Predictive maintenance for thirteen gas compressors across five sites
A proof of concept on one compressor, then scaling to twelve more, proposed for a gas producer with no usable history and few sensors.
Challenge
A gas producer wanted to monitor 13 compressors on five sites to find reliability issues and performance degradation early enough to act, with existing systems giving only an overview, no historical data to train on and gaps in instrumentation.
Solution
Arundo proposed a two-month proof of concept on one compressor: connect the live PLC data through the edge, apply existing compressor model templates, tune with the site engineers, identify the sensors still needed, and size the value from avoided downtime and maintenance; then scale the models to the remaining twelve compressors with local tuning.
Impact
A path from a single-compressor proof of concept to fleet-wide failure prediction, with the business case built on the 0.35–0.45% of annual production that compressor failures had cost a comparable operator.