Related Experiment Video
Updated: Apr 16, 2026
08:05
Design and Analysis for Fall Detection System Simplification
Published on: April 6, 2020
11.3K
Application of a Rule-Based Approach in Real-Time Crash Risk Prediction Model Development Using Loop Detector Data.
Ali Pirdavani1,2, Ellen De Pauw1, Tom Brijs1
1a Transportation Research Institute (IMOB), School for Transportation Sciences, Hasselt University , Diepenbeek , Belgium.
Traffic Injury Prevention
|March 21, 2015
Summary
This study developed a real-time crash risk prediction model using traffic data to enhance motorway safety management. The model accurately predicts crash likelihood, improving traffic safety systems.
Area of Science:
- Traffic Engineering
- Road Safety Management
- Predictive Modeling
Background:
- Dynamic safety management systems increasingly utilize real-time crash risk prediction.
- These models integrate crash data with real-time traffic surveillance data from sources like loop detectors.
Purpose of the Study:
- To develop a real-time risk model for predicting crash likelihood on motorways.
- The model is intended for integration into traffic management systems.
Main Methods:
- Utilized traffic-related characteristics as prediction variables.
- Employed a rule-based approach for model development, comparing it with binary logistic regression and decision trees.
- Data collected on the E313 motorway in Belgium from June 2009 to December 2011.
Main Results:
- Traffic flow characteristics like volume, average speed, and speed variations significantly predict crash occurrence.
- The final classifier achieved 70% accuracy in predicting crashes and 90% accuracy in predicting non-crash instances.
- A 10% false alarm rate was observed.
Conclusions:
- The developed model can predict the likelihood of crashes on motorways.
- Findings support the implementation of this model within dynamic safety management systems.
