Drop a screenshot here:assets/screenshots/frac-hit-mitigation.png
An ML pipeline that detects frac hits between neighboring oil wells in real time, reaching over 90% classification accuracy.
Drop a screenshot here:assets/screenshots/frac-hit-mitigation.png
Overview
A frac hit occurs when hydraulic fracturing at one well disrupts a neighboring well, an event that matters for both safety and cost. This project combines Hidden Markov Models with XGBoost to detect and classify frac hits from time-series sensor data, reaching over 90% classification accuracy.
The pipeline is deployed as a full-stack system, with a Flask API serving predictions and a React dashboard for real-time visualization and monitoring.
Built with