Kriston Capps is a staff writer for CityLab covering housing, architecture, and politics. He previously worked as a senior editor for Architect magazine.
CrimeRadar is the world’s first publicly available crime-forecasting tool based on open-access data. But will it work?
Crime, both real and imagined, has been the subject of intense scrutiny during the Rio 2016 Summer Games. The intense security theater surrounding this Olympics may be unrivaled. When the event winds down, crime may return to high levels across Rio. What that means, though, is poorly understood by city residents and the Olympics-watching world alike.
The Igarapé Institute, a Rio-based think tank, is preparing to launch a new tool to help people to understand how and where crime happens in Rio. After the conclusion of the Olympics, the group will formally unveil its CrimeRadar, a map-based application that uses open data to track incidences of crime across metropolitan Rio de Janeiro.
With data from 42 police precincts on crimes committed between January 2010 to March 2016, CrimeRadar tracks some 14 million different crime events. But the app goes beyond mapping historical crimes: Through machine learning and predictive analysis, CrimeRadar will also map out future crime trends—like an open-gov pre-crime heat map.
“A lot of our work has been done with governments and private actors,” says Robert Muggah, research director and program coordinator for citizen security at the Igarapé Institute. “Increasingly now, we wanted to work more with citizens, to enhance their own agency to take steps to protect themselves in these really complicated cities where we live and work.”
Muggah says that, with CrimeRadar, the Igarapé Institute wanted to build out a tech solution for a clear problem: personal security. In Latin America and southern and central Africa, the places where Igarapé operates, these answers often take the form of privatized solutions or databases, not public ones.
“Now that we finally have digitized data, now that we’re helping to clean the information, at source, in the various precincts around the city, now that we have a better sense of the integrity of the data—we felt, couldn’t we make some of this public facing?” Muggah says. “Especially when the government is required by law to make a lot of this data available. It tends to get buried in PDF files and Excel files.”
Muggah says that Igarapé struck a deal with the Institute for Public Security, a state government agency, to build a public-facing mobile app that would show the distribution, intensity, and typologies of crimes across metro Rio. The researchers analyzed data centralized with the ISP along with data from Rio’s 190 system (like 911 in the U.S.) and created 812 categories for crimes. Those break down into capital crimes and violent crimes (like armed assault or intentional homicide), less-intense crimes (thefts, burglaries), and “victimless” crimes (loitering, prostitution).
“We built out a model that uses three data points—the time, the location, and the event—by discriminating in geospatial polygons using these three tiers,” Muggah says. “This algorithm creates a score, a risk score, based on those three data points, for every 250-meter-by-250-meter square unit in the state. You group some of the hundreds of thousands of scores for each sector into deciles to create a simplified, color-coded risk rating, on a scale of 1 to 10.”
Igarapé worked with Via Science—a Boston-based “Big Math” firm with experience helping the Seattle Police Department launch its Real Time Crime Center—to add a predictive or “agile” component to CrimeRadar. Mosaico, a Brazilian venture capital group, contributed support for the design. Engineers at Igarapé put the whole thing together.
CrimeRadar boasts two primary modes: One that documents historical crime events and one that maps projected risk. The future mode uses historical data to build a forecast for crimes, by location and by severity. While other companies have developed predictive policing systems in other places, CrimeRadar is the world’s first such crime-forecasting tool that is publicly available.
“We have over an 85 percent accuracy of mirroring risk against actual events. The beauty of machine learning is that this improves over time,” Muggah says. “The more data, the more information you feed into it, the higher-resolution your risk projections are going to be.”
Muggah says that the institute hopes that CrimeRadar will inspire a more data-driven conversation among citizens in Rio, and counter the often sensationalist rhetoric surrounding crime there. Not that crime in Rio isn’t grisly: At least 50 people have been murdered since the start of the Olympics, most of them in the city’s favelas.
But hysterical headlines can obscure the fact that overall crime has dropped over the past 10 years, or that crime is not distributed exclusively or even primarily among the city’s low-income neighborhoods.
Is it useful for residents and tourists to know that crime in Rio is at its highest rate on Monday afternoons? Certainly. But can a forecast prevent crime?
“We don’t think it’s a panacea,” Muggah says. “Anyone living in a city like Rio ought to have multiple sources for information against which to make judgments about how to move, where to go, and when to go. We feel like it’s a really useful piece of information. Previously, this kind of data was locked up in the archives.”