Real-time process surveillance for the Intellect platform
Intellect Anomaly is the self-learning surveillance layer that watches every critical metric—flow, composition, pressures, temperatures, text logs, operator notes—and flags the subtle deviations that precede downtime or off-spec product. It runs wherever Intellect Server runs, so the same detectors you tune on the desktop can watch wells in Alberta, columns in the Gulf, or fabrication lines in Europe without rewriting code.
What Anomaly does
Fuses every data type
- Pulls historian tags, SCADA feeds, SQL data, and free-form text into a unified CVA dataset, using the same data handling templates as the rest of the suite.
- Synchronizes batch and real-time inputs, cleans them, and normalizes units so numeric signals and textual indicators can drive the same anomaly logic.
Designs critical-metric monitors
- Provides a structured workflow to identify the metrics, states, and contextual variables that matter for each asset.
- Lets engineers set threshold logic, statistical detectors, pattern recognizers, and escalation rules tailored to each unit operation.
Runs continuous surveillance
- Executes detectors inside Intellect Server’s task-oriented runtime, so alerts, recommendations, or automated actions can be triggered the moment behavior drifts.
- Supports multi-site deployments; a detector built for one well or train can be replicated “across or down” other assets using the suite’s replication tooling.
Feeds operators the story
- Integrates with Intellect’s web interfaces and reporting templates so anomaly alerts show the offending variables, context, and recommended next steps.
- Works alongside iImprove and Expert outputs—detectors can call optimizers, pause automation, or simply notify staff, depending on what the asset needs.
Engineering workflow
- Build the data foundation in Intellect Expert (CVA setup, cleaning rules, feature engineering).
- Use Anomaly to define detectors, thresholds, and escalation pathways for each process or equipment class.
- Publish detectors to Intellect Server, where they run continuously and can trigger alerts, iImprove optimizations, or human workflows.
- Replicate proven surveillance packages to new assets with minimal reconfiguration.
Benefits
- Earlier warning, fewer surprises – Detect sensor drift, process upsets, or equipment degradation before they hit production.
- Context-rich alerts – Every anomaly event includes the critical metrics, thresholds crossed, and the operational context that matters.
- Consistent surveillance across fleets – One detector design can run across dozens of wells, platforms, or lines, thanks to CVA-standardized data and replication tooling.
- Seamless with the rest of Intellect – Anomaly shares the same deployment, governance, and observability backbone as Expert, iImprove, Designer, and Intellect Server, so there’s no integration tax.