Color
Book a demo
Glossary
Operations and asset management

What is Predictive maintenance?

Definition

Predictive maintenance schedules repairs from trends in equipment condition data, such as vibration, temperature or current draw, so work happens before a failure rather than on a fixed calendar or after a breakdown.

Where it pays back

The approach suits critical rotating and refrigeration equipment where an unplanned stop is expensive: chillers, compressors, pumps, fans and air handlers in hospitals, data centers, airports, manufacturing lines and dairy cold chains. Reliability engineers and facility managers use it to move discretionary work to the moment the data says it is needed.

Savings reported by DOE

The DOE FEMP Operations and Maintenance Best Practices Guide, Release 3.0, estimates that predictive maintenance saves 8 to 12 percent over preventive maintenance, and 30 to 40 percent where a site relies heavily on reactive repair. The industry surveys it cites for working programs report 25 to 30 percent lower maintenance cost, 70 to 75 percent fewer breakdowns and 35 to 45 percent less downtime.

aQ Core onboards anything that emits telemetry, now including vibration and acoustic sensors next to power meters. aQ PredictIQ trends those signals per asset, and the electrical twin adds a second view: a motor drawing more kVA for the same duty, or a feeder whose power factor slips, can show wear before a breakdown.

Questions

Predictive maintenance: common questions

Predictive maintenance is a strategy that uses measured equipment condition to decide when to service it. Instead of servicing on a calendar or after a failure, a team watches signals that drift as wear builds and acts at the point the data shows rising risk. Aqueduct's approach is on the predictive maintenance page.

Enough history to show normal behavior for each asset. Most programs begin with points already on the BMS and add vibration or acoustic sensors to the few machines whose failure hurts most, then widen coverage once the first alerts have proved their value.

No. Statutory inspections and warranty tasks keep their calendars. Predictive data mostly changes the timing of discretionary work such as bearing replacement or coil cleaning, which is where condition-based maintenance and predictive methods overlap.

The CMMS stays the system of record for work orders, assets and labor. Predictive analytics decides when work is needed and raises the alert, and the CMMS schedules and tracks it. Aqueduct raises maintenance alerts from asset data, as covered under CMMS and EAM.

See Aqueduct on your own facility.

Bring a site, a utility bill or a question. We will show you where the loss is.

Book a demo Talk to our team