Marine Fleet Management Software for Marine Condition Monitoring

A marine machinery condition monitoring system tracks the condition of engines, generators, propulsion and auxiliary machinery over time, so developing faults are found early and diagnosed quickly. Most discussions stop at early detection. For a ship manager, the harder question is what happens next. When an alarm sounds or a machine fails at sea, how quickly can the shore team understand what happened and help the crew decide what to do?

This page answers that question in five parts:

• Which monitoring techniques catch which failures
• Which equipment to prioritize
• How shore-based fault support works step by step
• How condition monitoring connects to class surveys under IACS UR Z27
• What to check when choosing a system

For how alert thresholds and shore connectivity work in general, see our remote vessel monitoring guide. For the data foundation underneath, see our vessel data integration guide.

What a Machinery Condition Monitoring System Actually Covers

The Six Functions Defined in ISO 13374

ISO 13374 splits condition monitoring into six functional blocks. The framework is useful when comparing products, because systems sold under the same label can cover very different parts of it.

Many products marketed as condition monitoring stop at state detection. They tell you that something is abnormal, not what it is or what to do. The later functions, from health assessment to advisory generation, are the ones that matter when a ship calls for help. In practice they are delivered by software and people together, which is why the shore support process in this guide matters as much as the software.

Where It Sits in Failure Response

Several systems take part when something goes wrong, each with a different role:

• The alarm and monitoring system (AMS) protects machinery with fixed limits and triggers immediate action on board.
• The condition monitoring system assesses how the machine's condition has developed and what is likely to be wrong.
• The planned maintenance system holds maintenance history, spare parts and work orders.
• The remote monitoring link makes all of this visible ashore.

Fault support works best when these are connected. For example, a condition monitoring alert that opens directly onto the component's maintenance history saves the first half hour of any investigation.

Matching Monitoring Techniques to Failure Modes

Each technique sees a different set of failure modes. Choosing techniques by equipment type ("we monitor pumps with vibration") misses the point. Choose them by the failure modes that matter for each piece of equipment.

Techniques also detect the same failure at different stages of its development. Layering them lengthens the warning time. For example, oil debris or vibration may reveal a bearing problem long before it shows as a temperature rise. How much warning time you need, and how that sets sampling and review frequency, is covered under the P-F interval in our remote vessel monitoring guide.

Continuous or Periodic Measurement?

Not every machine needs permanent sensors. The choice depends on two things: how fast its failures develop and how much a failure costs.

• Continuous monitoring suits critical equipment whose failures can develop within hours or days, such as main engine components, generators and critical pumps.
• Periodic measurement, such as portable vibration routes, oil samples and thermography surveys, suits the large population of less critical machines whose failures develop over weeks or months.

Periodic data should still go into the same system as continuous data. Otherwise, the shore team diagnosing a failure cannot see the last vibration route or oil report next to the live trends.

Which Equipment to Monitor First

Section 10.3 of the ISM Code already requires companies to identify equipment and technical systems whose sudden operational failure may result in hazardous situations, and to take specific measures to promote their reliability. That critical equipment list is the natural starting point for condition monitoring. It is not the whole answer, because it is built around hazard rather than commercial impact or detectability.

A more complete ranking looks at two dimensions: the consequence of failure (safety, off-hire, cargo, repair cost and spare parts lead time) and detectability (whether the failure gives a measurable warning).

The upper-right cell is often overlooked. Some of the most expensive failures, including many electrical ones, cannot be predicted. For those, condition monitoring pays off in diagnosis rather than prevention, which is why fault support deserves as much attention as early warning.

When an Alarm Sounds: How Shore-Based Fault Support Works

Most fleets already have shore support of some kind: the chief engineer calls the superintendent, who calls the equipment maker. A condition monitoring system does not replace that chain. It adds evidence and structure, so every step happens faster and with less guesswork. The process below applies whether the trigger is a condition monitoring alert, an AMS alarm or an actual failure.

Step 1: Confirm the Signal Is Real

A surprising share of alarms come from sensors, not machines. Before anyone reduces load or stops equipment, check three things:

• Do related signals move consistently? A bearing temperature rising while load, lube oil temperature and vibration stay flat points toward the sensor.
• Does the signal behave physically? Sudden jumps, flat lines or values at the edge of the sensor's range are typical of sensor faults.
• Has the sensor recently been calibrated or replaced? This is only visible if the system keeps sensor history.

The crew can check locally. Shore can check history and compare with sister ships. Doing both in the first minutes avoids unnecessary slowdowns and, just as important, avoids dismissing a real fault as a sensor problem.

Step 2: Reconstruct What Happened

The shore team needs the detailed data from before and after the event, not one-minute averages. Averages hide exactly what diagnosis depends on: which value moved first, how fast it moved, and what else happened at the same moment. Systems that keep high-resolution data on board, for example one reading per second for several weeks, allow the shore team to request the relevant window and see the event as it developed.

