Spatiotemporal analysis and machine learning prediction of road traffic crash hotspots: a case of Amhara region, Ethiopia

Road traffic crashes remain a major public health and development challenge in Ethiopia. While previous studies in the Amhara region have examined crash trends, contributing factors, or specific road corridors, there is limited evidence on spatiotemporal hotspots and on predicting population-adjusted crash rates. This study aimed to investigate the spatial and temporal distribution of road traffic crashes in the Amhara region and to build machine learning models to estimate crash rates across administrative zones. The dataset comprised police-reported crash records from 2013 to 2022 along with spatial, demographic, traffic exposure, and school-related variables. Spatial clustering was assessed using the Getis-Ord Gi* statistic. Machine learning approaches were evaluated using repeated five-fold cross-validation. The North Shewa zone consistently emerged as a statistically significant cluster of reported crash burden, and the East Gojam zone appeared as an emerging hotspot during the later years of the study period. Of the models evaluated, Random Forest achieved the strongest cross-validated performance (RMSE = 0.866 ± 0.205; R 2 = 0.678 ± 0.266). The GIS and machine learning findings are complementary: GIS identifies geographic clustering of reported crashes, whereas machine learning estimates population-normalized crash rates. These results can support regional road safety planning by helping prioritize zones for enhanced monitoring, enforcement strategies, data quality improvements, and targeted corridor-level assessment. Future work should incorporate geocoded crash records and road segment-level data to enable the design of site-specific interventions.

Authors

Institutions

Publication Details

Journal
Scientific Reports
Published
2026-11-15
DOI
https://doi.org/10.1038/s41598-026-80840-3
Primary Topic
Traffic and Road Safety
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Spatiotemporal analysis and machine learning prediction of road traffic crash hotspots: a case of Amhara region, Ethiopia

Habtamu Zelelew, Jayesh R. Juremalani, Alamirew Mulugeta Tola, Abeje Tilahun Fetene
Scientific Reports
Traffic and Road Safety
article

Spatiotemporal analysis and machine learning prediction of road traffic crash hotspots: a case of Amhara region, Ethiopia

Habtamu Zelelew, Jayesh R. Juremalani, Alamirew Mulugeta Tola, Abeje Tilahun Fetene
article en

Abstract

Road traffic crashes remain a major public health and development challenge in Ethiopia. While previous studies in the Amhara region have examined crash trends, contributing factors, or specific road corridors, there is limited evidence on spatiotemporal hotspots and on predicting population-adjusted crash rates. This study aimed to investigate the spatial and temporal distribution of road traffic crashes in the Amhara region and to build machine learning models to estimate crash rates across administrative zones. The dataset comprised police-reported crash records from 2013 to 2022 along with spatial, demographic, traffic exposure, and school-related variables. Spatial clustering was assessed using the Getis-Ord Gi* statistic. Machine learning approaches were evaluated using repeated five-fold cross-validation. The North Shewa zone consistently emerged as a statistically significant cluster of reported crash burden, and the East Gojam zone appeared as an emerging hotspot during the later years of the study period. Of the models evaluated, Random Forest achieved the strongest cross-validated performance (RMSE = 0.866 ± 0.205; R 2 = 0.678 ± 0.266). The GIS and machine learning findings are complementary: GIS identifies geographic clustering of reported crashes, whereas machine learning estimates population-normalized crash rates. These results can support regional road safety planning by helping prioritize zones for enhanced monitoring, enforcement strategies, data quality improvements, and targeted corridor-level assessment. Future work should incorporate geocoded crash records and road segment-level data to enable the design of site-specific interventions.

Scientific Reports
Samara University (ET), Wollo University (ET), Kombolcha Institute of Technology (ET), Bahir Dar University (ET), University of Gondar (ET)
Traffic and Road Safety
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon