Manufacturing Maintenance & Reliability: A Practical Guide

A modern maintenance and reliability program may include:

• corrective maintenance
• preventive maintenance
• condition-based maintenance
• predictive maintenance
• reliability-centered maintenance
• equipment inspections
• lubrication
• calibration
• vibration monitoring
• thermal imaging
• oil analysis
• spare parts management
• maintenance planning and scheduling
• root cause analysis
• equipment criticality
• work-order management
• maintenance metrics
• CMMS software

The objective is not to eliminate every equipment failure.

That would usually be economically impossible.

The objective is to manage equipment risk so that the right assets receive the right maintenance at the right time.

Key Takeaways

• Reactive maintenance is appropriate for some assets, but dangerous or expensive for others.
• Preventive maintenance is based primarily on time or usage.
• Condition-based maintenance uses the actual observed condition of an asset to determine when action is required.
• Predictive maintenance uses condition data and analysis to anticipate developing failures.
• Reliability-centered maintenance selects maintenance strategies according to asset function, failure consequences, and failure behavior.
• CMMS software organizes maintenance work, assets, history, schedules, parts, and labor. It does not automatically create a good maintenance program.
• MTBF describes failure frequency for repairable equipment. MTTR describes how long restoration or repair typically takes.
• Availability depends on both how often equipment fails and how quickly it can be returned to service.
• Downtime cost includes more than lost machine time.
• Equipment criticality should influence maintenance priority.
• Spare parts are part of the reliability strategy.
• Production and maintenance systems should exchange information.
• Maintainability should be designed into new facilities and equipment before installation.

Maintenance and Reliability Are Not the Same Thing

Maintenance asks:

What work should we perform on this equipment?

Reliability asks:

What must be true for this equipment to perform its required function when the business needs it?

Maintenance is one method of improving reliability.

Others include:

• better equipment design
• better installation
• improved operating procedures
• proper lubrication
• better environmental conditions
• operator training
• improved spare parts
• reduced equipment abuse
• elimination of recurring defects
• redundancy
• condition monitoring

A machine that requires constant maintenance may have a reliability problem that maintenance alone cannot solve.

The Maintenance Strategy Spectrum

The U.S. Department of Energy commonly describes four broad maintenance approaches: reactive, preventive, predictive, and reliability-centered maintenance.

Most plants need some combination of them.

Strategy	Basic Approach
Reactive	Repair after failure
Preventive	Maintain according to time or usage
Condition-Based	Maintain when observed condition indicates need
Predictive	Use condition data and analysis to anticipate failure
Reliability-Centered	Select the appropriate strategy based on function and risk

The goal should not be to make every asset predictive.

The goal is to choose the maintenance strategy that makes sense for each asset.

Reactive Maintenance

This is sometimes called:

• run-to-failure maintenance
• breakdown maintenance
• corrective maintenance

Reactive maintenance is often criticized, but it is not always wrong.

It can be appropriate when:

• the asset is inexpensive
• failure creates little production impact
• replacement is fast
• spare equipment is available
• failure creates no significant safety or environmental risk
• monitoring would cost more than the failure

For example, replacing an inexpensive noncritical ventilation fan after failure may make more sense than developing a sophisticated predictive maintenance program for it.

The decision should be economic and risk-based.

The Problem With Excessive Reactive Maintenance

Reactive maintenance becomes expensive when critical equipment is allowed to fail unexpectedly.

Consequences can include:

• lost production
• overtime
• expedited parts
• emergency contractor charges
• secondary equipment damage
• scrap
• missed deliveries
• quality problems
• safety exposure
• disrupted schedules

Emergency repairs are also difficult to plan efficiently.

Technicians may spend significant time locating:

• documentation
• spare parts
• tools
• vendor support
• replacement equipment

A plant dominated by emergency maintenance usually has very little control over maintenance workload.

Preventive Maintenance

Triggers may include:

• calendar time
• operating hours
• machine cycles
• production quantity
• mileage

Examples include:

• replacing filters every three months
• lubricating a bearing every 500 operating hours
• inspecting belts every month
• testing a safety device annually
• replacing a wear component after a defined number of cycles

Preventive maintenance works well when failure probability has a useful relationship with age, operating time, or usage.

Preventive Maintenance Can Also Be Overdone

More maintenance is not automatically better maintenance.

Unnecessary preventive work can create:

• labor cost
• unnecessary parts replacement
• equipment downtime
• maintenance-induced failures
• premature component replacement

Every intervention introduces some possibility of:

• incorrect reassembly
• contamination
• incorrect adjustment
• damaged connections
• installation error

Preventive maintenance intervals should therefore be based on evidence where practical.

Useful sources include:

• manufacturer recommendations
• regulatory requirements
• equipment history
• failure data
• operating conditions
• engineering analysis

Condition-Based Maintenance

Instead of replacing a component simply because six months have passed, the organization examines indicators of its actual condition.

Examples include:

• vibration
• temperature
• lubricant condition
• electrical characteristics
• pressure
• flow
• ultrasound
• wear measurements
• visual inspection

Maintenance is triggered when the condition indicates deterioration.

Predictive Maintenance

The practical objective is to detect deterioration early enough that maintenance can be planned before functional failure.

Potential benefits include:

• fewer unexpected failures
• better maintenance scheduling
• fewer unnecessary preventive tasks
• better parts planning
• increased equipment availability
• earlier detection of developing defects

Predictive maintenance works best when the failure produces a measurable signal before functional failure occurs.

Not every failure behaves that way.

Condition-Based vs. Predictive Maintenance

These terms are often used interchangeably.

A useful practical distinction is:

Condition-based maintenance responds to the measured condition of the asset.

Predictive maintenance analyzes condition and trends to anticipate how the condition is likely to develop.

For example:

A vibration measurement exceeds a defined threshold and generates a maintenance work order.

That is condition-based maintenance.

Vibration data shows a developing bearing-frequency pattern and analysis indicates that maintenance should be scheduled during the next planned outage.

That is predictive maintenance.

The boundary is not always sharp, and vendors frequently use the terminology differently.

Focus on what the system actually does.

Reliability-Centered Maintenance

RCM does not assume that every failure should be prevented.

It asks questions such as:

• What function must the equipment perform?
• How can that function fail?
• What causes those failures?
• What happens when the failure occurs?
• Does the failure affect safety?
• Does it affect the environment?
• Does it stop production?
• Can the failure be detected before it occurs?
• Is scheduled maintenance technically effective?
• Is preventive action economically justified?
• Should the asset simply be allowed to run to failure?

This produces a maintenance strategy based on consequence rather than habit.

Equipment Criticality

Factors may include:

• safety
• environmental impact
• production impact
• quality impact
• repair cost
• replacement lead time
• customer impact
• availability of backup equipment

A simple criticality framework might classify assets as:

Critical

Failure immediately threatens safety, compliance, major production, or customer commitments.

Important

Failure significantly affects production but alternatives or short-term workarounds exist.

Noncritical

Failure causes limited operational impact and can be repaired or replaced without significant consequence.

Criticality helps determine:

• maintenance priority
• preventive maintenance frequency
• monitoring investment
• spare parts inventory
• redundancy
• response expectations

A $2,000 component can be more critical than a $500,000 machine if its failure stops the entire facility and replacement takes six months.

Failure Modes

A maintenance strategy should consider how an asset can fail.

Failure modes can include:

• bearing wear
• contamination
• overheating
• lubrication failure
• misalignment
• electrical insulation deterioration
• loose connections
• corrosion
• fatigue
• sensor failure
• software failure
• seal leakage
• clogged filters
• mechanical wear

The same asset can have multiple failure modes.

Each may require a different detection or prevention strategy.

Failure Mode and Effects Analysis

For maintenance purposes, FMEA can help identify:

• likely failure modes
• failure consequences
• detection methods
• preventive actions
• monitoring opportunities

FMEA does not replace maintenance history.

It helps provide structure when determining where risk exists and how that risk should be managed.

CMMS

Typical CMMS functions include:

• asset registry
• equipment hierarchy
• work requests
• maintenance work orders
• preventive maintenance schedules
• labor tracking
• parts usage
• spare parts inventory
• vendor information
• maintenance procedures
• equipment documents
• downtime records
• maintenance history
• inspection routes
• maintenance KPIs

A CMMS creates a system of record for maintenance activity.

CMMS Does Not Fix a Bad Maintenance Process

Installing CMMS software does not automatically improve reliability.

If the organization has:

• poor asset data
• incomplete equipment lists
• meaningless PM tasks
• no equipment criticality
• inaccurate spare parts
• weak work-order discipline
• poor failure codes
• no planning process

the CMMS will faithfully preserve those problems in a database.

The process should be defined before software configuration becomes the focus.

Go deeper: CMMS for Manufacturing: Complete Buyer’s and Implementation Guide

Related articles:

• CMMS vs. Spreadsheet Maintenance Tracking
• When Has Your Plant Outgrown Excel for Maintenance?
• How to Build a Preventive Maintenance Schedule
• How Maintenance Data Should Feed Production Operations

The Equipment Hierarchy

A useful CMMS needs a logical asset structure.

For example:

Facility 1

→ Machining Department
→ CNC Cell 2
→ Machine 17
→ Spindle System
→ Spindle Motor

Or:

Facility 1

→ Electrical Distribution
→ Main Switchgear
→ Distribution Panel DP-14
→ Circuit 17
→ Machine 17

The correct level of detail depends on how maintenance work, cost, and failure history need to be tracked.