Two comparisons usually narrow the problem quickly:

• Against the machine's own baseline: is this a sudden change or the end of a slow drift?
• Against sister ships or sister units: is this unit behaving differently from identical equipment?

Step 3: Narrow Down the Cause

Diagnosis is faster when the knowledge needed for it is organized in advance rather than rebuilt on each call:

• Symptom-to-cause tables and fault trees for critical equipment, from maker manuals and the fleet's own experience
• A searchable record of past cases: what was seen, what the cause turned out to be, what fixed it
• Troubleshooting procedures linked to alerts, so the engineer on watch sees the relevant first checks the moment an alert appears

The last point is where condition monitoring moves from state detection to advisory generation in ISO 13374 terms. It also turns the knowledge of experienced engineers into something every crew can use.

Step 4: Support the Operating Decision

The practical question is always what to do now. The options are usually some version of these:

• Continue with closer observation
• Reduce load
• Isolate the unit and switch to standby
• Stop for repair at sea
• Change the voyage plan

Shore can contribute what the ship often cannot see:

• How fast the condition is deteriorating
• What happened in similar cases
• The maker's advice
• Whether spares are available
• The consequences for the voyage and charter

The decision itself stays on board. IACS UR Z27 states this directly for classed condition monitoring schemes: condition monitoring does not replace routine surveillance or the chief engineer's responsibility for decisions based on their own judgment. Good shore support improves the decision without taking it over.

Step 5: Organize the Repair

Once the cause is known, shore coordinates what the ship cannot arrange from sea:

• Sending the maker a data package instead of a verbal description, so their specialists start from evidence
• Arranging spares and service engineers for the next suitable port
• Checking whether the damage must be reported to the classification society. For machinery under a class-approved condition monitoring scheme, UR Z27 requires damage to be reported and repairs to be carried out to the surveyor's satisfaction

Step 6: Learn From Every Failure

A failure the monitoring system did not catch is information about the system. UR Z27 builds this into classed schemes. It requires them to identify defects and unexpected failures that condition monitoring did not prevent, to review failure data, and to have a method for amending the scheme. The same discipline is worth applying whether or not you seek class credit:

• Add or refine alert rules. Often a combination of readings would have shown the problem earlier than any single limit.
• Record every threshold change, with who changed it and why. Z27 requires a way to identify where alarm and warning limits have been modified.
• Update the troubleshooting procedure linked to that alert.
• Apply the lesson to sister ships and sister units, not just the ship where it happened.

Example: Tracing a Generator Trip or Blackout

Electrical events show why the steps above depend on the right data. A typical sequence runs like this:

1. One generator trips on a protection function, such as reverse power, overcurrent or under-frequency.
2. Its load transfers to the remaining generators.
3. If they cannot carry it, non-essential consumers are shed or the ship blacks out.

The whole chain can play out in seconds.

Reconstructing it requires:

• High-resolution data for load, voltage, frequency and breaker status on every generator
• The alarm and event log from the power management system
• Consistent timestamps across all of these sources

With that data, the shore team can tell a generator that tripped because it failed from one that tripped because another failed first. With one-minute averages, the cause and the effect look identical. Electrical diagnosis is also a specialty many fleets lack in-house, so a system that helps analyze electrical events, or a vendor that supports it, is worth asking about specifically.

Organizing Shore Support: Three Models

Most fleets end up combining them. A common pattern is in-house first response and triage across all equipment, with maker services for specialist equipment such as propulsion or engines under service agreements. The in-house layer only works if it can see data from every maker's equipment in one place. See the vessel data integration guide for how that foundation is built.

Whichever model you choose, measure it. Useful measures include:

• Time from alert to first shore response
• Time to diagnosis
• Mean time to repair
• The share of alerts that led to action
• Service attendances and off-hire avoided
• Repeat failures of the same type

Condition Monitoring and Class: What IACS UR Z27 Requires

IACS Unified Requirement Z27, Condition Monitoring and Condition Based Maintenance, applies where condition monitoring results are used to influence the scope or frequency of class surveys. IACS societies apply it to survey schemes approved on or after January 1, 2020. It applies only to ships operating an approved planned maintenance survey scheme (see UR Z20). Its main requirements are:

• Limits: alarm and warning limits are based on the maker's guidelines or a recognized international standard. Software using complex algorithms or machine learning may work independently of maker limits, with approval based on maker recommendations, industry standards and the society's experience.
• Confidence: the system must give at least as much confidence in machinery condition as traditional survey or maintenance techniques.
• Remote diagnosis: where data is analyzed remotely, the system must meet applicable cyber safety and security standards and continue to operate on board if communication is lost.
• Change control: the system must identify where alarm and warning limits have been modified, and changes to software and parameters must be recorded.
• Failure feedback: the scheme must identify defects and failures that condition monitoring did not prevent, and include a method for amending the scheme.
• Documentation: the requirements cover baseline data, the list of monitored parameters, sensor calibration records, all condition monitoring data since the machine was last opened, and the qualifications of the people and company analyzing the results.
• Responsibility: the chief engineer is the responsible person on board.
• Surveys: an installation survey with baseline readings, an implementation survey (no earlier than six months after installation and no later than the first annual survey), and an annual audit. Approval is reconsidered if the ship is sold, changes manager or transfers class.

