Police Department Modernizes Policing with ArcGIS Enterprise
The St. Petersburg Police Department fights crime and provides public safety services for the city of St. Petersburg along the Florida gulf coast. The Information and Technology Services (ITS) team includes a GIS specialist, a systems analyst and a data specialist, along with a team of five analysts in the Intelligence-Led Policing (ILP) Unit supervised by Dr. Richard Ferner, Jr. In collaboration, these teams administer the ArcGIS platform and create maps and dashboards used by more than 560 officers, detectives, and supervisors in the department to make data-driven decisions. The ITS Unit manages the ArcGIS Enterprise portal and all associated products; while the ILP Unit is a data-driven center that provides support to tactical, strategic, and operational initiatives. Through innovative use of Esri technology, the department has modernized and improved its policing methods.
For years, the police department relied on text-based information. Crime metrics were collected in spreadsheets, with information coming from paper reports. “Everything was text based,” said Frank Ullven, Systems Analyst in the ITS Unit. “We didn't have any maps; it was all street names and addresses.” Ullven remembers how police officers had to read addresses in columns to figure out where incidents were occurring. Matching crime to addresses was difficult without visual representation especially when new streets were added or street names were changed. “Officers don’t have to memorize all that, like this used to be 2nd Street South, but now it’s University Way South.”
St. Petersburg Police Department
Overwhelming volume of text-based information collected in spreadsheets
ArcGIS Enterprise, ArcGIS Pro with Crime Analysis toolset, ArcGIS Insights, Operations Dashboard for ArcGIS, ArcGIS Web AppBuilder, Survey123 for ArcGIS, ArcGIS Monitor
Better policing with data-driven maps, dashboards, and analysis
Dr. Ferner, Supervisor of the ILP Unit, said that department stakeholders such as the Chief of Police and Command staff were overwhelmed by the sheer volume of information, which made it difficult to gain meaningful insights to make decisions and do their jobs efficiently.
Back then, the police department implemented an older technology which lacked flexibility to support custom, user-friendly visualizations, explained Dr. Ferner. “Those solutions did not promote robust and relevant visualizations. Users had little incentive to utilize those tools when deliberating on a course of action.” Possible actions include proactive patrol assignments or developing leads in identifying suspects.
Five years ago, the arrival of a new Chief of Police — Chief Anthony Holloway — marked the St. Petersburg Police Department’s transition to a data-driven organization. Chief Holloway advocated to adopt CompStat as a management model. CompStat, or Computer Statistics, is a policing method that relies on timely and accurate information to efficiently combat crime while improving police accountability.
The team realized they needed to move away from a desktop environment and find a way to deliver content in an interactive manner. “The minute I heard Esri developed an enterprise solution that could allow the user community to interact with the content we publish, I knew, unequivocally, that that was the solution we needed,” Dr. Ferner said.
In 2016, the department implemented ArcGIS Enterprise 10.5 and ArcGIS Pro, and kept pace with each upgrade, steadily adding products such as ArcGIS Insights, Operations Dashboard for ArcGIS, and more.
ArcGIS Enterprise was key in supporting the department’s need for a secure, behind-the-firewall enterprise platform that powered their data management and analysis needs, especially in the context of law enforcement data.
Ullven said: “There’s a level of comfort in knowing that it’s our data on our in-house system and not somewhere that we don’t have control over who sees it or what’s being accessed.”
Before ArcGIS Enterprise, analysts created static content that was distributed through email and posted to a file share and Microsoft SharePoint. Disparate tables, charts, and graphs did not tell the whole story and there was no way to customize dashboards to visualize data for a common operating picture.
Since moving onto the ArcGIS platform, analysts can now create dashboards and story maps that focus specifically on what each unit needs to know, so they are not overwhelmed with irrelevant information.
Dr. Ferner said: “Now that we’ve evolved onto the Esri platform, we can carve out highly nuanced, relevant data that matters and answers questions. It helps the staff and supervisors carve out a strategy and set of tactics for immediate application.”
These days, data is refreshed 45 minutes before each shift, allowing Watch Commanders to detect emerging crime trends and evaluate initiatives on the go. Analysts use the Crime Analysis toolset in ArcGIS Pro and ArcGIS Insights to analyze data, while sharing out the interactive content via story maps and operations dashboards via ArcGIS Enterprise.
“Canvassing a neighborhood no longer requires a 6-foot-long paper map and tons of hard-to-decipher markings,” Kevin Christy, GIS Specialist in the ITS team, said. ArcGIS Web AppBuilder made the process more targeted and efficient, while Survey123 for ArcGIS enhanced the department’s Eagle Eye program that geocodes addresses for an up-to-date camera locator application.
In one example, the command team was looking for insights around the destruction of parking meters downtown St. Petersburg. The analysts pulled up crime data around parking meters, made a prediction on crime hot spots using ArcGIS Pro, and published it on a map within ArcGIS Enterprise so staff could use those predictions in their operations planning.
Data for the prediction was acquired from the records management system where a detailed account of each parking meter location was documented. In turn, analysts geocoded each parking meter location in ArcGIS Pro and then subsequently published the content as a hosted feature layer in ArcGIS Enterprise.
In turn, patrol officers, armed with this analysis, patrolled the risk area identified in the prediction and encountered the suspects and subsequently arrested them as they were preparing to commit more crimes. Such analysis and predictions were not possible before.
More and more officers are also coming to the ILP and ITS units to request specific dashboards. For example, they want to see what ArcGIS can do, and when they get a tour of the technology, their eyes light up, said Christy.
“I see their wheels turning. The big thing is tailoring it for exactly what the end user wants. Whether it’s ‘I want these metrics in my dashboards,’ or ‘I want these colors,’ or ‘I want an app that does XYZ,’ it’s all about giving them what they want. If they don’t get exactly what they want, they’re less likely to use it.”
Dr. Ferner noted that a growing younger police force has contributed to a critical mass of users in the department. “One of our biggest challenges was the cultural dynamic in giving them access to these products, and assigning accountability to the metrics,” he said. “The workforce here is also becoming younger, and we’ve discovered they are more adept with using different technologies. Even if we had a product 10 years ago that is as sophisticated as this one today, I don’t think the staff then would have been so accepting of these technologies.”
The move from static maps and static data that officers couldn’t fully engage with to a more interactive mapping platform has transformed the department.
Christy said: “Esri allows [officers] to have a customized product that really presents them with geospatial data that prompts questions now. They can look at the data, they can ask questions, and now they are getting more insights than they had in the past. It’s really helping to drive better policing.”
One thing that no other product in the past helped us address is what I call a broken feedback loop between the analysts, detectives, and patrol. Now that we have [ArcGIS Enterprise], we are working as co-producers of the content. People that write the reports, call takers from the call center, and ultimately analysts that produce the nuanced analysis downstream. These discussions were notably absent…now the feedback loop has been put into place where it never really existed before.