Turn thirty years of records into analytics.

Engagement detail

Sector

Energy

How we worked

Project delivery

Stack

Cloud infrastructure, field data-collection tools, audio analytics, Tableau

Headline

30 years of records digitised

The problem

A maintenance-engineering firm with more than 31 years of experience in in-plant maintenance for critical infrastructure — lubrication consultancy and specialised welding wherever steam is used — wanted to move from reactive maintenance to intelligent, data-driven leak prevention. Decades of operational data existed but in unorganised form, scattered across formats and departments, with no consolidated system for identifying and prioritising equipment losses before they became costly failures.

What we built

We set up cloud infrastructure to host and manage over 30 years of previously unorganised data, digitising structured records and unstructured audio with enhanced quality monitoring using a mix of off-the-shelf and custom-built field tools. On top we built an audio-analytics capability for automatic leak-type and leak-size detection from the frequency signature of ultrasound waves, plus an analytics-based decision framework used across prominent oil refineries to reduce losses from 11 distinct types of equipment aberration. A Tableau environment and internal processes completed the solution.

Impact

• Automatic classification of air leaks by type and severity across plant departments.
• A consolidated, 30+-year data foundation supporting ongoing loss-reduction analysis.
• A repeatable audio-analytics leak-detection capability applicable across industrial sites.