Scheduled or Condition-Based Maintenance: Which Assemblies Suit Each Approach ART-412

Short answer

Keep a task on the scheduled plan when the manufacturer specifies it, when an assembly has a defined service life, or when skipping the task could have serious consequences. This is especially important for safety-related functions. Condition monitoring is appropriate when a change can be observed objectively and repeatedly, and when the decision does not depend on one person’s impression. For most laser and metalworking systems, the practical answer is a combined plan: mandatory scheduled checks supported by a consistent condition and event log.

There is no honest universal interval for an entire machine. The correct interval depends on the model, manufacturer documentation, operating hours, environment, materials, number of shifts, configuration, and consequences of failure. General guidance can explain how to make a decision, but it cannot replace the operating manual for the installed equipment, the service schedule, or a site-specific risk assessment.

How the approaches differ

Scheduled preventive maintenance is triggered by time, cycle count, operating hours, or another condition defined for a particular assembly. Its purpose is to complete specified work before risk becomes unacceptable. A manufacturer may, for example, require an inspection or replacement after a defined interval even when the operator has not yet noticed a problem.

Condition-based maintenance starts with a changing sign. That sign may be a repeated system message, deterioration of a test part, increasing vibration, a temperature change, or another parameter whose measurement is allowed by the equipment documentation. This approach does not mean waiting for a breakdown. It makes it possible to plan an action after a trend has been confirmed.

Predictive maintenance is a more demanding form of this logic: decisions rely on accumulated data and analysis of trends. A small production site does not have to begin with a complex analytical platform. A well-designed log, a stable inspection criterion, and a clear escalation route already create a useful basis for better decisions.

How to choose an approach for an assembly

Assess four characteristics rather than relying on the name of the part. First, consider the consequence of failure: safety, quality, downtime, damaged material, or harm to another assembly. Second, check whether the manufacturer gives a mandatory requirement. Third, determine whether a reliable condition signal can be obtained without unsafe intervention. Fourth, account for the time needed to obtain a spare, diagnose the issue, and restore operation.

When the consequence of failure is high or the manufacturer gives a firm requirement, the scheduled plan remains mandatory. When an approved signal can reveal condition, add observation to that plan. If the signal is weak, subjective, or appears only after damage has already occurred, the assembly should not be maintained only according to condition.

| Assembly characteristic | Main approach | Reason | What must be confirmed | |---|---|---|---| | Failure could create a hazard | Scheduled work plus a function check | Waiting for a symptom is unacceptable | Operating manual, risk assessment, and authorized specialist’s role | | The manufacturer specifies a life or interval | Scheduled | The limit is defined for the model | Current document revision and the event from which the interval is counted | | A stable measurable signal exists | Condition-based plus a log | The trend matters | Measurement method, instrument, and response criterion | | Load changes substantially | Combined | A calendar does not reflect actual operation | Operating hours, materials, shifts, and environment | | Replacement is simple and the consequence is low | Inspection when justified | Date-based replacement may add no value | Downtime cost, spare availability, and inspection limit | | The assembly affects output quality | Combined | Both a schedule and output verification are needed | Test part, measurement, and acceptance criterion |

This is a decision matrix, not a universal list of components for every laser machine. The assembly names and permitted actions must be checked against the documentation for the installed model.

What normally remains scheduled

The scheduled plan includes tasks explicitly required by the manufacturer and tasks connected to machine safety. Depending on the specific equipment, these may include verification of protective functions, guards and interlocks, utilities, cooling, extraction, lubrication, or other systems. Every action must remain within the equipment instructions and the competence of the person carrying it out.

Some inspections are needed even when no symptom is present: reviewing the alarm history, inspecting the work area, confirming that emergency information is available, checking labeling, or verifying the shift-handover process. Such inspections do not authorize a person to open covers or work on electrical parts, optics, or the laser path without an approved procedure.

The advantage of a calendar is predictability: a service window can be reserved in advance. Its weakness is that it may replace a component that is still serviceable, while a fault can also emerge between two planned visits. Scheduled maintenance therefore does not remove the need to report abnormalities during normal operation.

Where condition monitoring is useful

Condition monitoring is useful when a signal is connected to the result and can be reproduced. In production, useful signs may include deterioration of a test part, a change in repeatability, a growing number of system messages, an unusual cycle time, or repeated stops. A technical specialist may add approved measurements when the manufacturer permits them and an appropriate instrument is used.

Do not record only “works” or “does not work.” Record the context: machine model, program, material and thickness, date, shift, operating hours, the event before the symptom, and the verification method. Without this context, a change in result may be wrongly attributed to one assembly when the actual cause is material, process settings, environment, or another part of the line.

The operator’s log should record a sign, not present an unsupported diagnosis. The operator’s role is to notice a deviation, follow the permitted safe-stop procedure, and pass the facts to the responsible person. Repair, calibration, access to hazardous areas, and changes to service parameters belong to another level of work.

How to build a combined plan

Begin with an assembly register, not a generic calendar template. For each row, record the function, possible failure, consequence, basis for the requirement, permitted monitoring method, responsible role, and action after a deviation. If a requirement has no confirmed basis or understandable criterion, mark it for clarification rather than presenting it as settled.

Divide the work into three levels. The first level contains tasks performed according to the manufacturer’s schedule. The second contains observations that an operator or shift supervisor can make without entering hazardous assemblies. The third contains diagnostics and service tasks requiring authorized specialists. This division prevents a simple visual inspection and a task requiring special competence from appearing as equivalent items in one list.

