This article is based on research conducted as a part of the CA POST Command College. It is a futures study of a particular emerging issue of relevance to law enforcement. Its purpose is not to predict the future; rather, to project a variety of possible scenarios useful for planning and action in anticipation of the emerging landscape facing policing organizations.
The article was created using the futures forecasting process of Command College and its outcomes. Managing the future means influencing it — creating, constraining and adapting to emerging trends and events in a way that optimizes the opportunities and minimizes the threats of relevance to the profession.
Explore this article to uncover insights on these issues:
- The critical role of crime analysts in identifying trends and supporting law enforcement strategies.
- The impact of Real-Time Crime Centers (RTCCs) in providing instant, actionable intelligence to officers.
- The future potential of AI-driven wearable devices for instant data access and enhanced policing efficiency.
- The importance of integrating multiple data sources for comprehensive crime analysis.
- The balance between leveraging AI technology in policing and addressing accuracy and privacy concerns.
By Captain Cassandra Wilkerson
Until the first two-way police radio was introduced in New Jersey in 1933, patrol officers could not directly talk to a dispatcher. [1] There were no mobile data terminals that would display their calls for service. Officers had access to pin maps, static reports and lists of wanted persons, but crime analysis data was not as readily accessible to a patrol officer as it is today.
Access to crime analysis data, however, has gotten better over the years. Automated records management systems (RMS) and computer-aided dispatch (CAD) have become crucial for generating data used by applications such as GIS to help create hotspot policing reports. This evolution of technology over time began to change the landscape of officers’ responses.
Today, law enforcement officers have body-worn and fleet cameras, city surveillance and drones. As we look further into the future, we can see even more technology-equipped patrol officers who have instant access to and interpretation of all police systems, similar to what you see in today’s real-time crime centers (RTCCs). As RTCCs become ubiquitous, officers will have the ability to instantaneously receive similar data in the field. This will be enhanced by the work of crime analysts using AI technologies to facilitate real-time actionable intelligence that can transform the effectiveness of the police in ways we may not even imagine today.
The work of crime analysts
When we say “crime analysis,” it can mean a variety of things. The role of crime analysts depends on how many a department has employed, what their focus is and how the analysis of data is applied. There are three primary types of analysis:
- Intelligence analysis is the study of organized criminal activity linking people with events to assist in identifying and apprehending individuals.
- Strategic analysis focuses on operational strategies and seeks solutions to ongoing problems evaluated weekly, monthly, quarterly or yearly, such as drug activity or auto theft.
- Tactical analysis focuses on pattern recognition by looking at recent crimes reported, such as burglaries and robberies, to assist personnel in identifying investigative leads.
Generally, all three types aim to identify crime trends, patterns and investigative leads and support agencies’ informed decisions in their efforts to reduce crime. [2]
Crime analysts sift through crime data utilizing software programs that pull from various data sources to create reports. According to the International Association of Crime Analysts, included in the list of their daily tasks is to find series, patterns, trends and hotspots as they happen; research and analyze long-term problems; provide information on demand; and develop and link local intelligence. [3] Crime analysts read police reports and daily shift reports and network with analysts from surrounding agencies to compile their reports.
The various responsibilities of crime analysts are time-consuming tasks. If we consider how long it takes one analyst in a midsize department to gather this information, we can see the challenges of getting it into the hands of field officers expeditiously to allow a swift response. [4]
The solution to expediting these tasks lies in technologies we have yet to see. To better see the future, we must consider what our analysts are working with today.
Where we are now
Crime analysis today widely employs geographic information systems (GIS) for data collection and spatial analysis, also known as crime mapping and hotspot analysis. In Alexandria, Virginia, a film crew followed crime analyst supervisor Joe Ryan, who explained the best way to define GIS is by placing data into a geographic context. [5] GIS helps him analyze where things happen in relation to each other. With GIS, analysts can put incidents on a map and look at the relationship between them — where a crime occurred, for example, along with where a vehicle was towed and where a person was contacted.
All GIS platforms are based on the various systems the information is being fed from, most prominently the RMS-CAD systems commonly used by police agencies. Prior to GIS, analysts often put maps on the wall and used color-coded pins to mark where incidents occurred. The time consumed in gathering data, even utilizing more current GIS-fed analytics, would limit how much information is dispersed to field officers. Fortunately, there are solutions emerging that can enhance the robustness of analysis without adding time or complexity to that work.
