The four established techniques
Planned maintenance answers "when is this due?". Condition monitoring answers "does this actually need doing?". The second question is worth money in both directions — it prevents overhauls of healthy machinery and catches failing machinery before it becomes a casualty.
| Technique | Detects | Typical cadence |
|---|---|---|
| Vibration analysis | Bearing wear, misalignment, imbalance, looseness | Monthly to quarterly on rotating machinery |
| Lube oil analysis | Wear metals, contamination, additive depletion, water ingress | Per sampling schedule, typically monthly |
| Thermography | Electrical connection faults, insulation, bearing heat | Quarterly on switchboards and motors |
| Performance monitoring | Efficiency degradation across systems | Continuous |
None of these are new. What has changed is the cost of collecting the data continuously rather than periodically, and the ability to send it ashore.
Where it genuinely pays
The value concentrates where three conditions coincide: failure is expensive, failure develops gradually, and the equipment is not fully redundant. Main engines, turbochargers and auxiliary engines meet all three. A duplicated seawater pump does not — running it to failure and switching to the standby is a rational strategy.
That distinction is the single most useful filter when a vendor proposes monitoring everything.
Making it work at sea
1. Establish a baseline. A vibration reading is meaningless without knowing what the machine looked like when it was healthy. Baseline after overhaul, and after any change.
2. Consistent measurement. Same point, same mounting, same load condition, same operator technique. Inconsistent measurement produces trends that are artefacts.
3. Trend, do not threshold. A single reading below an alarm limit tells you little. A reading that has doubled over four months tells you a great deal.
4. Close the loop with the PMS. Condition findings must generate PMS jobs, and job completion must reset the baseline. Monitoring that lives in a separate system is a hobby.
5. Give the analysis to someone competent. Data without interpretation is noise. Either train a shore-side analyst or contract one, and make sure their findings reach the chief engineer in usable form.
6. Act on findings. The most common failure of condition monitoring programmes is not technical. It is that a finding is raised, the spare is not available, the job is deferred, and the machine fails anyway.
Where class fits in
Where a vessel operates under a class-approved planned maintenance scheme, condition monitoring can be credited — allowing survey intervals to be extended for machinery whose condition is demonstrably being managed. That requires:
- Class approval of the monitoring arrangement
- Competent, named personnel
- Genuine, retrievable trend data
- Documented action on findings
The prize is reduced opening-up surveys and less downtime. The obligation is that the data must be real. A class-approved scheme built on fabricated readings is a far worse position than no scheme at all.
Predictive versus condition-based
The terms are used loosely. Usefully distinguished:
| Approach | Basis |
|---|---|
| Reactive | Fix on failure |
| Preventive / planned | Fix on calendar or running hours |
| Condition-based | Fix when measured condition indicates |
| Predictive | Model estimates remaining useful life from patterns in data |
Most maritime programmes today are condition-based with some predictive elements. True predictive modelling requires large volumes of consistent data and enough failure examples to learn from — which is why it works best across a fleet of sister vessels rather than on a single ship.
Avoiding the dashboard trap
A pattern repeats across the industry: sensors are fitted, a dashboard is delivered, everyone is impressed, and eighteen months later nobody opens it.
Preventive measures:
- Define, before purchase, which decisions the data will change
- Name the person who reviews it and how often
- Set alert thresholds that produce a manageable number of alerts — alert fatigue kills programmes
- Route findings into the PMS as jobs, not into an inbox
- Report avoided failures and unnecessary overhauls avoided; that is the return
- Review the programme annually and switch off what is not used
class-approved maintenance schemes vary by society — confirm requirements with your class. Value chart is an indicative model. Reviewed by the Zeaclub Editorial Team, 24 August 2026.
Frequently asked questions
Does predictive maintenance replace the PMS?
No. It changes *when* jobs are done. The PMS remains the record of equipment, requirements and completed work.
What equipment should be monitored first?
Main engine, turbochargers and auxiliary engines — where failure is expensive, develops gradually and redundancy is limited.
Do we need continuous sensors?
Not necessarily. Periodic handheld vibration and oil analysis deliver much of the value at a fraction of the cost. Continuous monitoring adds most on high-value, high-consequence machinery.
Can class credit condition monitoring?
Yes, within a class-approved planned maintenance scheme, subject to approval of the arrangement, competent personnel and genuine documented data.