Predict screenouts before they happen
ScreenOut uses a large, labeled dataset of real screenout events to build a prediction model that gives your frac crew advance warning — turning reactive shutdowns into proactive adjustments.
Features
Key Capabilities
Screenout Prediction Model
A machine learning model trained on thousands of labeled screenout events, learning the pressure signatures that precede failures so your crew can act before they happen.
Early Warning Alerts
Get advance notice when pressure patterns indicate a developing screenout — giving your team time to adjust rates, concentrations, or fluid chemistry before the event.
High-Resolution Pressure Data
40,000 samples per second at 0.002 PSI accuracy captures the subtle pressure changes that precede screenouts — detail that traditional gauges miss entirely.
FracScore Integration
Combines with FracScore stage diagnostics to correlate screenout risk with cluster performance, stage difficulty, and completion design parameters.
Building on the Largest Labeled Screenout Dataset in Completions
We've labeled thousands of real screenout events from our high-resolution pressure data. This dataset is now training a prediction model designed to catch developing screenouts before they force a shutdown — giving frac crews the lead time to adjust and keep pumping.
Roadmap
From Detection to Prediction
ScreenOut is progressing from a labeled dataset to a deployed prediction model — each phase building on proven data.
Data Collection & Labeling
Thousands of screenout events identified and labeled from high-resolution pressure data across multiple basins and completion designs.
Model Training
Machine learning models training on labeled events to recognize the pressure signatures that precede screenouts — learning patterns invisible to human analysis.
Real-Time Deployment
Deploy the trained model alongside live FracScore data streams, delivering early warning alerts directly to frac crews during operations.
Get Early Access to ScreenOut
ScreenOut is currently in beta. Join the early access program to test prediction capabilities on your well data.