Creating an asset record for every inexpensive replaceable item can create unnecessary administrative work.

Creating only one record for an entire production line may hide important failure history.

The hierarchy should support decisions.

Maintenance Work Orders

A useful work order should capture enough information to understand both the maintenance activity and the failure.

Typical information includes:

• asset
• problem description
• requested date
• priority
• failure code
• technician
• labor hours
• parts used
• work performed
• cause
• downtime
• completion date
• follow-up requirements

Poor descriptions create poor maintenance history.

A record saying:

Fixed machine.

has almost no long-term analytical value.

A useful record might say:

Replaced failed drive-end bearing after vibration trend indicated increasing outer-race defect frequency. Alignment checked and corrected after installation.

Maintenance history becomes an engineering resource only when the information is meaningful.

Maintenance Planning vs. Scheduling

These are different activities.

Maintenance Planning

Planning determines:

• what work needs to be done
• required procedures
• labor skills
• parts
• tools
• permits
• safety requirements
• estimated duration

Maintenance Scheduling

Scheduling determines:

• when the work will occur
• which technicians will perform it
• when equipment will be available
• how maintenance fits around production

Planning makes work ready.

Scheduling determines when ready work should be performed.

A maintenance job scheduled without the correct parts or instructions is not really ready to execute.

The Maintenance Backlog

The maintenance backlog contains work that has been identified but not completed.

A backlog is not automatically bad.

It provides work that can be planned and prioritized.

The problem occurs when the backlog is:

• unknown
• unprioritized
• continually growing
• filled with duplicate requests
• dominated by old work that nobody intends to perform

Useful backlog management considers:

• criticality
• safety
• production impact
• required shutdown
• parts availability
• labor
• age
• estimated effort

MTBF

A common calculation is:

MTBF = Total Operating Time / Number of Failures

Suppose a machine operates for 2,000 hours and experiences four qualifying failures.

MTBF = 2,000 / 4 = 500 hours

MTBF can be useful for:

• reliability trends
• comparing similar assets
• evaluating improvements
• identifying deteriorating equipment

It should be used carefully.

MTBF is an average.

It does not mean the equipment will fail every 500 hours.

Define What Counts as a Failure

MTBF becomes meaningless if the organization does not consistently define failure.

Does failure include:

• complete machine shutdown?
• reduced production rate?
• loss of quality capability?
• operator reset?
• five-minute interruption?
• sensor fault?

Definitions should be established before comparing the metric over time or between facilities.

MTTR

A common calculation is:

MTTR = Total Repair or Restoration Time / Number of Repairs

If four failures require a total of 12 hours to restore:

MTTR = 12 / 4 = 3 hours

Organizations should explicitly define when the clock starts and stops.

Possible definitions include:

• failure to repair completion
• failure to return to production
• technician arrival to repair completion

Those are different measurements.

MTBF vs. MTTR

MTBF answers:

How frequently does the equipment fail?

MTTR answers:

How quickly can we recover when it fails?

Both matter.

A machine that rarely fails but takes three days to repair may create serious production risk.

A machine that fails frequently but can be restored in five minutes creates a different problem.

Reliability and maintainability should be considered together.

Availability

For a simple repairable system under common assumptions, inherent availability can be approximated by:

Availability = MTBF / (MTBF + MTTR)

If:

MTBF = 500 hours
MTTR = 5 hours

Availability is approximately:

500 / 505 = 99.0%

This formula is useful conceptually, but real operational availability may also be affected by:

• waiting for technicians
• waiting for parts
• administrative delays
• planned maintenance
• logistics
• production scheduling

The exact availability definition should match the business question.

Downtime

Downtime may be:

• planned
• unplanned

Planned Downtime

Examples include:

• preventive maintenance
• calibration
• inspection
• scheduled shutdown
• planned changeover

Unplanned Downtime

Examples include:

• equipment failure
• electrical failure
• tooling failure
• control system failure
• utility interruption
• emergency maintenance

The distinction matters because planned downtime can be coordinated with production.

Unplanned downtime disrupts the schedule.

The True Cost of Downtime

Downtime cost is rarely just:

Hourly machine rate × hours stopped

Potential costs include:

• lost throughput
• idle labor
• overtime recovery
• scrap
• rework
• expedited material
• expedited shipping
• contractor charges
• replacement parts
• missed customer commitments
• downstream starvation
• upstream congestion

The financial impact also depends on whether the failed equipment is actually the production constraint.

If a non-bottleneck machine fails and sufficient inventory exists downstream, immediate plant throughput may not change.

If the bottleneck fails, every lost hour may directly affect output.

