University of British Columbia
Real-time crash risk modelling and its applications in intelligent transportation systems
Abstract
dc:descriptionTraditional safety evaluation methods rely on historical crash data, which are limited by underreporting, sparsity, and long observation periods. The emergence of high-resolution trajectory data, enabled by advances in sensing technologies, has facilitated the development of real-time crash risk measures derived from near-misses, supporting a wide range of real-time safety applications. This thesis develops a comprehensive framework for real-time crash risk modelling and its integration into various Intelligent Transportation Systems (ITS) applications. A Bayesian Hierarchical Extreme Value model is utilized to derive crash risk metrics from traffic conflicts, providing a dynamic safety measure that can estimate safety over very short time intervals. This dynamic nature of crash risk warrants dynamic, real-time applications. Four key applications are examined. The first application focuses on network safety and dynamic changes in safety levels, introducing a dynamic spatial modelling framework for identifying and ranking hazardous locations using short-term and long-term crash risk indices. The second extends this methodology to route planning, presenting a real-time safest route algorithm that quantifies and minimizes exposure to high-risk conditions, while examining trade-offs between safety and mobility. The third application demonstrates the use of autonomous vehicle data to evaluate cyclist safety and assess the influence of infrastructure characteristics on crash risk. The final application applies real-time crash risk metrics within a reinforcement learning framework to optimize traffic signal operations at a corridor level, jointly improving safety and mobility. These frameworks are data-agnostic and are validated using real-world data derived from drones, autonomous vehicles, and roadside cameras. The proposed approaches reduce crash risk and delay at the network, route, and intersection levels and form the foundation of future ITS systems supporting data-driven decision-making for safer and more efficient transportation networks.
Degree
thesis:*- Name thesis:degree_name
- Doctor of Philosophy - PhD
- Level thesis:degree_level
- doctoral
- Discipline thesis:degree_discipline
- Civil Engineering
- Grantor dc:publisher
- University of British Columbia
- Year dc:date
- 2026
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Ghoul, Tarek
- Contributors dc:contributor
-
- Sayed, Tarek, 1965-
Rights
dc:rights- Statement dc:rights
-
- Attribution-NonCommercial-NoDerivatives 4.0 International
- Language dc:language
- eng
Identifiers
dc:identifier.*- Handle dc:identifier
- http://hdl.handle.net/2429/95726