On most vessels, the maintenance plan exists. It is often a substantial document. The problem is not its absence, it is its schedule compliance: tasks permanently overdue, a crew that has given up trying to catch up, and audit preparation that consists of reconstructing what should have been recorded in real time.
The cause is almost always the same: the plan was designed as though it applied to a workshop ashore. Yet a vessel imposes three constraints no factory ever faces. The crew that carries out the plan changes every two or three months. The missing part does not arrive within the hour but at the port call, sometimes the one after. And the result has to be provable in front of an ISM auditor or a classification society surveyor.
This article sets out the four-step method for building a plan that survives those three constraints 鈥 what the shipping industry calls a PMS, a planned maintenance system.
Step 1 鈥 Build the inventory and rank it
The asset hierarchy comes first
Nothing can be built without a physical description of the vessel. You break it down into systems 鈥 propulsion, electrical generation, steering, fire safety, water treatment, deck 鈥 then into equipment, then into sub-assemblies. Each node carries its technical data sheet, its maker's documentation and its associated spares. That is the role of the Equipment module.
One rule is worth setting down now: sister vessels must share the same asset hierarchy and the same coding. Without that, no fleet-wide comparison is possible, and you will discover the problem two years later, when it costs ten times more to put right.
The mistake to avoid: maintaining everything the same way
Giving an alleyway fan as much attention as the steering gear produces an unmanageable plan, and therefore an abandoned one. Criticality ranking is not a refinement, it is what makes the plan workable with the crew actually on board.
Three questions are enough to classify a piece of equipment. Does a sudden failure put people or the environment at risk? Is the equipment without redundancy, to the point that losing it means reducing speed or cancelling a port call? Is the full cost of a breakdown out of all proportion to the cost of maintaining it?
| Category | Share of the asset base | Examples on board | Maintenance strategy |
|---|---|---|---|
| A 鈥 Critical | 15 to 20% | Main engine, steering gear, generator sets, sea water and cooling pumps, fixed fire-extinguishing systems, starting air compressors | Reinforced planned maintenance and condition-based maintenance, with recorded testing of standby arrangements |
| B 鈥 Important | Around 30% | Separators, transfer pumps, deck winches and hydraulics, air conditioning, fresh water generation | Systematic, triggered by calendar or running hours |
| C 鈥 Secondary | 50 to 55% | Alleyway lighting, small fittings, non-critical redundant equipment | Corrective: replaced on failure |
One important caveat. This ranking is economic. Safety equipment 鈥 liferafts, lifejackets, extinguishers, fire detection, life-saving appliances 鈥 is also subject to regulatory intervals that apply whatever its economic criticality. It enters the plan as an obligation, not as a trade-off.
The link with the ISM Code
Chapter 10.3 of the ISM Code asks for precisely this work: identifying equipment whose sudden failure may result in a hazardous situation, and providing specific measures for it, including the regular testing of standby arrangements. Your list of category A assets is, in practice, the answer to that requirement. It is worth building once and using for both.
Step 2 鈥 Define the tasks and their triggers
Once the assets are ranked, you have to decide what to do and what triggers it. It is on the second point that shore-based plans fail at sea.
Three trigger logics
- The calendar. Monthly inspection of extinguishers, weekly test of the emergency generator, annual servicing of liferafts. Easy to schedule, but completely disconnected from actual use: a vessel alongside for two months burns through the same due dates as one in continuous service.
- The counter. Running hours, number of starts, miles run, winch cycles. This is what reflects genuine wear and what makers prescribe. It also assumes that someone reads the counters 鈥 hence the value of a running hours module fed by the engine room watch rather than by a monthly entry from memory.
- Condition. Oil analysis, vibration measurement, thermography. Reserved for category A assets, it allows you to intervene before the failure rather than on a fixed date.
A solid PMS combines all three. The classic error is to put everything on the calendar because it is easier to write, then to be surprised that the plan is at once too heavy on lightly used vessels and insufficient on the others.
Writing tasks that can actually be carried out
A PMS task must fit into a job procedure that can be read by someone who has never done the work before. It carries the procedure, the estimated duration, the skills required, the tools, the safety measures 鈥 isolation, permit to work, certifications, a subject covered in detail in our article on lock-out tag-out and certifications 鈥 and the list of parts to be consumed.
That last point is not a detail. A task that cannot be carried out for want of a spare on board generates permanent delay, and a permanently overdue plan ends up being ignored. Linking the plan to stock is what prevents that drift.
Step 3 鈥 Schedule against the real windows
This is the step industrial methods handle worst, because ashore you can almost always stop a machine. At sea, you cannot.
The four constraints to respect
- The operating window. A task that requires the main engine to be shut down is not scheduled on passage. It is set against a port call, an anchorage or a drydocking. The plan therefore has to be built around the operating programme, not the other way round.
