Skip to main content
ICTD 2025 Main banner
Final Program


Icon Legend
This session is not in your schedule.
This session is in your schedule. Click again to remove it.
Back

ICTD 2025

Freight & Logistics

POSTER RECEPTION (View list of posters here)

AI-Enabled Data Disaggregation for Freight Analysis Framework Commodity Flow

Tuesday, June 10, 2025
4:45 PM – 6:00 PM MT

    Primary Author(s)

  • SK

    Sakib Khan, PhD

    Intelligent Transportation Systems Principal
    MITRE
    Bristow, VA, United States

    • Co-Author(s)

  • SK

    Sakib Khan, PhD

    Intelligent Transportation Systems Principal
    MITRE
    Bristow, VA, United States

    • Presenter(s)

  • SK

    Sakib Khan, PhD

    Intelligent Transportation Systems Principal
    MITRE
    Bristow, VA, United States

    • Co-Author(s)

  • KC

    Keith Campbell, n/a

    Data Scientist, Principal
    MITRE
    Mclean, Virginia, United States

    • Presenter(s)

  • KC

    Keith Campbell, n/a

    Data Scientist, Principal
    MITRE
    Mclean, Virginia, United States

    • Co-Author(s)

  • LA

    Laurence Audenaerd, n/a

    Multi-Discipline Sys Eng,Lead
    MITRE
    Mclean, Virginia, United States

  • Billy Baden, n/a

    Applied Operations Research, Lead
    MITRE
    Mclean, Virginia, United States

  • DB

    David Bodoh, n/a

    Software Engineering, Lead
    MITRE Corporation
    Mclean, Virginia, United States

    • Submitter(s)

  • SK

    Sakib Khan, PhD

    Intelligent Transportation Systems Principal
    MITRE
    Bristow, VA, United States

Abstract: In the U.S., understanding commodity flows across different modal routes, such as rail and maritime, has long been challenged by the lack of granular, mode-specific data. This research utilizes AI-driven disaggregation techniques to overcome this limitation and infer detailed county-level commodity movements. The process begins with the known socioeconomic and commodity data at the Freight Analysis Framework (FAF) zone level, and AI-based regression models are applied to develop relationships between these data points and commodity movements. Once established, these models are extended to the county level, allowing for a more precise analysis of multimodal flows across the national commodity network. Additionally, infrastructure datasets, including ports, rail nodes, coal plants, and stone quarries, are integrated to improve the accuracy and reliability of the outputs. This comprehensive approach offers new insights into the distribution of commodity flows and provides a framework for addressing gaps in multimodal transportation data. The presentation will delve into key findings, discuss the methodological challenges, and share critical lessons learned, highlighting how AI methods can facilitate the data disaggregation of rail and maritime commodity flows across the U.S. transportation landscape. This expanded model will serve as a tool for policymakers to see the impact of any large-scale disruptions on the national multimodal freight movement and anticipate future transportation demands.

Learning Objectives:

Attendees can expect to learn the following from this session:
  • Upon completion, participants will understand the applicability and gaps associated with the US commodity flow movement.
  • Upon completion, participants will be able to understand how data driven model can help data disaggregation.
  • Upon completion, participants will understand the methodological challenges, and share critical lessons learned with the publicly available US rail and commodity flow data.