Machine Learning for Slope Monitoring

Remote sensing technologies (e.g. lidar and photogrammetry) have demonstrated applications for slope monitoring and risk assessments. Application of remote sensing technologies are used in the context of asset management and are supplementing traditional visual inspections. Many agencies with slope assets have begun to invest in the development of remote-sensing-based monitoring programs, including Colorado Department of Transportation (CDOT). As the application of remote sensing technologies to slope monitoring has grown, significant efforts have been devoted to improving and automating different parts of the data processing pipeline necessary to convert raw data into useable results (e.g. Lague et al., 2013; Bonneau et al., 2019; Kromer et al., 2019; Schovanec et al., 2021). However, most automated processing pipelines end at the point of producing a “change map”, which shows areas where the 3D point clouds obtained for different time periods differ from one another. This change map must then be manually interpreted to identify where actual notable slope changes of practical interest have occurred. For example, BGC Engineering is currently monitoring a rock slope in Manitou Springs with a fully automated data collection and processing pipeline that requires an engineer to manually check automatically generated change maps to develop a report of where rockfalls have occurred on the slope. As the amount of data collected continues to grow, the automation of the interpretation of change maps represents a critical opportunity to unlock the full potential of remote-sensing-based monitoring. Given that this is a data rich problem and no obvious direct algorithmic approaches exist, machine learning represents a natural avenue to explore.

    Language

    • English

    Project

    • Status: Programmed
    • Funding: $150000
    • Sponsor Organizations:

      Colorado Department of Transportation

      Applied Research and Innovation Branch
      Denver, CO  United States  80204
    • Managing Organizations:

      Colorado Department of Transportation

      Applied Research and Innovation Branch
      Denver, CO  United States  80204
    • Project Managers:

      Tran, Thien

    • Performing Organizations:

      Colorado School of Mines

      1500 Illinois St
      Golden, CO  United States  80401
    • Principal Investigators:

      Walton, Gabriel

    • Start Date: 20250501
    • Expected Completion Date: 0
    • Actual Completion Date: 0

    Subject/Index Terms

    Filing Info

    • Accession Number: 01930280
    • Record Type: Research project
    • Source Agency: Colorado Department of Transportation
    • Files: RIP, STATEDOT
    • Created Date: Sep 16 2024 8:26AM