OUR APPROACH
Precision
over blanket intervention
Prevention
over reaction
Longevity
over replacement
Early acoustic change
precedes failure
Crack development and emerging leaks change acoustic behaviour before catastrophic rupture.
Adelitics identifies and classifies these early changes to enable structured leak-before-break decision-making.
This is leak before break: operationalised
Sensing-agnostic by design
Our analytics interpret data from acoustic loggers, distributed fibre-optic sensing, and hybrid monitoring systems.
The platform is vendor-agnostic, ingesting and analysing data from multiple sensor providers and monitoring environments.
The focus is not more hardware,
but extracting earlier, actionable insight from acoustic data — whether from existing monitoring systems or sensing deployed as part of an Adelitics solution.
​
From signal to decision
-
Network data (sensor, fibre or hybrid)
-
Advanced acoustic analytics
-
Actionable insights to support targeted intervention
​
Delivery models
Data as a Service (DaaS)
Ongoing analysis of acoustic and network data, delivering prioritised insights on leakage and emerging failure risk.
Turnkey delivery
End-to-end deployment: sensing (where required), analytics, integration and operationalisation, delivered directly or through partners.
Advisory
Targeted, bespoke support to design, adopt and embed leak-before-break approaches, from pilots through to scaled operational use.
Adelitics team undertaking night time sensing verification activities
EVIDENCE
Adelitics’ analytical methods have been developed and validated through more than a decade of peer-reviewed research and utility-scale deployment.
The findings below summarise results from permanent monitoring studies, machine learning validation and transmission scale field trials.
Validating in operating networks
Water distribution systems are large, distributed and constantly changing environments.
In practice:
Monitoring points may be spaced more than 100 metres apart
Different pipe materials transmit acoustic signals differently
Network conditions vary with demand and background noise
Rather than relying solely on correlation between sensor pairs, Adelitics analyses how acoustic behaviour evolves over time at each monitoring location.
​
This enables two complementary monitoring approaches:
Sentinel mode – a single sensor monitors its surrounding acoustic zone, reducing the number of sensors required.
Paired sensing – where sensors are deployed in pairs to support correlation and localisation when required.
Changes in acoustic behaviour are analysed across multiple time scales and signal characteristics to identify patterns consistent with emerging leaks or structural deterioration.
Reducing false positives in operational data
Water networks generate constant background acoustic activity.
Traffic, air conditioning units, irrigation systems, rainfall, pumps and routine water use all contribute to this background noise.​
​
Machine learning models were trained and validated using acoustic data collected from operating water networks across multiple utilities.
​
​This allows the analytics to reliably distinguish between background acoustic activity, intermittent environmental signals and emerging leak signatures, as described in published validation studies.³
​
Many utilities operating permanent sensing systems experience high alert volumes that are difficult to operationalise.
​
Adelitics’ analytical framework is designed to convert raw sensing data into a manageable stream of prioritised, actionable alerts, supporting confident field decision-making.
Operational impact:
Greater confidence in field decisions
Fewer
unnecessary excavations
More efficient
use of constrained resources
Locating condition change at transmission scale
Once an anomaly is detected, accurate location becomes critical.
Field studies demonstrate that low-impact, continuous acoustic signals can be used to assess pipe condition. ​​This supports long-range, non-destructive screening before targeted inspection.
​
In field validation:​​​​​​
​
​
​
​​
​
​
​
Developing faults were located within around 1% of pipeline length over kilometre-scale sections
Material transitions and closely spaced pipe features were reliably distinguished from emerging defects
What this enables
Spotting developing problems earlier
before they turn into major failures
Reliable detection
in operating networks
​
Pinpointing where condition changes are occurring
even in long pipelines
Leak-before-break is demonstrated as a structured, measurable operational process.
​
Stephens, M. L., Gong, J., Zhang, C., Marchi, A., Wilson, A., & Lambert, M. F. (2020).
Leak-Before-Break Main Failure Prevention for Water Distribution Pipes Using Acoustic Smart Water Technologies: Case Study in Adelaide.
Journal of Water Resources Planning and Management. DOI: https://doi.org/10.1061/(ASCE)WR.1943-5452.0001266
​​
Gong, J., Lambert, M. F., Stephens, M. L., Cazzolato, B. S., & Zhang, C. (2020).
Detection of Emerging Through-Wall Cracks for Pipe Break Early Warning in Water Distribution Systems Using Permanent Acoustic Monitoring and Acoustic Wave Analysis.
Water Resources Management, 34(8), 2419–2432. DOI: https://doi.org/10.1007/s11269-020-02560-1
​
Zhang, C., Alexander, B. J., Stephens, M. L., Lambert, M. F., & Gong, J. (2022).
A Convolutional Neural Network for Pipe Crack and Leak Detection in Smart Water Networks.
Structural Health Monitoring. DOI: https://doi.org/10.1177/14759217221080198
Zeng, W., Nguyen, S. T. N., Lambert, M., & Gong, J. (2024).
Field Study on Proactive Pipe Condition Assessment Using Hydroacoustic Noise.
Structural Health Monitoring, 24(4), 2052–2063. DOI: https://doi.org/10.1177/14759217241284729
​
Full peer-reviewed publications available on request: info@adelitics.com.au