Predictive Analytics
Predict-It
Predict-It monitors critical equipment in real-time process environments and uses advanced predictive computational methods to recognize potential future outages
Unleashing Predictive Analytics: Unlocking Future Insights
Enhance Equipment Reliability with Real-Time Predictive Monitoring by Predict-It
Read more about Predict-It?
This suite comprises four robust modules that enable users to identify, diagnose, and manage developing equipment malfunctions at the earliest possible stage. These modules also support troubleshooting and management of equipment recovery plans.
See how AGL reaped the benefits of ECG's Predictive Analytics platform, Predict-It
Predict-It Anomaly Detection
Dual detection capability for complete asset coverage
Predict-It detects equipment issues in all modes of operation, providing coverage from start-
up, through normal operation, and when shutting down. Detection of developing equipment
issues are flagged weeks to months ahead of traditional monitoring techniques, providing
time to plan corrective actions on your schedule.
Predict-It’s Advanced Pattern Recognition (APR) module uses robust algorithms that track
trends in process variables on a continuous basis and compare them to historical operation.
This allows for real-time monitoring that gives O&M teams a look at the operation of
machines during continuous processes.
Start-ups and shut-downs account for much of the wear and tear that a machine will undergo
in its lifecycle. Predict-It’s Pathway module introduces a new way to monitor your equipment
during the most vulnerable stages of operation using multivariable analysis techniques.
Awareness of changes in key variables can make all the difference in optimizing operations
and getting more productive hours out of your assets.
APR Module
- Continual Operation – long runs
- Data snapshot every 5-15 min.
- Enhanced Similarity Based Modeling
(SBM) - Snapshot of expected vs. current value
- DBSCAN outlier filtering
Pathway Module
(Batch Processes)
- Batch Processes – short runs
- Data snapshot every 1-10 sec.
- Univariate and Multivariate Models (MVDA)
- Overlay of process maturity events
- Batch model support
- Dynamic time warping
Predict-It Diagnostic Reasoner
A digital doctor for your process equipment
Predict-It’s Diagnostic Reasoner is a dynamic diagnostic engine that
capitalizes on Causal Asset Networks (CANs). These networks are
designed for the specific make and model of the assets they evaluate
— resulting in more precise and interactive diagnostic results. CANs
power the Diagnostic Reasoner, suggesting multiple root causes for
failures and assigning exact probabilities for each outcome.
Anomalies from both APR and Pathway feed into the Diagnostic
Reasoner to provide evidence and start the analysis process.
Subject Matter Experts (SME) can add knowledge of typical asset
behavior, which is captured by Predict-It and applied across a fleet
of assets. Real-time observed symptoms from tech exam reports
can be added to the diagnostic engine to facilitate a higher level of
certainty. Symptoms derived from multiple sources and probabilistic
calculations allow pinpoint accuracy in Predict-It’s diagnostics, taking
the guesswork out of apparent cause fault identification.
Diagnostic Reasoner Benefits
- Tailored Asset Diagnostics
- Intuitive Fault Troubleshooting Logic
- Backed by Bayesian Open Architecture
- Configured Diagnostics Available from ECG
- SME Input Support
- Fault Probabilistic Ranking
Learn more on how automating asset diagnosis can add value to your maintenance practices
Product Video
Product Support
Access guides, manuals, software downloads, training videos, and more in our product support center.