Universidad de Málaga (ATREIDES project)

El Español (26 April 2026) reports that researchers at the University of Málaga have developed an AI system that designs police patrol routes intended to anticipate and prevent crime. It runs on a digital twin of the city built from geographic, social and crime-history data, uses multi-agent reinforcement learning so that each patrol is given a different, less predictable route, divides the city into 50 m by 50 m cells and uses a "coverage index" to estimate the number of officers needed. The article says the work was done in collaboration with the Comisaría Provincial of the Policía Nacional and is still at laboratory stage, not operational. It describes the work as part of Juan Palma-Borda's doctoral thesis within the ATREIDES project of the national R&D plan, funded by the Junta de Andalucía and FEDER funds, and quotes researcher Eduardo Guzmán saying that less predictable routes can deter criminal activity. The researchers' paper (Palma-Borda, Guzmán-de-los-Riscos and Belmonte-Martínez, Engineering Applications of Artificial Intelligence, vol. 166, 2026, article 113706) reports tests on three districts of Málaga using actual city crime data, which it does not publish, with coordinated routes covering more than 90 per cent of the top 3 per cent of crime-incidence locations; the paper does not name a police partner.