Correctional agencies are building some of the most sophisticated geospatial intelligence programs in public safety. At this year’s Esri Safety and Security Summit, Pennsylvania’s Department of Corrections (PADOC) showed how a modern corrections organization is using spatial analytics and 3D mapping to move from after‑action reporting to anticipatory risk management.
From Data Overload To Operational Insight
Damien Mscisz, Intelligence Analyst Supervisor with the PADOC, described his world as a “challenging job” where something new arrives every day and where the stakes are measured in violence prevented, not just incidents investigated. When he joined the intelligence bureau in 2021, he found a system overflowing with data yet starving for connection, context and visibility.
Pennsylvania supervises roughly 38,000 inmates across 24 facilities and about 28,000 parolees in the community. That population generates more than two million recorded phone calls per month plus e‑messages, tablet activity, video visits, in‑person visits, GPS traces, financial transactions and investigation reports. Historically, staff used this information reactively: a fight or assault would occur, investigators would go back to phone calls, visits and human intelligence to ask “why did this happen” and then reconstruct an after‑action narrative. As Mscisz put it, “We didn’t have a data problem. We have tons of data. We just didn’t always know where to go to get it, how to bring it together and what to do with it.”
The core challenge was fragmentation. Housing assignments, GPS points, financial flows, confidential tips and K9 search results all existed, but no one could see them together in one operational picture. To solve that, Mscisz’s unit set out to connect, integrate and visualize that information using ArcGIS Enterprise, building a corrections geospatial intelligence program that serves prison and parole staff as well as federal, state and local law enforcement partners. The team invested in extract‑transform‑load processes to consolidate disparate systems into “one true source of data” and then layered spatial analytics, 3D facility mapping and network analysis on top. “A picture is worth a thousand words,” he said. “A map is worth ten thousand words.” The goal was straightforward: give decision makers a living operational picture instead of pages of disconnected reports.
3D Facilities As Intelligence Surfaces
That picture starts inside the facilities. Correctional institutions function like small cities where infrastructure, movement, behavior and community ties all intersect. To understand that complexity, the PADOC modeled entire prisons in 3D, transforming static floor plans into interactive geospatial environments that show who is housed where, what gangs are present and where risk may concentrate.
Inside one facility, staff can toggle layers to see where gang members live, identify cells shared by associates and spot “dark corners” where certain inmates may try to cluster. The 3D environment has become a practical tool for daily problems. When a staff member says, “I’m looking for this person, I can’t find them” and only has a nickname, they no longer have to dig through multiple databases. Instead, they query the nickname inside the 3D map, get every instance of that alias and narrow the population in seconds. As Mscisz explained, if an investigator is “looking for a ‘shiz’ who sells drugs in B unit at this jail,” they can narrow the search to a handful of candidates in just a couple clicks.
The same spatial thinking supports targeted violence prevention. One warden noticed recurring fights between inmates from different cities, particularly “Philadelphia versus the world,” and asked for a simple hotspot product that would reveal where inmates from those rival cities were being housed. The intelligence team built an experience that lets staff visualize where people from Philadelphia, Erie and Pittsburgh are placed within the facility. When the map shows clusters of Philadelphia and Erie inmates in close proximity, leadership can quickly see “opportunities for calamity” and redeploy or separate individuals to head off future violence. “This is what it’s about,” Mscisz explained. “It’s about anticipating the risk, about learning so that we can make better decisions as we go forward. It’s not always just reactive… How can we prevent it from happening again?”
Networks, Money Flows And Hidden Weapons
Beyond physical placement, the bureau uses scripting and network analysis to map invisible relationships. They ingest phone call logs and identify inmates who call the same number, receive visits from the same person or share senders for commissary funds, then visualize those ties as a network graph.
Within one of the state’s newest facilities, staff can pull up an individual inmate and see the cluster of other prisoners connected through shared contacts and financial flows. Some of those units are known problem pods, and overlapping networks can reveal gang coordination, illicit economies or emerging threats that were hidden in raw transaction data. Financial activity alone accounts for roughly 60,000 senders and about 20 million dollars in commissary payments per month, generating names, addresses and phone numbers that can feed geospatial analysis. “Think about what we can do with all that,” Mscisz said.
