Citation

Muratore G, Vannelli A, Micheli D. Transp. Eng. (Amsterdam) 2023; 12: e100172.

Copyright

(Copyright © 2023, Elsevier Publishing)

DOI

10.1016/j.treng.2023.100172

PMID

unavailable

Abstract

Rome municipality publishes geo-referenced vehicle accidents reports and detailed road traffic counters that, in conjunction with Open Street Map roads details, enable deep analyses of the tragic phenomenon. This study focuses on the full year 2021, during which a total of 62,081 vehicles casualties were registered by Rome traffic authorities. The broad Rome territory, overcoming a bunch of Italian chief towns, the wide and capillary set of arteries and the municipality roads extension greater with respect to many European capitals, enhance the statistical relevance of the analyzed data. Artificial intelligence method takes into account three potential accidents predictors as the level of road traffic, the event timing, and the driving complexity of the different zones. The proposed method has the capability of linking time/traffic/complexity indexes of a specific zone of a city with the accident riskiness, moreover, is applicable to whatever urban complex scenario and takes care of the result format in order to facilitate road safety decision makers in smoothing the tragic sequence of accidents. Further possible improvements directions are discussed.

Language: en

Keywords

Management decisions; Open data; Random forest; Road safety; Rome; Smart city; Vehicle’ accidents