Propeller shaft surveys follow their own scheme under IACS UR Z21. Where conditions such as regular analysis of stern tube lubricating oil are met, shaft withdrawal intervals can be extended. Check the exact conditions with your classification society.

Even if you never seek survey credit, Z27 is worth reading as a specification for a well-run condition monitoring system. Baselines, recorded threshold changes, onboard operation without the link, qualified analysts, and a feedback loop from failures are exactly what makes fault support reliable.

Standards Worth Knowing

Limitations

• Sudden failures stay sudden. Where a failure mode has no measurable warning, monitoring can only speed up diagnosis and recovery.
• The first months produce more false alerts. Baselines need time to cover the ship's normal operating range. Plan for tuning, and record every change.
• Interpretation is a bottleneck. Vibration spectra and oil reports need skilled analysts. Software reduces the workload but does not remove the need for them.
• Periodic data depends on the crew. Portable routes and oil samples are only as good as the discipline with which they are taken.
• Maker-specific systems create gaps. Data locked inside one maker's platform is hard to combine with the rest during a failure investigation.

Selecting Marine Fleet Management Software for Condition Monitoring: FAQ Guide

About Marine Condition Monitoring

ISO 13372 defines condition monitoring as the acquisition and processing of information and data that indicate the state of a machine over time. In marine use, that means tracking engines, generators, propulsion and auxiliary machinery through sensor data, periodic measurements and analysis, so maintenance and operating decisions are based on actual condition rather than elapsed time alone. The most useful systems go beyond flagging abnormal values to diagnosing causes and recommending action, as described in the six ISO 13374 functions.

Benefits of Marine Condition Monitoring

The direct benefits are fewer unplanned stoppages, maintenance timed to actual condition, and faster diagnosis when failures do occur. Two further benefits are easy to underestimate. First, under a class-approved scheme following IACS UR Z27, condition monitoring results can influence the scope and frequency of class surveys. Second, every diagnosed failure adds to a record the whole fleet can learn from. See what Z27 requires.

Challenges Solved by Marine Condition Monitoring

Condition monitoring addresses three recurring problems. The first is that knowledge depends on individual engineers: rules and procedures linked to alerts keep that knowledge in the fleet when people rotate. The second is unexpected stoppages, which early detection turns into planned work. The third is slow shore support caused by a lack of evidence: high-resolution event data lets shore teams diagnose from facts rather than descriptions. See how shore-based fault support works.

Choosing a Marine Condition Monitoring System

Start from your failure modes and critical equipment, not from the number of sensors a system supports. Then check how the system performs when something actually goes wrong. Does it keep detailed event data? Does it give ship and shore the same view? Can your own team add rules and link alerts to procedures? Does it handle data from all makers' equipment? The checklist below covers these points in detail.

Use Cases for Marine Condition Monitoring

Typical uses include:

• Trend monitoring of main engines and rotating machinery, with findings scheduled into port calls or dockings
• Prioritizing attention across a fleet, so shore teams focus on the ships and equipment that need it
• Supporting crews when alarms and failures occur, from confirming the signal to organizing the repair
• Supporting class survey arrangements for approved schemes

For fleets short of experienced engineers, the fault support role often delivers the most visible value.

Checklist for Choosing a System

Implementation: Start With the Failures That Cost You Most

1. Review the failure history. List the failures of the past few years by cost, off-hire and repetition. They show where monitoring and fault support will pay off first.
2. Map failures to techniques. For each important failure mode, decide which technique would detect it and whether continuous or periodic measurement fits.
3. Define the shore response before go-live. Decide who triages alerts, who is contacted when, how the ship is informed, and how makers are involved.
4. Allow a baseline period. Expect tuning in the first months, and record every change.
5. Talk to class early if you want survey credit. Z27 schemes need approval, documentation and an implementation survey.
6. Review after every significant failure. Close the loop described in Step 6.

See also How to Test a Ship Management System Before Fleet-Wide Commitment, Retrofit vs Newbuild and 7 Ship Management System Implementation Mistakes to Avoid.

Related Articles

• Remote Vessel Monitoring Systems: How Shore Teams See What's Happening On Board
• Vessel Data Integration Platforms: How to Give Ship and Shore the Same Data
• Performance Monitoring
• Fleet Management Software vs Predictive Maintenance Software