Field Demonstration of Advance Landslide Warning Index for Railroad Tracks on Amtrak’s Harrisburg Line in Pennsylvania
A previous CIAMTIS project, performed jointly by Penn State University and the University of Delaware looked at the use of “Artificial Intelligence for Advance Landslide Warning along Railroad Tracks in Pennsylvania and Delaware”. This activity, which is in the process of being completed this academic year, used video footage from Amtrak’s track geometry inspection car to examine several segments of track along Amtrak’s Harrisburg right of way for approximately 10 out of the total 195-mile route. Approximately 17 videos, taken over a period of 4 years were examined. This data included high landslide risk zones, as defined by Amtrak’s geotechnical engineer and before and after videos of a recent landslide on the line (June 2021). The results of the previous project allowed for the development of an AI based landslide risk index, based on the analysis of approximately 14,000 frames from the right-of-way video. The methodology used these video files together with data from other sources to include railway identified geo-hazards (and attributes) and USGS open-source data. Based on discussions with Amtrak’s geotechnical engineer, there is very serious interest in this methodology, which has the potential for allowing Amtrak to assess landslide risk across its entire range of national operations; approximately 21,000 miles of track. Since it is based on analysis of video images taken from Amtrak’s current track geometry inspection car, it can be readily implemented by Amtrak on all its routes, since Amtrak inspects all of its US and Canadian operating routes with its track geometry inspection car. In order to refine and validate the methodology and to provide Amtrak with this ability, it is proposed that a comprehensive landslide risk evaluation be performed on the entire Harrisburg line, approximately 195 route miles in length (double track the entire way). Amtrak has agreed to provide the additional video data files (it provided the 17 files used in the previous activity) and to work with our team to help determine risk criterion that can be applied. Using this data, it will be possible to further extend the existing analysis to provide a more comprehensive risk index and to determine if landslide precursors can be developed such as: leaning trees, tension cracks, overtopping, debris, slope angle (pitch), slope composition, cut height, etc.
-
Supplemental Notes:
- DEL Fed/Match56250/56082PSUFed/Match56250/56250
Language
- English
Project
- Status: Active
- Funding: $224832
-
Contract Numbers:
69A3551847103
-
Sponsor Organizations:
Office of the Assistant Secretary for Research and Technology
University Transportation Centers Program
Department of Transportation
Washington, DC United States 20590 -
Managing Organizations:
Pennsylvania State University
University Park, PA United States 16802 -
Project Managers:
Donnell, Eric
-
Performing Organizations:
University of Delaware, Newark
College of Engineering
Newark, DE United States 19711Pennsylvania State University, University Park
Thomas D. Larson Pennsylvania Transportation Institute
Research Office Building
University Park, PA United States 16802-4710 -
Principal Investigators:
Zarembski, Allan
Qiu, Tong
Shen, Chaopeng
- Start Date: 20220101
- Expected Completion Date: 20230801
- Actual Completion Date: 0
- USDOT Program: University Transportation Centers
Subject/Index Terms
- TRT Terms: Landslides; Rail (Railroads); Railroad tracks; Warning systems
- Subject Areas: Environment; Railroads; Security and Emergencies;
Filing Info
- Accession Number: 01835077
- Record Type: Research project
- Source Agency: Center for Integrated Asset Management for Multimodal Transportation Infrastructure Systems (CIAMTIS)
- Contract Numbers: 69A3551847103
- Files: UTC, RIP
- Created Date: Jan 31 2022 3:16PM