Software companies offer platforms officers can access from their mobile devices to review data like city camera images, body-worn camera footage, arrest and vehicle records, and automated license plate reader (ALPR) data. [6] Such software pulls from multiple data sources integrated into the system. These can also include the department RMS and other records. The officer can use the search tools to locate information relative to their current needs, whether it be geospatial information for locating wanted persons or vehicle information based on partial plates. This still requires officers to pause their responses, particularly during live calls, to search for information; however, this is near-real-time information, limited to what the officer is specifically searching for. Under stress, an officer may not think of a particular search parameter that could provide valuable information.
Software company, SoundThinking, created CrimeTracer, an advanced analytics tool that generates immediate investigative leads to field officers and detectives. In a 2012–2019 Oakland Police Department case study, the department faced staffing shortages and partnered with Forensic Logic (now CrimeTracer) to access more than a billion law enforcement records in one centralized interface. [7] As a result of that success, Oakland PD now encourages its field officers to use this software as a tool during the initial investigation to compile leads while the crime is still fresh. As they integrated this analysis into their work, their patrol officers weren’t just responding to the crime but were identifying and apprehending those responsible for the crime and ultimately putting a dent in the violent crime rate.
In the five years after implementing this approach, Oakland saw a 50% decrease in shootings, 42% drop in homicides and 38% decline in robberies. [7] This is an example of how an agency suffering staffing constraints can effectively fight crime despite it.
Of course, the information collected from these technologies is only as good as the information entered into them. The more agencies that buy in and agree to share their information, the more data is available to compile robust reports and assist in investigative analysis across jurisdictions. Ultimately, there is one approach that can accomplish this: regional live data sharing. A critical factor is ensuring the private technology selected for any RTCCs ensures interoperability with surrounding agencies. If the technology is not compatible, it creates a real challenge for sharing data.
Agencies like Flagler County (Florida) Sheriff’s Office and others have implemented real-time crime centers staffed with analysts and sworn officers that monitor LPR cameras and city surveillance to aid officers in the field with live incidents. RTCCs provide field officers with enhanced assistance in response to calls where dispatchers don’t otherwise have the time or access to this technology to do so. [8] Nonetheless, these centers must be staffed by humans, at a time when staffing police departments is already challenging. In the future, however, that may not necessarily be the case.
Where we want to go
What law enforcement needs from AI is accuracy and reliability. For example, if data sources such as facial recognition are to be integrated for field officers who can take swift action, the accuracy of that information needs to be foolproof to eliminate the risk of mistaken identity and liability.
According to the “Washington Post,” several California cities have banned the use of facial recognition technology (FRT), including Oakland, San Francisco, Alameda, Berkley and Santa Cruz. Among concerns that led to the bans was the misidentification of individuals leading to wrongful incarceration. [9] Beyond these localized bans on FRT, Assembly member Phil Ting introduced AB-1814, which would prohibit judges from signing search or arrest warrants where the sole basis for the probable cause is an FRT-generated match. Currently California has no state limitations on how law enforcement uses FRT. AB-1814 is still under legislative consideration. [10]
To overcome these objections, law enforcement will need to chart a path forward for the use of AI in policing. Understanding AI is here to stay and only advancing, there is an inevitable shift to leverage this technology to the benefit of safer communities. There are a variety of ways leaders can change minds, and they will be unique to each jurisdiction based on its political climate and views of police and technology.
As AI infiltrates existing technology, crime analysis software will be a significant beneficiary. We can already experience instantaneous responses from AI when seeking information from OpenAI sources like ChatGPT-4o, OpenAI’s the newest flagship model that offers users features including analyzing data, creating charts, providing responses from the web, and assistance in summarizing, writing and analyzing uploaded files, to name a few. ChatGPT-4o developers are also testing a more helpful memory experience in which the user controls the memory bank. For example, the user can request ChatGPT-4o to remember specific conversations the user has with it and ask it to forget certain portions, saving the user from having to repeat information later. [11]
Adapting platforms like ChatGPT-4o raises the possibility of officers in the field no longer needing a remote, human-staffed RTCC. They can merely use a wearable device like Humane’s Ai Pin, a small device that can clip to the outside of your clothing within earshot of your voice. Using generative AI, it acts as a personal assistant ready to take its owner’s verbal requests. It can take photos, send texts, project a visual interface onto your hand and act as a virtual assistant ready to search the internet and communicate back to you. [12] Now think of a wearable device capable of searching police databases and camera systems specific to law enforcement that provides officers with the real-time information RTCCs provide, but at a much faster rate.