Define the route after a signal appears: who reviews the record, when operation stops, who appoints a diagnostic check, where the report is stored, and who authorizes return to production. Do not invent universal periods such as “once a month” without documented justification. If the frequency is still unconfirmed, label it as a working assumption and obtain confirmation for the specific model.

What the condition log should contain

A minimum record must allow the event to be reconstructed. Include the date and time, equipment identifier, shift, program or operation, material, symptom, conditions before the sign appeared, permitted check performed, result, decision, and responsible person. For a numerical parameter, add the instrument, units, and measurement method.

The log is useful for more than troubleshooting. It shows which deviations repeat, which actions actually remove a symptom, whether load is changing, and where downtime arises. The records must, however, be comparable. A cycle time recorded today for one material cannot be compared directly with a different material tomorrow without accounting for the change.

Without a consistent history, do not create the appearance of predictive maintenance. First establish stable fields and responsibility for records. Then decide whether the accumulated data are sufficient to move from observation toward forecasting.

The roles of the manufacturer and service partner

The manufacturer defines requirements for a particular model and configuration: mandatory work, allowable intervals, performer competence, and the permitted limits of intervention. Read these requirements together with the documentation for the installed machine. A schedule from one model must not be transferred automatically to another, even when the equipment has a similar purpose.

A service partner can help adapt requirements to the site’s actual operating pattern, perform diagnostics, review trends, and document the result. Before agreeing on the work, clarify which tasks are mandatory, which are recommendations, what information is provided after the visit, and what the site’s own staff may perform. The boundaries of scheduled and condition-based service always depend on the equipment, its documentation, and the service agreement.

A simple rule for assigning work

Ask three questions for every task. Has the manufacturer specified a time or trigger? If so, do not remove the task from the calendar merely because no symptom is visible. Is there a safe, repeatable signal that can be monitored without accessing a hazardous assembly? If so, condition monitoring can supplement the schedule. Would failure cause major downtime or quality risk? If so, provide an alternative route, planned diagnostics, or service escalation even while the current indicator remains normal.

If none of the questions has a confirmed answer, do not invent a maintenance method. Add the assembly to an open-issues list and request model-specific documentation. That is safer than creating a local rule that conflicts with the manufacturer’s requirements or gives employees a false sense of control.

How to revise an interval

Do not shorten an interval after every isolated failure. First check whether operating load, shifts, material, environment, operator, configuration, or procedure changed. A failure may have followed an external event, and a more frequent inspection will not remove its cause.

Lengthening an interval also requires evidence: a repeatable history, absence of critical events, and confirmation that the manufacturer does not impose a stricter requirement. Record the date, justification, responsible person, review trigger, and approval status. A temporary assumption must not be presented as a new official schedule.

After repair or replacement, verify whether the symptom disappeared, but do not attribute the entire result to one action without evidence. If material, program, or operator also changed, record that change as a factor affecting the conclusion.

Common mistakes

• using one calendar for every model and configuration;
• replacing a part by date without considering the manufacturer’s requirement and actual condition;
• waiting for a breakdown and calling it condition-based maintenance;
• treating a subjective impression as a numerical measurement;
• failing to record material, program, operating hours, and the conditions in which a symptom appeared;
• assigning diagnostics or hazardous intervention to an operator without an approved procedure;
• shortening intervals after an isolated event without analyzing the cause;
• applying a manufacturer’s schedule for another model to the installed equipment;
• writing “continue monitoring” without a parameter, responsible role, and review date.

Checklist before approving the program

• [ ] Every assembly has a documented basis for its requirement or a “confirmation required” status.
• [ ] Scheduled work, operator observations, and service diagnostics are separated.
• [ ] Consequences of failure for safety, quality, and downtime have been assessed.
• [ ] Every condition check has a parameter, method, instrument, and response criterion.
• [ ] The log identifies the model, date, conditions, symptom, and responsible person.
• [ ] An escalation route and a stop-work condition are defined.
• [ ] Hazardous work is not assigned to personnel without the necessary procedure and competence.
• [ ] Intervals are not described as universal when equipment documentation does not confirm them.
• [ ] The program version has a date, owner, and review process.

What cannot be decided without data

Exact service intervals and a task list for a specific laser machine cannot be determined without its model, configuration, operating hours, work pattern, environment, manufacturer documentation, and event history. Nor can anyone guarantee that an assembly is safe to maintain only according to condition without a reliable signal and an agreed method.

General industry practices provide a framework but do not replace the operating manual for the installed model. If documents give conflicting requirements, a competent technical owner should resolve the issue with the manufacturer or an authorized service partner. This article is not a repair, calibration, electrical, optical, gas-system, or laser-safety procedure.

Summary

Scheduled maintenance and condition monitoring are not mutually exclusive. The calendar covers work required by the manufacturer, tied to a defined life, or carrying a high consequence if skipped. Condition monitoring reveals trends where there is a reliable signal and a clear decision criterion. For laser and metalworking equipment, a combined program with an assembly register, consistent log, assigned roles, and service escalation is usually the practical model.

The quality of the program is not measured by the number of checklist items. It depends on whether the company knows what is checked, why it is required, who may do it, what the limits are, and what happens after a deviation. Confirm the requirements of the specific machine first, then adapt the plan to the site’s real operating conditions.

Need a service consultation?