Data-driven road safety in Germany: a hybrid analysis of fatal traffic accidents using business-intelligence and machine learning

This study presents an integrated visual and predictive analysis of 16914 fatal road traffic accidents in Germany from 2018 to 2024, using data sourced from the German Unfallatlas. A hybrid methodology combining Microsoft Power BI dashboards and supervised Machine Learning (ML) algorithms in MATLAB was employed to uncover temporal, spatial, and contextual risk patterns. Temporal classification segmented accidents into peak (29.81%), night off-peak (24.02%), and day off-peak (46.17%) intervals, revealing that fatality rates are disproportionately high during peak commuting hours despite their shorter duration. Descriptive analyses highlighted persistent accident clusters in North Rhine-Westphalia, Bavaria, and Baden-Württemberg (Germany), with elevated risks during summer months, Fridays, and Mondays. Predictive modelling with kernel Logistic Regression (LR), kernel Support Vector Machine (SVM), and fine k-Nearest Neighbours (KNN) demonstrated high classification performance, with fine KNN yielding the best results across accuracy, precision, recall, and F1-score. SHAP analysis confirmed Hour of accident and Light conditions as the most influential predictors, especially during peak periods. The results underscore the utility of combining interactive geovisualization and explainable ML to support Germany′s 2030 Road Safety goals by enabling data-driven, time-sensitive, and region-specific interventions.

Authors

Institutions

Publication Details

Journal
Transport
Published
2026-11-19
DOI
https://doi.org/10.3846/transport.2026.28998
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

Data-driven road safety in Germany: a hybrid analysis of fatal traffic accidents using business-intelligence and machine learning

Kamilė Taujanskaitė, Hüseyin Serdar Geçer, Muhammed Fatih Pekşen
Transport
Traffic and Road Safety
article

Data-driven road safety in Germany: a hybrid analysis of fatal traffic accidents using business-intelligence and machine learning

Kamilė Taujanskaitė, Hüseyin Serdar Geçer, Muhammed Fatih Pekşen
article en

Abstract

This study presents an integrated visual and predictive analysis of 16914 fatal road traffic accidents in Germany from 2018 to 2024, using data sourced from the German Unfallatlas. A hybrid methodology combining Microsoft Power BI dashboards and supervised Machine Learning (ML) algorithms in MATLAB was employed to uncover temporal, spatial, and contextual risk patterns. Temporal classification segmented accidents into peak (29.81%), night off-peak (24.02%), and day off-peak (46.17%) intervals, revealing that fatality rates are disproportionately high during peak commuting hours despite their shorter duration. Descriptive analyses highlighted persistent accident clusters in North Rhine-Westphalia, Bavaria, and Baden-Württemberg (Germany), with elevated risks during summer months, Fridays, and Mondays. Predictive modelling with kernel Logistic Regression (LR), kernel Support Vector Machine (SVM), and fine k-Nearest Neighbours (KNN) demonstrated high classification performance, with fine KNN yielding the best results across accuracy, precision, recall, and F1-score. SHAP analysis confirmed Hour of accident and Light conditions as the most influential predictors, especially during peak periods. The results underscore the utility of combining interactive geovisualization and explainable ML to support Germany′s 2030 Road Safety goals by enabling data-driven, time-sensitive, and region-specific interventions.

TransportVol. 41(2)
Sakarya University (TR), Vilnius Gediminas Technical University (LT)
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