Earlier detection.
Evidence-based classification.
Auditable for every call.
Deterministic classification of fluid losses, packoff, wellbore instability, and geometry restriction. Validated against a documented test suite. The supporting evidence ships with every call.
One platform. Six outcomes.
From the moment a sensor reading arrives to the moment a drilling engineer files the post-well report — Sentinel keeps your operation visible, classified, and audit-ready.
Detect risk earlier
Evidence-based classification of the four developing risks that drive NPT.
→Single-well operations
Live View dashboard built for the company man and the drilling engineer.
→Fleet-wide oversight
Multi-rig overview for operations managers and superintendents.
→Tune alerts per user
Alert routing per person. Hold windows, escalation chains, and false-positive flags — all logged.
→Reconstruct every run
Step back through any minute. Export the whole record for audit.
→Integrates with your data feeds
Compatible with existing rig-site telemetry systems.
→Five surfaces. One source of truth.
Every value is auditable. Every classification carries a structured trace. Every minute is preserved.
One well, in real time.
The Live View is what an operator sees during an active drilling operation. A primary risk card on the left, a channels grid on the right, a 60-minute timeline rail at the top, and the Why this call panel below it all.
- Risk heroPrimary classification, severity, confidence, trend, sparkline.
- ChannelsOperator-configurable; SPP, torque, ROP, hookload, flow, ECD, pit active.
- Why this callNarrative on the left. Structured classification record on the right.
- Timeline60-min review timeline with severity-colored event markers.
The whole wellbore, in 3D.
A real-time WebGL scene of the well — derrick at surface, formation slabs by casing interval, casing tube along the trajectory, drillstring split into drilled (solid) and planned (semi-transparent), bit halo at the current MD. Auto-orbits when idle; drag to rotate. Same scene the engineering team sees in Setup.
- TrajectoryYour well through every survey station, with average-angle integration. Smoothed for display.
- Casing programSurface · intermediate · production · liner — each with its own tone and shoe marker.
- Drilled vs plannedLive drillstring split at the current MD — drilled solid, planned semi-transparent.
- InteractiveDrag to rotate. Auto-orbit when idle. No plug-in needed.
Your whole program, on one screen.
Built for the operations manager who needs to know which wells are on track, which need attention, and how the model is performing across the program. Four KPIs at the top. The Active Wells table below. The Fleet Risk Heatmap at the foot.
- KPI stripTotal Wells · Drilling · High Risk · At Warning Level.
- Active wellsPer-rig drilling status, depth context, current risk and severity.
- Heatmap48 thirty-minute buckets across up to six wells.
- Trust metricsLead time and false-positive rate visible per rig.
Every event recorded. Only the right person interrupted.
Sentinel separates the act of detecting an event from the act of waking someone up. Every event is recorded the moment it happens. Whether someone gets a call is a second decision — made per person, checked against the alert's live state when their hold window expires.
- Hold windowSet your own. If the event resolves before it expires, no one wakes you. The event is recorded either way.
- EscalationIf no one acknowledges in time, the call escalates up the chain — regardless of personal preference.
- FeedbackDrilling engineers flag false alarms. The false-positive rate per rig is visible to ops managers.
- AuditEvery callout decision is recorded — who, when, why, and whether the call was made or held.
Every minute of every run, preserved for review.
Step back to any minute and the platform reproduces exactly what you saw at that moment — risk score, channels, severity, confidence, evidence, and the natural-language rationale. Every run produces a single sealed evidence bundle.
- Run timelinePhase bars, daily risk band, event markers, scrubbing playhead.
- SnapshotLive View as it was, paired with the classification record.
- BundleTelemetry · signals · audit trail · alerts · validation · sealed bundle.
- Audit-readyOne downloadable artefact for review, regulator, or model-governance.
Every drilling event in one of five families.
Most drilling-monitoring tools fire when a sensor crosses a threshold. They tell you a number is high, not what is happening. Sentinel classifies the mechanism — the root cause that produces the signal pattern.
Normal
Baseline operation. No mechanism active.
Fluid Losses
LCM and ECD adjustment.
Packoff
Circulation and spot pills.
Wellbore Instability
Mud-weight tuning, exposure-time control.
Geometry Restriction
Reciprocation and rotation.
How it works.
Sentinel analyzes over 70 channels of rig telemetry, sourced from standard industry feeds. Classification runs every minute. Your teams act on developing risk earlier — and every decision is preserved for review.
Your data, your boundary
Your data stays in a tenant boundary your team controls. Only your nominated users can see it. Full security pack on request.
Audit-grade evidence
Every call carries an audit trail. Every run is exportable as a single sealed bundle. Same input produces the same answer, every time.
What you connect
Cloud-hosted on AWS. No rig-side installation. Ingests directly from existing WITSML, WITS Level 0, and OSDU streams.
Who we are
UK company, registered in England and Wales. Cyber Essentials Certified. Hosted on AWS. AWS Activate Member. Vendor due-diligence pack on request.
Designed around how drilling teams operate.
Sentinel does not assume one screen and one user. Three distinct roles, each with a different physical location, a different connectivity profile, and a different idea of what useful looks like.
Company Man
- Live View on the rig-floor screen
- Severity-coded primary classification
- Channels grid with current values and 12-min trend
- Why-this-call rationale in plain English
Drilling Engineer
- Personal hold windows for callouts
- Acknowledgement and false-positive flagging
- Full audit trail for every alert
- Post-well review with sealed evidence bundles
Operations Manager
- Fleet Overview KPI strip
- Heatmap of risk score across up to 6 wells
- Lead-time and false-positive rate per rig
- Independent escalation chain
Built around what drilling engineers ask for.
Drilling Sentinel Ltd is a UK-registered company building real-time risk classification for drilling operations. We separate detection from delivery, classify the mechanism (not the threshold), and publish the supporting evidence with every call we make.
Data review
Historical wells reviewed. Performance baseline established. Success metrics agreed in writing.
Shadow mode
Sentinel runs alongside existing monitoring. Alerts visible to the Sentinel team only. Zero operational change.
Pilot well
Live deployment on one well. Embedded engineer. Lead-time tracking. Phase-end FP review.
Expansion
Fleet rollout. Calibrated thresholds. Dedicated support. Quarterly performance review.
Initiate a pilot enquiry.
Drilling Sentinel produces lead-time numbers from your first well, measured against your own drilling reports. Pilot enquiries, commercial discussions, and technical diligence all begin at the same address.