One of the most striking examples of this approach is a “hidden weapon analysis” built on twelve million rows of historical housing data. The team tracked where each inmate lived over time, with whom they were celled and what those cellmates did in the future. By looking at patterns where an inmate never personally gets caught with a weapon yet a significant number of past cellmates do, analysts can flag potential “producers of weapons” or facilitators who teach others how to make or hide contraband. Instead of waiting for an assault and reacting after injuries occur, staff can use that information to deploy K9 units, conduct focused searches and intervene before violence. “What your cellmate does in the future might say something about what happened when he housed with you in the past,” Mscisz explained. In other words, longitudinal spatial data can turn seemingly mundane housing moves into intelligence signals.
Community‑Scale GPS Intelligence And Partnerships
The same philosophy extends into the community. GPS monitoring data has gone from static trace logs to dynamic, analyzed insights. Pennsylvania consumes live GPS feeds through an API, updating every hour and maintaining a rolling window of seven days of movement for roughly 450 individuals on GPS. Raw GPS points can be noisy, but even simple filters can surface actionable patterns. GPS analysis, for example, can help flag parolees assigned to non‑Philadelphia offices who travel into Philadelphia regularly, potentially for illicit purposes they are not cleared to undertake.
The bureau goes further by analyzing the data every six hours for co‑travelers. Yellow points on the map highlight places where parolees from different agents and offices move together. In the past, a parole agent could only look at one person’s GPS points at a time through the vendor portal and then write “I didn’t see anything unusual” based on isolated traces. Today, co‑traveler analytics add the missing context: who else was there, how long they stayed and whether they repeatedly visit one another’s addresses. “We’ve never been able to do this before,” he said. “This solves that problem. It adds context to the data.”
Context also transforms how PADOC works with external partners. The team fuses GPS data with city open data to explore time‑space relationships between parolee movement and reported crime. In Philadelphia, they map crime incidents and look for individuals on GPS who were within fifteen minutes and in the same location when a crime was reported. The output can generate a large number of leads, so partnerships and access control are key. Using ArcGIS Enterprise, the bureau shares tailored views of the data with task forces and law enforcement groups, allocating access to the right resources that can investigate specific patterns. The same approach extends to specialized overlays such as visits to gun stores, federal firearms licensees and mapped gun ranges, which interns compiled and analysts now use as buffer layers.
Culture Shift And Strategic Impact
For Mscisz, geospatial intelligence in corrections is part cultural shift and part technical project. He spent years “knocking on doors” as a state parole agent and freely admits he does not always speak in perfect technology terms, but he knows the use cases and where the pain points are for staff in the field. “We’re kind of new to the scene and I’m new to the business,” he admitted. “It’s all about going from reactive to proactive.”
His core lesson: identify one data element that, if connected to everything else, would give a better picture and help solve a real problem. “We’re doing it with one thing at a time,” he said. “Phones, emails, video visits, in‑person visits, all these data points that we have. We’re bringing them in one at a time to our enterprise and working through the problems.”
For Pennsylvania, corrections is a three‑billion‑dollar enterprise statewide. Every better decision about how to deploy K9 units, security staff and investigative resources carries operational and fiscal weight. The geospatial intelligence program is designed to give leadership faster, clearer insight into risk so they can anticipate threats inside facilities and across communities rather than simply documenting what went wrong after the fact. That operational picture now spans 3D facility maps, GPS‑based community surveillance, network graphs of calls and money, and risk models built on years of housing history. By turning siloed data into interactive maps, dashboards and analytics, Mscisz’s unit supports violence prevention, smarter parole supervision and more focused planning for searches, K9 deployments and special operations.
For the autonomy and unmanned systems community, the takeaway is the same. Correctional agencies that embrace geospatial intelligence are discovering that the same principles driving advanced air mobility, autonomous security patrols and mission planning for unmanned systems can transform how they manage small cities behind the walls and the communities beyond them. Data volume is not the problem. Connection, context and visualization are.