- Parts availability. Check the stock before scheduling, not when you open the toolbox. On long voyages, anticipation becomes the main subject 鈥 we covered it in our guide to spare parts management at sea.
- Crew numbers and skills on board. A plan sized for three engineer officers becomes unrealistic when there is only one. Group tasks by area and by skill to limit walking and repeated isolations.
- Crew rotation. Avoid concentrating heavy tasks in the crew change week. That is when the competence on board is most fragile and data entry errors are most frequent.
Smooth the load rather than stack it
A poorly smoothed plan produces peaks: forty tasks in the first week of the month, six in the third. The crew drops out at the first peak and never catches up. Spread the load, even if it means shifting non-regulatory intervals slightly, and check the weekly workload per vessel before freezing the plan.
Step 4 鈥 Measure and adjust
A PMS is never finished. Four indicators are enough to steer it, and it is better to follow four of them seriously than fifteen for appearances.
| Indicator | What it measures | What it reveals when it drifts |
|---|---|---|
| Schedule compliance | Share of planned tasks carried out within their window | Below 80%, the plan is too heavy for the crew on board or badly matched to the port calls |
| Backlog | Number and age of overdue tasks | A backlog growing in a straight line signals an unrealistic plan, not a careless crew |
| MTBF on category A assets | Mean time between failures on critical items | If it falls despite the plan, the intervals are too far apart or the tasks do not target the right failure mode |
| Share of corrective work | Corrective hours as a proportion of total maintenance hours | A share that stays high after a year shows that the plan is not targeting the breakdowns that actually happen |
These four figures must be readable per vessel and consolidated for the fleet, which is the function of the dashboard. They also provide the material for the management review required by chapter 12 of the ISM Code.
The most useful reflex
When schedule compliance stalls, the instinctive reaction is to call the crew to order. It is almost always the wrong one. Start by lightening the plan: remove duplicate tasks, stretch the non-regulatory intervals that nothing justifies, merge those carried out in the same place. A realistic plan achieved at 95% protects infinitely better than an exhaustive plan achieved at 55%.
The case of the class-approved PMS
Beyond internal organisation, a well-kept maintenance plan can carry direct regulatory value. Under a planned maintenance scheme approved by the classification society, the records and job reports can be used to credit certain machinery surveys, in place of systematic opening up.
The conditions vary from one society to another, but they converge on one point: the file must be documented, dated, attributed and consistent with the counters. That is exactly what a PMS kept in a CMMS produces, and what a paper binder struggles to demonstrate. Talk to your classification society before you freeze the plan: the benefit in survey time and in dismantling avoided is rarely anticipated.
Five mistakes that make a PMS fail on board
- Taking the maker's manual as it stands. Maker's intervals are calibrated on average use. On a vessel running 300 hours a year they produce an absurd plan; on one running 6,000, they are not enough.
- Confusing completeness with quality. A plan of 2,000 tasks of which 700 are overdue is worth less than a plan of 600 tasks that are kept up to date. Every task written down is an audited promise.
- Forgetting familiarisation. Chapter 6 of the ISM Code requires documented familiarisation on every embarkation. Without it, the new arrival works the plan the wrong way for his first three weeks.
- Letting the plan live outside the CMMS. As soon as a parallel spreadsheet appears on board, two versions of the truth coexist and the auditor will find the wrong one.
- Never revising it. A PMS that has not moved in three years is not stable, it is dead. The annual review, fed by the indicators, is part of the system.
Frequently asked questions
What is the difference between a preventive maintenance plan and a PMS?
The two describe the same thing. "Preventive maintenance plan" is the generic term; planned maintenance system, or PMS, is the expression used in the shipping world and in classification society documents. It is the one you will meet in an audit.
How long does it take to build a PMS?
For a first vessel, allow several weeks between the equipment inventory and a plan that is genuinely worked to day by day. The vessels that follow go much faster once the fleet coding has been settled.
Should everything go into the plan from the start?
No, and that is in fact the surest way to fail. Start with category A equipment and the regulatory tasks, reach a high level of schedule compliance, then extend. A plan that works on 30% of the asset base is worth more than a theoretical plan covering 100%.
Does the PMS replace the engine room logbook?
No, it feeds it. Closed work orders and counter readings feed the engine room logbook, which remains the reference record at survey.
In summary
An effective preventive maintenance plan is not judged by its thickness but by its schedule compliance. Rank the assets, choose the right trigger for each of them, schedule against the real operating windows, and steer with four indicators rather than with intuition. The rest 鈥 completeness, the fineness of the job procedures, condition-based maintenance 鈥 comes afterwards, and only once the foundation holds.
That foundation is what 91麻豆精品 is designed to carry: a single asset hierarchy per vessel, calendar and running hours triggers, work orders that can be closed offline and a history that stands up in an audit. The marine CMMS is developed in Marseille by seafarers and deployed on more than 700 vessels. Request a demonstration on your own fleet, or read the complete guide to marine CMMS.
