Holistically Identifying Road Complexity and Relating it to Fatal Crashes [supporting dataset]

Details

• Creators:
  Wang, Meng ; Roberts, Shannon C
• Corporate Creators:
  New England University Transportation Center ; University of Massachusetts at Amherst
• Corporate Contributors:
  United States. Department of Transportation. Office of the Assistant Secretary for Research and Technology ; United States. Department of Transportation. University Transportation Centers (UTC) Program
• Subject/TRT Terms:
  Artificial Intelligence Automatic Data Collection Systems Data Files Evaluation Fatalities Large Language Models Naturalistic Driving Perception Roadway Complexity Scene Perception Evaluation Traffic Lanes

  More +
• DOI:
  https://doi.org/10.7910/DVN/ZW3V1D
• Resource Type:
  Dataset
• Right Statement:
  Open Access; https://creativecommons.org/publicdomain/zero/1.0/
• Geographical Coverage:
  United States
• Corporate Publisher:
  New England University Transportation Center
• Abstract:
  Understanding the context of crash occurrence in complex driving environments is essential for improving traffic safety and advancing automated driving. Previous studies have used statistical models and deep learning to predict crashes based on semantic, contextual, or vehicle kinematic features, but none have examined the combined influence of these factors. In this study, we term the integration of these features ``roadway complexity''. This paper introduces a two-stage framework that integrates roadway complexity features for crash prediction. In the first stage, an encoder extracts hidden contextual information from these features, generating latent complexity features. The second stage uses both original and latent complexity features to predict crash likelihood, achieving an accuracy of 87.98% with original features alone and 90.46% with the added latent complexity features. Ablation studies confirm that a combination of semantic, kinematic, and contextual features yields the best results, which emphasize their role in capturing roadway complexity. Additionally, complexity index annotations generated by the Large Language Model outperform those by Amazon Mechanical Turk, highlighting the potential of AI-based tools for accurate, scalable crash prediction systems.
• Content Notes:
  This item is made available under the terms of the Creative Commons Attribution 1.0 Universal (CC0 1.0) license https://creativecommons.org/publicdomain/zero/1.0/. Use the following citation: Shannon Roberts, 2025, "Holistically Identifying Road Complexity and Relating it to Fatal Crashes", https://doi.org/10.7910/DVN/ZW3V1D, Harvard Dataverse, V1, UNF:6:7oLQrrF/emtvw/obCpA5Ew== [fileUNF]
• Format:
  CSV
• Funding:
  69A3552348301
• Collection(s):
  University Transportation Centers
• Main Document Checksum:
  urn:sha-512:f2b967615e2fb992436b931f3d5a17e3631ef53993d8dfa40a30b99be330a75dad8bf9d9f5eb8738e27c21d03669cd625fc6af7ad072d9533be4fea466ae4a1d
• Download URL:
  https://rosap.ntl.bts.gov/view/dot/86899/dot_86899_DS1.pdf
• File Type:
  [PDF - 128.02 KB ]