AI-native from the foundation. Grounded in engineering.
SibylCor is being developed as an AI-native platform rather than conventional monitoring software with AI added as a feature layer. The architecture is designed to combine multiple forms of engineering and machine intelligence.
One platform. Five layers of engineering and machine intelligence.
Select a layer to see what it does.
-
Explains the diagnosis, answers operator questions, and turns the result into workflow. Large language models and natural language processing are designed to read maintenance records, work-order history and operating logs to enrich each asset’s model.
-
When the model cannot explain a drift away, an AI agent opens the case automatically—analyzing data, testing hypotheses and identifying root cause.
-
Probabilistic graphical models represent the dependencies among operating conditions, sensor signals and failure modes for each asset. Competing failure hypotheses are tested against live evidence, and the confidence in each is quantified.
-
Machine learning for diagnostics; genetic algorithms to optimize dynamic remaining-useful-life estimates as operating conditions change; probabilistic reasoning, state models and optimization—each applied where it fits.
-
Physics-based equipment models calculate parameters that cannot be directly measured and reconcile them with measured operating data. Built from physics and OEM documentation on day one, then tuned by AI to each machine as it runs.
SibylCor Platform
Dispatches itself, reasons over the evidence, and identifies the failure.
The model tests many possible causes against all available data. Paths that fit the evidence get higher probability.
-
Observed
A relationship drifts that the model cannot explain away. No sensor is in alarm.
-
Reconciled
Evidence that remains normal reduces support for competing faults.
-
Remaining explanation
The cause the evidence still supports becomes the highest-probability explanation, with its confidence.
Your infrastructure is already the sensor.
Through Electric Signature Analysis (ESA), SibylCor is designed to read the voltage and current waveforms your motors already produce, using the current and potential transformers (CTs and PTs) already in place. Combined with vibration analysis and any additional sensors your facility operates, it builds a continuously updating picture of equipment health.
SibylCor combines Electric Signature Analysis with physics-informed probabilistic inference to identify developing failure modes, affected components and estimated remaining operating life.
Starts with your existing sensors. Every signal added sharpens the model.
- Motor current and voltage via CTs and PTs (existing infrastructure)
- Physics-based signal analysis
- Probabilistic graphical model inference
- Failure mode, severity and component
- Remaining useful life, work order, total cost of risk and spare-parts forecast
Electric Signature Analysis
Current and potential transformers capture motor electrical signatures. ESA is designed to identify electrical and mechanical failure modes, including stator winding faults, rotor bar defects, air-gap eccentricity and bearing faults in inverter-driven equipment.
Vibration analysis
Traditional vibration monitoring as a complementary signal layer. Fusing vibration data with ESA is designed to cross-validate diagnostics—reducing false positives and catching failure modes either method alone would miss.
Process & performance signals
Mass and energy balance, fuel flow, power output and efficiency provide operational context that condition monitoring alone cannot—linking equipment health to process performance and identifying degradation that reduces output before it causes failure.
Environmental & thermal sensors
Particulate, thermal, oil-quality and acoustic sensors sharpen the diagnostic picture. Additional streams are ingested through open standards as instrumentation expands.
From operating signals to earlier understanding and planned action.
-
Monitor
Establish how the asset is operating under current conditions.
-
Detect
Identify meaningful deviations from expected operating behavior.
-
Diagnose
Reason across operating data and engineering models to identify likely failure mechanisms.
-
Anticipate
Evaluate condition progression and estimate the remaining intervention window.
-
Act
Translate condition intelligence into a planned maintenance or operating response.
A decision, not an alarm.
Most platforms tell you something changed. SibylCor is designed to tell you which failure mode is developing, in which component, at what severity, and when to act.
-
Failure mode
What is failing, and why.
-
Component
The specific component affected.
-
Confidence
How sure, given the evidence.
-
Evidence
The signals, trends and records behind the conclusion.
