Condition Monitoring & Predictive Maintenance Management

Reliability Operating Centre

Condition Monitoring & Predictive Maintenance Management

Condition monitoring that ends in a work order. We manage the programme rather than the routes: which technique detects which failure mode, what the alarm limits should be, and whether findings turn into planned work in time to matter.

Why condition monitoring underdelivers

The technology is rarely the problem. The programme around it usually is.

Technology bought before strategy

Sensors installed and routes established without a specific failure mode they were selected to detect. The data arrives, nobody can say what it is supposed to prevent, and it becomes a compliance activity.

The loop breaks at the end

A finding is recorded. It does not become a scheduled work order, or it does but too late to be planned properly. The asset fails anyway, and the programme gets blamed for something that was actually detected.

Alarm limits set once and never revisited

Limits configured at commissioning, never tuned against actual failures. False positives accumulate, confidence erodes, and eventually people stop opening the alerts. A real one arrives and it looks the same as the last thirty.

A finding that does not become a scheduled work order before the P-F interval runs out is an expensive record of a failure you were going to have anyway.

What we do

Match the technique to the failure modeVibration, oil analysis, thermography, ultrasonic and motor circuit analysis each detect particular things. We work from your failure modes to the detection method — the same failure-mode-first logic as RCM — not from the technology catalogue backwards. New to condition monitoring? Start with our plain-language guide.

Set the interval against the P-F curveThe monitoring interval has to be shorter than the warning period the failure mode actually gives, and long enough to be affordable. Get this wrong and you are either paying for data you cannot use or missing failures you paid to detect.

Set and tune the alarm limitsInitial limits from the standards and the OEM, then tuned against what your equipment actually does. Reviewed whenever a failure is missed or a false positive is raised, so the limits improve rather than calcify.

Govern the loop from finding to work orderMOST OFTEN SKIPPEDEvery finding tracked from detection through to planned work, execution and confirmation of what was found on teardown. This is where most programmes leak, and it is administrative rather than technical, which is precisely why a busy site engineer never gets to it.

Report programme effectivenessNot how many readings were taken. Whether the programme is catching failures early enough to act on, and how often it cries wolf.

Two numbers tell you if it is working

Most operations cannot produce either. We report both, every month.

Detection rateThe share of failures on monitored assets that were detected before functional failure, with enough warning to plan the work. This is the only measure of whether the programme is doing its job.

False positive rateHow often the programme raised an alarm that teardown did not support. High rates destroy confidence faster than missed detections, because they cost money and credibility at the same time.

Why both together. Either number alone can be gamed. Tighten the limits and detection improves while false positives climb. Loosen them and the reverse. Reported as a pair, the programme has nowhere to hide, including from us.

What is in scope, and what is not

We are deliberately not your condition monitoring contractor. That distinction matters when it is time to hold someone to account for the data.

We do

Condition monitoring strategy, by asset class and failure mode
Technique selection and interval setting
Alarm limit setting, review and tuning
Interpretation and review of findings
Governance of the finding to work order loop
Technology and vendor assessment
Monthly effectiveness reporting

We do not

Run your vibration or oil sampling routes
Own or operate your sensors and hardware
Replace your existing CM contractor
Perform the maintenance work the findings generate

If you have a condition monitoring provider, we make their output useful and hold the programme to account for results. If you do not, we help you select one on the basis of the strategy rather than the sales pitch. Where monitoring reveals the strategy itself has drifted, that feeds maintenance task optimisation.

Frequently asked questions

What is condition monitoring management?

Managing the programme around condition monitoring rather than collecting the data: matching technique to failure mode, setting intervals against the P-F curve, tuning alarm limits, and governing every finding through to a planned work order. The routes can stay with your existing provider — the management layer is what usually is missing.

Do you run vibration or oil analysis routes?

No, deliberately. We do not run routes, own sensors, or replace your condition monitoring contractor. We make their output useful and hold the programme to account for results — which only works if we are not also the ones being held to account for the data.

What is the P-F interval and why does it matter?

The window between the point a failure becomes detectable (P) and functional failure (F). The monitoring interval has to be shorter than that window or you are paying to detect failures too late to plan the work — which is how a programme records failures instead of preventing them.

Why do condition monitoring programmes fail?

Three patterns: technology bought before strategy, so nobody can say what a sensor is meant to prevent; the loop breaking at the end, where findings never become planned work orders; and alarm limits set at commissioning and never tuned, so false positives bury the real alerts.

How do you measure whether condition monitoring is working?

Two numbers together: detection rate (share of failures on monitored assets caught early enough to plan) and false positive rate (alarms teardown did not support). Either alone can be gamed by moving alarm limits; as a pair the programme has nowhere to hide.

Find out what your programme is actually catching

The readiness check includes a first look at detection and false positive rates. Most sites have never had these calculated.

Book a fit and data readiness check