Cases/CROSS/C17
Cross-Domain SignalsSmart water / gas / energy audit / predictive maintenance

Pipeline Leak Acoustic Detection

01

Pain Points

  • !Gas/water/compressed-air pipeline leaks emit ultrasound (20–40 kHz) inaudible to the human ear.
  • !Long-term monitoring must auto-identify anomalies above the baseline and alert — too much for manual watching.
  • !Leak-intensity trends lack a continuous record.
02

The Tinia Approach

Ultrasonic band-pass + energy level + baseline modeling + Z-Score deviation alerting — anomaly-detection operators applied to pipeline security, rerun automatically over the long term.

Ultrasonic band-passEnergy levelBaseline modelingZ-Score alerting
03

Nodes Used

IIR Filteriir_filter
20–40 kHz ultrasonic band-pass
Sound Level Meterlevel_meter
Energy level
Baseline Statisticsbaseline_stats
Model the normal-state distribution
Z-Score Anomaly Detectionzscore_anomaly
Baseline-deviation alerting

Core nodes; add or remove steps as needed for your data and standards — the node graph is always editable.

04

Expected Output

  • Intensity trend relative to baseline (dB above baseline).
  • An anomaly-event list with alerts.
  • A monitoring workflow that reruns automatically over the long term.
Reference Standards / Engineering PracticeIndustry practice (ultrasonic leak-detection framework)

Want to run this workflow on your own data?

Request a Trial
CROSS

Other cases in this domain