A bi-level framework for real-time crash risk forecasting using artificial intelligence-based video analytics

This study proposes a bi-level framework for real-time crash risk forecasting (RTCF) for signalised intersections, leveraging the temporal dependency among crash risks of contiguous time slices. At the first level of RTCF, a non-stationary generalised extreme value (GEV) model is developed to estima...

Full description

Saved in:
Bibliographic Details
Main Authors: Fizza Hussain, Yasir Ali, Yuefeng Li, Md. Mazharul Haque
Format: Default Article
Published: 2024
Subjects:
Bi-level
Framework
Crash risk
Online Access:https://hdl.handle.net/2134/26039038.v1
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1818164499261161472
author Fizza Hussain
Yasir Ali
Yuefeng Li
Md. Mazharul Haque
author_facet Fizza Hussain
Yasir Ali
Yuefeng Li
Md. Mazharul Haque
author_sort Fizza Hussain (8139471)
collection Figshare
description This study proposes a bi-level framework for real-time crash risk forecasting (RTCF) for signalised intersections, leveraging the temporal dependency among crash risks of contiguous time slices. At the first level of RTCF, a non-stationary generalised extreme value (GEV) model is developed to estimate the rear-end crash risk in real time (i.e., at a signal cycle level). Artificial intelligence techniques, like YOLO and DeepSort were used to extract traffic conflicts and time-varying covariates from traffic movement videos at three signalised intersections in Queensland, Australia. The estimated crash frequency from the non-stationary GEV model is compared against the historical crashes for the study locations (serving as ground truth), and the results indicate a close match between the estimated and observed crashes. Notably, the estimated mean crashes lie within the confidence intervals of observed crashes, further demonstrating the accuracy of the extreme value model. At the second level of RTCF, the estimated signal cycle crash risk is fed to a recurrent neural network to predict the crash risk of the subsequent signal cycles. Results reveal that the model can reasonably estimate crash risk for the next 20–25 min. The RTCF framework provides new pathways for proactive safety management at signalised intersections.
format Default
Article
id rr-article-26039038
institution Loughborough University
publishDate 2024
record_format Figshare
spelling rr-article-260390382024-02-19T00:00:00Z A bi-level framework for real-time crash risk forecasting using artificial intelligence-based video analytics Fizza Hussain (8139471) Yasir Ali (13768642) Yuefeng Li (521173) Md. Mazharul Haque (3643810) Bi-level Framework Crash risk This study proposes a bi-level framework for real-time crash risk forecasting (RTCF) for signalised intersections, leveraging the temporal dependency among crash risks of contiguous time slices. At the first level of RTCF, a non-stationary generalised extreme value (GEV) model is developed to estimate the rear-end crash risk in real time (i.e., at a signal cycle level). Artificial intelligence techniques, like YOLO and DeepSort were used to extract traffic conflicts and time-varying covariates from traffic movement videos at three signalised intersections in Queensland, Australia. The estimated crash frequency from the non-stationary GEV model is compared against the historical crashes for the study locations (serving as ground truth), and the results indicate a close match between the estimated and observed crashes. Notably, the estimated mean crashes lie within the confidence intervals of observed crashes, further demonstrating the accuracy of the extreme value model. At the second level of RTCF, the estimated signal cycle crash risk is fed to a recurrent neural network to predict the crash risk of the subsequent signal cycles. Results reveal that the model can reasonably estimate crash risk for the next 20–25 min. The RTCF framework provides new pathways for proactive safety management at signalised intersections.<p></p> 2024-02-19T00:00:00Z Text Journal contribution 2134/26039038.v1 https://figshare.com/articles/journal_contribution/A_bi-level_framework_for_real-time_crash_risk_forecasting_using_artificial_intelligence-based_video_analytics/26039038 CC BY 4.0
spellingShingle Bi-level
Framework
Crash risk
Fizza Hussain
Yasir Ali
Yuefeng Li
Md. Mazharul Haque
A bi-level framework for real-time crash risk forecasting using artificial intelligence-based video analytics
title A bi-level framework for real-time crash risk forecasting using artificial intelligence-based video analytics
title_full A bi-level framework for real-time crash risk forecasting using artificial intelligence-based video analytics
title_fullStr A bi-level framework for real-time crash risk forecasting using artificial intelligence-based video analytics
title_full_unstemmed A bi-level framework for real-time crash risk forecasting using artificial intelligence-based video analytics
title_short A bi-level framework for real-time crash risk forecasting using artificial intelligence-based video analytics
title_sort bi-level framework for real-time crash risk forecasting using artificial intelligence-based video analytics
topic Bi-level
Framework
Crash risk
url https://hdl.handle.net/2134/26039038.v1

Similar Items