-
Prognosis
Health and reliability indices, and a dynamic remaining-useful-life estimate that updates as operating conditions change.
-
Recommended action
Work window, people, parts and work orders.
Deep expertise across rotating machinery.
SibylCor’s domain expertise spans motor-driven and rotating machinery.
Gas turbines
Operating condition, combustion efficiency, and compressor and turbine stage health for power generation and propulsion.
Diesel engines
Cylinder health, injection-system performance and fuel-efficiency diagnostics for propulsion and power generation.
Electric motors
Fixed-speed and variable-frequency-drive motor diagnostics via ESA, including inverter-driven equipment where test points are hard to reach.
Generators & alternators
Standby and primary power-generation assets, supporting condition-based maintenance planning.
Compressors
Centrifugal and reciprocating compressors for upstream and midstream oil and gas, industrial processing and refrigeration.
Chillers & cooling plant
Chillers, cooling towers and air-handling units, including data-center cooling infrastructure.
Propulsion systems
Commercial propulsion and coupled drive systems, with equipment and operating-context analytics.
Domain specialization
Motor-pump intelligence
SibylCor’s core domain expertise is in motor-pump systems. The platform is designed to deliver diagnostics and prognostics across pump types—from small process pumps to large centrifugal and positive-displacement systems.
Discuss motor-pump reliability- Centrifugal
- Positive displacement
- Submersible
- Vertical turbine
- Gear
- Diaphragm
- Axial flow
- Screw
- Lobe
- Cooling water
- Fuel transfer
- Hydraulic
Works with what you already have.
-
In
The data already in place
Sensors, data acquisition, historians, controls and maintenance records. No proprietary hardware.
-
SibylCor
AI-native diagnosis, prognosis and maintenance planning
Deployed inside your network.
-
Out
Delivered into your maintenance system
Through SibylCor’s own workflows or integration with the site’s CMMS, EAM and ERP systems.
Value is tied to understanding the equipment—not selling a proprietary sensor or OEM hardware estate.
Maintenance management
Designed to act as the diagnostic intelligence source for your CMMS—generating failure-mode-specific work orders, task lists and part requirements in your existing workflow.
Instrumentation & sensor networks
Connects to existing instrumentation—CTs, PTs, vibration transducers, temperature sensors and flow meters. Higher sensor density improves diagnostic precision.
Alerts & notifications
Context-rich alerts, not threshold breaches: each notification carries the failure mode, affected component, severity and recommended action.
Open standards
Data and recommendations available through a REST API for dashboards, BI tools, ERP systems and operational applications, without proprietary lock-in.
Supported data acquisition standards
- NI DAQ
- OPC UA / DA
- RS-485
- Modbus / Modbus TCP
- AVEVA PI
- PI Data Framework
- PI Asset Framework
- REST API
If your system uses a standard protocol, SibylCor is designed to connect to it.
What does an unplanned failure actually cost?
SibylCor is designed to compute the Total Cost of Risk for every monitored asset—not just maintenance cost.
Process unavailability
Lost productive capacity while the asset is down.
Degraded efficiency
Output and efficiency lost as a fault develops.
Emergency procurement
Premium cost of parts and services sourced in a hurry.
Missed operating windows
Planned windows that pass before the problem is understood.
A competent ally, not just a chat overlay.
It dispatches itself
The model cannot reason the drift away, so an AI agent opens the case automatically.
It keeps you in the loop
“How sure are you?” returns the confidence and the evidence—not another alert.
You stay in command
The operator chooses the window. SibylCor queues the work and notifies Engineering.
It explains the decision, shows the evidence, and leaves the decision where it belongs.
Better decisions start with understanding the condition earlier.
If you are responsible for the reliability, performance, or lifecycle of critical rotating equipment, we would like to hear about the problems you are working to solve.
Or email info@SibylCor.com
Start with one critical asset class
One critical asset class, one narrow engagement—starting where the consequence of failure is already understood, with a low-risk path to validate SibylCor.