This repository contains the source files for the AsphaltMine Database web application. AsphaltMine is an Asphalt Test Result Database for Performance Prediction. It serves as a platform for collecting asphalt test results in an organized online database and for using this data to train machine learning models to predict asphalt performance.
The web application is built upon the NIST MDCS project. Significant modifications include:
- Implementation of a unique form for AsphaltMine called the GV Web Form. This form is capable of saving, downloading, loading, and viewing single and multiple test records.
- Adding a comprehensive tutorial explaining various aspects of AsphaltMine and how to use the platform.
- Addition of a performance prediction module, using trained ML models to predict asphalt performance from mix design inputs (currently including Marshall mix design, with more models planned).
- Addition of a Live Insights visualization page, providing aggregated statistics and histograms over curated records (per-workspace or sitewide).
- Addition of a Bulk Upload tool, allowing multiple test records to be extracted, validated, and saved from AsphaltMine's predefined Excel template.
- Addition of an Excel export option, using AsphaltMine's predefined Excel template, to the existing record download functionalities.
- Several major and minor modifications for compatibility across libraries.
- Removal of certain aspects of the original NIST MDCS project, including the curate, query, and form pages, along with several other features.
These modifications have been ongoing since June 2025; for the detailed nature and dates of individual changes, see this repository's commit history.
For the complete original NIST MDCS project, see ORIGINAL_README.md and visit the GitHub link: https://github.com/usnistgov/mdcs
- Original NIST License: NIST_LICENSE.md
- Third Party License: THIRD_PARTY_LICENSES.md