A Comprehensive Analysis of Road Crashes at Characteristic Infrastructural Locations: Integrating Data, Expert Assessments, and Artificial Intelligence

• Published in: Infrastructures (Basel) Vol. 9; no. 8; p. 134
• Main Authors: Ivanišević, Tijana, Vujanić, Milan, Senić, Aleksandar, Trifunović, Aleksandar, Čičević, Svetlana
• Format: Journal Article
• Language: English
• Published: Basel MDPI AG 01-08-2024
• Subjects: Artificial intelligence; Automobile safety; Bridges; Construction costs; Crashes; Data analysis; Data collection; expert assessments; Fatalities; Infrastructure; Injury analysis; Pandemics; Railroad crossings; Railway tunnels; road crashes; Roads; Roads & highways; Software engineering; Traffic control; Traffic engineering; Traffic flow; Traffic safety; Trends; tunnels; Vehicles; Serbia
• ISSN: 2412-3811; 2412-3811
• DOI: 10.3390/infrastructures9080134

Road crashes, although random events, frequently occur on roads. However, certain characteristic infrastructural locations require detailed analysis regarding the frequency of road crashes. This study examines the dynamics of road crashes at characteristic infrastructural locations in Serbia from 2018 to 2022, focusing on bridges, tunnels, railroad crossings, and road work zones. Using data on road crashes from official reports, the analysis includes trends in crash rates, fatalities, injuries, and material damage during the above-mentioned time frame. In addition to the data analysis, 22 experts from the fields of traffic engineering ranked the mentioned characteristic infrastructural locations in terms of road safety. The same questions were asked to six different artificial intelligence software programs. The findings reveal significant variations in crash rates across different infrastructures, with bridges and road work zones having the highest number of crashes. Expert assessment is in line with the analysis of the results, while artificial intelligence gives a completely opposite assessment.