Mission:

Build a Repository of Subsurface Investigation Data from Publicly Funded Sources and Derive Valuable Geotechnical Engineering Tools for The Mutual Benefit of The Data Owners and The Geotechnical Engineering Profession.


Vision:

To extract the maximum value and clarity from historic geotechnical data in an open format and save our users time and money by using cutting edge technology on aggregated geotechnical data.


Non-Profit:

Geosetta is a registered Non-profit company, based in Maryland. We are committed to building a shared resource for the geotechnical engineering profession. Geosetta is providing this data in an open format to ensure sharing and collaboration. This nonprofit structure ensures that we develop this tool such that both the data owners and users can benefit. Geosetta needs your support to remain an independent steward of this valuable data resource.


What is Geosetta?

  • The Geosetta name gets its inspiration from the concept of a Rosetta stone for Geotechnical data.
  • Provides a platform for hosting subsurface investigation/geotechnical data from various publicly funded sources throughout the United States.
  • Provides a preliminary understanding of anticipated subsurface conditions.
  • Geosetta developed geospatial and visualization tools, with machine learning techniques applied
  • Geosetta is NOT a substitute for site-specific subsurface investigation

In partnership with public agencies and private entities, Geosetta, Inc. is creating a “living” geotechnical database that will be routinely updated with additional data as it becomes available. This will involve running customized routines on large sets of files, including csv, gINT, database snapshots, and other formats. Geosetta only extracts data that is needed both for machine learning (predictive models) and for the general geotechnical user for planning purposes.


A historic data point or predicted model will typically include test boring information such as location coordinates; SPT data including depth/elevation, blow-count for each interval, recovery; major soil types; moisture content; rock type, core recovery, and RQD; and groundwater levels.


Using Geosetta’s predictive models, the user will gain an initial understanding of expected subsurface conditions that will help in planning an optimized, focused geotechnical exploration program. Thus, achieving our mission of saving the practitioners’ time and money.



Research & Publications

Geosetta was founded on research conducted for the Maryland State Highway Administration. Below are the foundational reports and subsequent publications that have utilized or contributed to Geosetta's development:

Foundational Research

Machine Learning Techniques for SPT Based Geotechnical Subsurface Modeling

Maryland State Highway Administration (2021)

Report No. MD-21-SHA/UM/5-23

This groundbreaking report established the machine learning methodology that forms the core of Geosetta's predictive capabilities, demonstrating how neural networks can effectively predict subsurface conditions from SPT data.

View Report
Developing a GIS-Based Platform for Managing Boring Log Requests and Geotechnical Data

Maryland State Highway Administration (2018)

Report No. MD-18-SHA/UM/4-52

This report laid the foundation for Geosetta's platform architecture, outlining the need for and design of a centralized system for managing and sharing geotechnical data across agencies.

View Report

Publications Using Geosetta

A Methodology for Comparison of Algorithm-Based Subsurface Predictions with Geotechnical and Geophysical Data

Ghimire, A., Yost, K.M., Ph.D., M.ASCE, Cutts, R., M.ASCE, and Zhu, T., Ph.D.

Pennsylvania State University

This study validates Geosetta's AI-based predictions by comparing them with high-quality geophysical data from Multichannel Analysis of Surface Waves (MASW) testing and traditional geotechnical boring data collected at Pennsylvania State University. The research investigates methods for comparing predictions including direct strata delineation comparison and Site Class computation based on VS30.

Citing Geosetta

If you use Geosetta in your research or professional work, please cite it as:

Geosetta, Inc. (2024). Geosetta: A Comprehensive Geotechnical Database Platform. Available at: https://geosetta.org

For specific datasets, please include the data source (e.g., VDOT, MDOT, MnDOT) and access date in your citation.


Have you published research using Geosetta? We'd love to feature your work! Please contact us at research@geosetta.org with your publication details.