jurisdiction. Being able to forecast the times and areas more likely to suffer additional calls or events is useful in determining appropriate staffing and deployment. Animated density is yet another helpful way to analyze crime. Using animated densities, the movement of calls or crimes over time can be observed. Whether the user chooses to analyze the all call type or only particular types of calls or self-initiated activity over a day, weeks, months or even years, an animated density provides the visual cues, along with the supporting statistical information regarding the activity.
Despite being the focus of this article, analysis cannot be limited to the analysis of crimes, calls for service and self-initiated data. Rather, analysts should be able to analyze many different types of data for analysis purposes. This includes events data, such as crimes and crashes, as well as information on offenders of all types, property recovered and field interview data collected. Being able to compare the call locations on false burglary alarms to actual residential burglary incidents or the locations of juvenile contacts compared to graffiti incidents can be very useful in both deployment and investigation. Similarly, using call and crime information to conduct a geo profile of a serial offender can help guide investigations and prioritize leads. Keyword searches across all data sources, including report narratives, help investigators locate linked cases in their own jurisdiction and from jurisdictions surrounding. Link analysis tools within the program allow investigators to visualize the connections and links between cases and people.
It was but a short time ago that conducting meaningful analysis on calls for service and other police data relied on a full-time analyst on staff to answer even the simplest of questions. While analysts are still key to effective analysis, including jurisdictional knowledge, software exists and is readily available and easily understood and utilized for all members of a department to analyze data and answer questions about that data. Using those answers, decision-makers from the highest to the basic level can make decisions about the allocation of resources, the boundaries of districts, beats and zones and much more than ever before. This capability frees up the crime analyst to conduct more in-depth studies at a much higher level to support tactical and strategic analysis within the jurisdiction.
Sean Bair is president of BAIR Analytics, a Colorado-based corporation dedicated to providing analytical services and innovative software solutions, including ATAC, ATACRAIDS, and RAIDS Online, to the law enforcement, intelligence, and defense communities. Mr. Bair was the Assistant Director at the National Law Enforcement and Corrections Technology Center as well as a former Crime Analyst and Police Officer for the Tempe, Arizona Police Department. He is the creator and developer of ATAC (Automated Tactical Analysis of Crime) analytical software which is currently used in hundreds of law enforcement, defense agencies and academic institutions worldwide. He has trained thousands of analysts, officers and investigators around the world in the analysis of crime.
Susan C. Smith is the Vice President of Operations for BAIR Analytics, a Certified Law Enforcement Analyst (CLEA) through the International Association of Crime Analysts, an adjunct policy analyst for RAND, Corporation and teaches at several universities, including the University of Missouri-Kansas City, Tiffin University and Regis University. She previously served as a crime analyst for the Overland Park and Shawnee, Kansas Police Departments, retiring in January 2013 after 22 years of service. Susan is the current President of the International Association of Crime Analysts(IACA) and a past president of the Mid-America Regional Crime Analysis Network (MARCAN).
For more information, see http://www.bairanalytics.com/
Photo from 9-1-1 Magazine archive.