Learning Transportation Insecurity from Location Intelligence Data

Details

• Creators:
  Nie, Yu (Marco) ; Chen, Ying ; Murphy, Alexandra K. ; Dai, Tianxing ; Yan, Xiaoyu ; Kivestu, Peeter
• Corporate Creators:
  Northwestern University (Evanston, Ill.) ; University of Michigan, Ann Arbor
• Corporate Contributors:
  Center for Connected and Automated Transportation (CCAT) ; 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:
  Classification Computer Security Data Communications Evaluation Feasibility Analysis Intelligence Mobile Applications Mobile Phone Data Mobile Telephones Monitoring Statistical Inference Surveys Trajectory Transfer Learning Transportation Equity Transportation Insecurity Transportation Safety

  More +
• Publication/ Report Number:
  CCAT-NU-2026-1
• Resource Type:
  Tech Report
• Geographical Coverage:
  United States ; Chicago (Illinois)
• Edition:
  Final Report (July 2024 – January 2026)
• Corporate Publisher:
  University of Michigan. Center for Connected and Automated Transportation
• Abstract:
  Transportation insecurity (TI)— a condition in which one is unable to regularly move from place to place in a safe or timely manner because one lacks the material, economic, or social resources necessary for transportation—remains a persistent barrier to economic opportunity and quality of life in the United States. While the Transportation Security Index (TSI) provides a validated measure of TI, its reliance on stand-alone surveys limits its scalability, temporal coverage, and geographic comparability. This report proposes a scalable alternative for measuring TI by leveraging large-scale location intelligence data and transfer learning methods. Building on evidence that mode use patterns are strongly associated with TI status, we develop a framework that infers TI from passively collected mobility trajectories. The approach first employs a semi-supervised machine learning pipeline to identify driving, transit, and active travel modes from sparse and unlabeled mobile phone data, integrating trip-level features with mode-specific travel characteristics from the Google Maps API. We then apply a TI classification model trained on survey data from the Detroit Metro Area Communities Study to location intelligence data from Chicago. Despite relying on simplified mode categories and census tract-level socio-demographic inputs, the model achieves approximately 70% accuracy and recall. While cross-region transfer introduces some bias, the resulting spatial patterns of inferred transportation insecurity align with well-documented areas of mobility disadvantage. These findings demonstrate the feasibility and promise of using location intelligence data to measure transportation insecurity at scale, enabling more systematic monitoring and evaluation of transportation equity outcomes.
• Format:
  PDF
• Funding:
  69A3552348305
• Collection(s):
  University Transportation Centers
• Main Document Checksum:
  urn:sha-512:c0610190db74720b2010e69d9ae0aa62846e0a919d5546faced9f1e8074dd6e64e7d491647a5552c55ddce659a48936b06b5d45f9be54bae09c7f79e380d2809
• Download URL:
  https://rosap.ntl.bts.gov/view/dot/89262/dot_89262_DS1.pdf
• File Type:
  [PDF - 1.50 MB ]