Page 9 - Case study: Crime prediction for more agile policing in cities – Rio de Janeiro, Brazil
P. 9
Crime forecasting models are another example of utilizing new technologies for more agile security:
promising data-driven and problem-oriented approaches that can speed up decision-making and
providing smart solutions that help reduce human biases and inefficiencies . In an era marked by
7
the fourth industrial revolution, AI, the Internet of Things and Big Data are available to help law
enforcement and criminal justice authorities adopt more effective policing strategies.
In 2016, the Igarapé Institute partnered with Via Science - a data analytics firm - to develop the
CrimeRadar app, a public-facing crime forecasting platform that evaluates relative crime frequencies
in different locations and times in the metropolitan of Rio de Janeiro.
2.2. Implementation
CrimeRadar is a digital platform that forecasts the probability of crime. It runs on smartphones and
desktop browsers. The software uses advanced data analytics to show real and relative crime rates
and risks for different neighborhoods at different times in the Rio de Janeiro municipality.
Figure 1: Illustrative CrimeRadar Platform
CrimeRadar visualizes the safety levels in specific locations and times. By making crime data more
accessible and transparent, it improves security for citizens.
The underlying crime data was retrieved from the state Institute for Public Safety and official crime
records produced by the state civil police. The platform was launched during the Rio Olympics in
August of 2016. Residents and tourists could access the website to view the predictions displayed
on an intuitive mobile heatmap.
Solution Development: CrimeRadar was conceived by the Igarapé Institute and developed with
partners Via Science and Mosaico. The Igarapé Institute provided expert knowledge about the region
and worked with various data providers to gather and verify the accuracy of the historical data.
Crime prediction for more agile policing in cities – Rio de Janeiro, Brazil 3