Chris Hsiung, who recently served as the undersheriff for San Mateo County, California, says his agency has integrated an AI agent that is not predictive but has proven effective in both current and cold cases. [13] The technology uses AI to “connect the dots” among multiple data sources, including calls for service, outstanding warrants, video surveillance and police reports. He explained, “AI took 1½ seconds to find a lead in a cold case that took a person three weeks to find.” [13] If this technology was implemented in the field, officers could prompt a wearable device like Ai Pin for specific information relative to their beat and potentially receive a vast amount of information from a multitude of data sources within seconds.
In another example of the potential for intelligent systems to enhance police effectiveness, Elk Grove, California manages an RTCC that incorporates live data for officers while calls are in progress. In one recent instance, a person was brandishing what was thought to be a firearm. The analyst began monitoring video surveillance cameras in the area and determined the individual described was holding a hairbrush. [14] This swift relay of information from the analyst to the responding officers allowed them to alter their response tactics and avoid a possible tragedy.
Certainly, as reluctance to use AI for policing ebbs, the positive impacts of its use will emerge. These include improved resource allocation, enhanced situational awareness, increased collaboration with regional agencies, enhanced decision making and a reduction in crime as perpetrators learn the police have tools that can identify them, track their movements and allow officers to already be there when they arrive.
Conclusion
The future of AI-led crime analysis can mean officers having access to real-time information faster than a human can search for and relay it, as they do in a RTCC. Imagine a time when officers are equipped with an Alexa-like device that recognizes their jurisdictional responsibilities and shift hours and proactively feeds them real-time situational awareness via crime analysis.
RTCCs are at the beginning of this curve as agencies learn how to respond to near-real-time information faster than a dispatcher typically delivers. The future of policing will be an even more fluid and dynamic space that will only become faster as we move further into the technological evolution.
References
1. Borelli F. The evolution of police communications (and what’s still ahead). Officer.com. July 2015.
2. Crime Tech Solutions. Criminal intelligence management: best practices. Crime Tech Weekly. May 2017.
3. International Association of Crime Analysts. About crime analysis. IACA.net. 2024.
4. Ratcliffe JH. Integrated intelligence and crime analysis: enhanced information management for law enforcement leaders. Police Foundation. 2007.
5. Fairfax Network-Fairfax County Public Schools. GIS – on the job: crime analyst. YouTube. 2016.
6. Nerbetski L. Peregrine 101: improving field operations for officers and their community. Peregrine. April 2023.
7. Forensic Logic Coplink. FL – a search engine and a crime miracle in Oakland. SoundThinking. March 2022.
8. IPVM team. National real time crime center association president Nikki North interview. IPVM. April 2024.
9. Charlton A. Oakland becomes third U.S. city to ban police use of facial recognition. Salon. July 2019.
10. Digital Democracy Calmatters. AB 1814: law enforcement agencies: facial recognition technology. Calmatters. 2024.
11. OpenAI.com. Memory and new controls for ChatGPT. OpenAI. February 2024.
12. Dave P. Humane’s AI pin is a $700 smartphone alternative you wear all day. Wired. November 2023.
13. Goldenberg P, Gips M. AI is set to revolutionize policing: are we ready? Police1. March 2024.
14. Calams S. A day in the life of a real-time information center analyst. Police1. May 2023. Available at:
About the author
Captain Cassandra Wilkerson has been in law enforcement for 23 years working for the Pittsburg Police Department in California’s Bay Area. She has worked a variety of assignments throughout her career and was recently promoted to captain, overseeing patrol services. She holds a Master of Science degree in criminal justice and is a graduate of POST Supervisory Institute and POST Command College.