Incident Duration

Predicting the duration of traffic incidents is a crucial area of research aiming to improve traffic management and commuter experience. Current research heavily utilizes machine learning, particularly gradient boosting machines (like XGBoost and LightGBM) and deep learning models (including transformers and recurrent neural networks), often employing a bi-level approach combining classification (short vs. long duration) with regression (precise duration estimation). Key factors influencing prediction accuracy include incident type, location, weather conditions, and road network characteristics. Accurate incident duration prediction enables proactive resource allocation, optimized route planning, and ultimately, reduced congestion and improved safety.

Papers

June 27, 2024

Predicting the duration of traffic incidents for Sydney greater metropolitan area using machine learning methods
Artur Grigorev, Sajjad Shafiei, Hanna Grzybowska, Adriana-Simona Mihaita
Machine Learning Traffic Safety Long Duration Duration Prediction Incident Duration Incident Duration Prediction

November 1, 2023

A Bi-level Framework for Traffic Accident Duration Prediction: Leveraging Weather and Road Condition Data within a Practical Optimum Pipeline
Rafat Tabassum Sukonna, Soham Irtiza Swapnil
Efficient Pipeline Traffic Accident Road Network Data Incident Duration

April 23, 2023

Machine learning framework for end-to-end implementation of Incident duration prediction
Smrithi Ajit, Varsha R Mouli, Skylar Knickerbocker, Jonathan S. Wood
End to End Machine Learning Framework Incident Report Incident Duration Incident Duration Prediction

March 21, 2023

DG-Trans: Dual-level Graph Transformer for Spatiotemporal Incident Impact Prediction on Traffic Networks
Yanshen Sun, Kaiqun Fu, Chang-Tien Lu
Graph Transformer Temporal Dependency Temporal Transformer Traffic Network Dg Tta Traffic Incident Incident Duration

September 19, 2022

Traffic incident duration prediction via a deep learning framework for text description encoding
Artur Grigorev, Adriana-Simona Mihaita, Khaled Saleh, Massimo Piccardi
Deep Learning Framework Traffic Accident Text Description Traffic Incident Incident Duration Incident Duration Prediction

May 10, 2022

Incident duration prediction using a bi-level machine learning framework with outlier removal and intra-extra joint optimisation
Artur Grigorev, Adriana-Simona Mihaita, Seunghyeon Lee, Fang Chen
Joint Optimization Traffic Congestion Incident Duration Incident Duration Prediction

Incident Duration

Papers

Predicting the duration of traffic incidents for Sydney greater metropolitan area using machine learning methods

A Bi-level Framework for Traffic Accident Duration Prediction: Leveraging Weather and Road Condition Data within a Practical Optimum Pipeline

Machine learning framework for end-to-end implementation of Incident duration prediction

DG-Trans: Dual-level Graph Transformer for Spatiotemporal Incident Impact Prediction on Traffic Networks

Traffic incident duration prediction via a deep learning framework for text description encoding

Incident duration prediction using a bi-level machine learning framework with outlier removal and intra-extra joint optimisation