EnergyMindAI / Drilling intelligence

From a drilling report to a reviewed finding

1. Bring the evidence. Import CSV, XLSX or structured text PDFs. Map source columns and timezones, correct errors and approve the import preview.
2. Account for the time. Separate productive time, planned non-drilling and unplanned NPT. Resolve overlapping intervals before comparing wells or calculating exposure.
3. Find the pattern. Inspect category Pareto, campaign trends and event details. Keep denominator, cost rate and data coverage visible.
4. Review the explanation. Advisory root-cause suggestions retain evidence and require a human decision. Record investigation actions and export the reviewed summary.

Eight connected engineering workspaces

• NPT Intelligence: reviewed events, rig time, configurable costs and reports.
• Offset-Well Intelligence: historical comparison, sourced parameters and lessons.
• Rig Readiness: evidence, owners, certification expiry and critical blockers.
• Stuck Pipe Guardian: read-only telemetry ingestion, replay and advisory alert episodes.
• Drilling Hydraulics: ECD and annular-velocity screening with stated assumptions.
• Drilling Performance: ROP and MSE screening from sourced inputs.
• Equipment Reliability: maintenance evidence, failures and observed reliability metrics.
• Engineering Copilot: engineering-document answers with exact source citations.

Post-well analytics and rig-floor alerts

Historical NPT analysis and telemetry alerts have separate workflows. Alert rules check rig state, source quality, freshness and sequence; persistence, hysteresis and cooldowns help distinguish sustained excursions from brief noise. Each episode retains the source observations and rule configuration for engineering review.

A signal excursion is an advisory observation. It does not automatically establish a stuck-pipe diagnosis or create an NPT event. A physical rig feed and site-approved limits must be commissioned before field use. DrillAstra has no automated rig-control or well-control outputs.

Your first NPT investigation

1. Create a campaign, register its rig and add the well. Set the campaign cost rate and reporting denominator.
2. Upload a DDR and verify timestamps, units, event boundaries and planned/unplanned mapping.
3. Correct or exclude invalid rows, save the review and approve the import.
4. Open the largest Pareto category. Check whether it contains one long event or many repeat events.
5. Inspect each event’s source evidence. Review classifications and document the rationale.
6. Generate an advisory root-cause investigation. Separate verified observations from proposed mechanisms and missing evidence.
7. Compare offsets with similar sections and operating context; examine readiness and equipment evidence.
8. Export the summary with cost assumptions and coverage. Assign follow-up ownership in your team’s action register and verify whether recurrence changes over later wells.

Reading NPT patterns responsibly

A large cost category: establish whether event duration, frequency or the configured rate explains its size. Estimated exposure is not verified financial loss.

A rising NPT percentage: inspect logged coverage and denominator first. Missing productive hours can make the percentage rise without a new operational failure.

Repeat events across wells: compare section, equipment, operation and source quality before treating a common label as a common mechanism.

A torque or pressure excursion: inspect the actual signal, rig state and gaps. A rule or statistical baseline detects departures; it does not establish their cause or a calibrated probability.

The application includes a 15-step walkthrough and an interactive, explicitly synthetic accounting example. Demo values are illustrative and are never represented as measured field data. Optional AI assistance requires engineering review.