A Predictive Service for Highway Hotspot Policing

Joao Vitor Venceslau Coelho, Jose De Souza Silva, Adelson de Araujo, Nelio Cacho, Frederico Lopes, Jose Alex De Medeiros

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

2 Citations (Scopus)
1 Downloads (Pure)


Brazil's road traffic mortality rates have been increasing since the 1960s and in 2003 road traffic crashes in Brazil were responsible for over 26 percent of fatal injuries. Some actions need to be taken to make it safer, hence, we have proposed, implemented, and deployed a support system that predicts and generates, based on the prediction, a program-card for a Federal agency responsible for patrolling highways which cross most of the big Brazilian cities. We used the data from the Federal Highway Police (PRF) about road accidents to test the performance of the generated models and demonstrate the designed dashboard. We obtained a better performance than the baseline of comparison on the three groups used on the study: Alcohol_drugs, lack_attention and aggressive driving, reaching a R2 score of 0.440, 0.688 and 0.453 respectively.

Original languageEnglish
Title of host publication2020 IEEE International Smart Cities Conference, ISC2 2020
ISBN (Electronic)9781728182940
Publication statusPublished - 28 Sept 2020
Externally publishedYes
EventIEEE International Smart Cities Conference, ISC2 2020: Smart Cities Solutions for New Challenges, Including a Pandemic - Virtual Conference, United States
Duration: 28 Sept 20201 Oct 2020


ConferenceIEEE International Smart Cities Conference, ISC2 2020
Abbreviated titleISC2 2020
Country/TerritoryUnited States
CityVirtual Conference
Internet address


  • Forecasting
  • Hotspot Prediction
  • Machine Learning
  • Public Security
  • Road Accident Prediction
  • n/a OA procedure


Dive into the research topics of 'A Predictive Service for Highway Hotspot Policing'. Together they form a unique fingerprint.

Cite this