Read: How to Calculate the True Cost of Manufacturing Downtime

Maintenance Metrics That Matter

Useful maintenance metrics can include:

Metric	What It Helps Explain
MTBF	Failure frequency
MTTR	Recovery speed
Availability	Ability to operate when needed
Planned Maintenance %	Planned vs. reactive workload
PM Compliance	Whether scheduled PMs are completed
Emergency Work %	Degree of reactive maintenance
Maintenance Backlog	Work waiting to be completed
Schedule Compliance	Whether planned maintenance was executed
Repeat Failures	Effectiveness of repairs
Downtime	Production loss
Maintenance Cost	Financial maintenance burden

No single maintenance KPI describes reliability.

A plant can have excellent PM compliance while repeatedly performing unnecessary PM work.

Measure outcomes as well as activity.

Repeat Failures Matter

One of the most useful reliability signals is recurrence.

If technicians repair the same failure repeatedly, the organization may be restoring function without removing the cause.

Repeat failures should trigger questions such as:

• Is the repair method correct?
• Is installation causing the problem?
• Is the replacement part correct?
• Is lubrication adequate?
• Is alignment correct?
• Is equipment being operated outside its design range?
• Is contamination entering the system?
• Is another failure causing this one?

Fixing the same problem faster is useful.

Preventing the problem is better.

Root Cause Analysis for Equipment Failure

Not every failure requires a major investigation.

Root cause analysis is most valuable for failures that are:

• expensive
• recurring
• safety-related
• quality-related
• environmentally significant
• production-critical

Useful methods can include:

• 5 Whys
• fishbone analysis
• fault tree analysis
• failure mode analysis
• physical failure analysis
• review of operating history
• review of condition-monitoring data

Avoid stopping at conclusions such as:

Bearing failed.

The bearing failure is the event.

The cause may involve:

• contamination
• misalignment
• improper installation
• incorrect lubrication
• overload
• electrical fluting
• poor storage
• inadequate sealing

The maintenance strategy should address the mechanism that created the failure.

Predictive Maintenance Technologies

Different failure modes create different warning signals.

A predictive maintenance program may therefore use several technologies.

Vibration Monitoring

What Does Vibration Monitoring Detect?

Vibration analysis can help identify mechanical problems in rotating equipment such as:

• bearing defects
• imbalance
• misalignment
• looseness
• resonance
• gear problems

Applications commonly include:

• motors
• pumps
• fans
• gearboxes
• compressors
• spindles

A single vibration measurement can provide useful information.

Trend data is often more valuable because it shows how equipment condition changes over time.

Read: Vibration Monitoring for Motors and Rotating Equipment

Thermal Imaging

It can help identify conditions such as:

• overheating electrical connections
• overloaded electrical components
• abnormal bearing temperatures
• insulation problems
• steam trap problems
• mechanical friction
• process temperature abnormalities

Thermography is most valuable when temperature differences can be interpreted in the context of:

• load
• ambient conditions
• equipment history
• similar components

A hot component does not automatically mean a component is failing.

Interpretation matters.

Read: Thermal Imaging for Predictive Maintenance

Lubricant and Oil Analysis

Oil analysis can provide information about both the lubricant and the equipment.

Testing can detect:

• contamination
• water
• wear particles
• viscosity changes
• lubricant degradation
• abnormal metal content

Applications can include:

• gearboxes
• hydraulic systems
• compressors
• engines
• large rotating equipment

Proper sampling technique is essential.

Bad samples produce bad conclusions.

Ultrasound

Ultrasonic inspection can help identify:

• compressed air leaks
• vacuum leaks
• steam trap problems
• bearing problems
• electrical discharge

Compressed air leak detection can be particularly useful because leaks may waste substantial energy while producing no obvious production failure.

Electrical Condition Monitoring

Maintenance programs may also monitor:

• current
• voltage
• power quality
• insulation condition
• motor current signatures
• breaker condition
• connection temperature

Electrical reliability deserves particular attention because electrical failures can affect many assets simultaneously.

Predictive Maintenance Needs Baselines

Condition monitoring is stronger when the organization knows what normal equipment looks like.

Baseline data may include:

• vibration
• temperature
• current
• pressure
• speed
• lubricant condition

Ideally, useful baseline measurements are captured when equipment is:

• newly commissioned
• operating correctly
• properly loaded
• correctly aligned

Without a baseline, the organization may know a measurement changed without knowing whether the starting point was healthy.

Spare Parts Management

Spare parts decisions should consider:

• equipment criticality
• part failure probability
• supplier lead time
• cost
• shelf life
• repairability
• interchangeability
• number of installed assets
• consequences of stockout

A rarely used $10,000 component with a one-year lead time may deserve to sit on the shelf.

A common $20 component available locally within an hour may not require dozens of spares.

Inventory decisions should consider risk rather than unit price alone.

Critical Spare Parts

A critical spare is typically a component whose absence could create unacceptable downtime or risk.

Critical spare analysis should ask:

• What equipment uses it?
• What happens if it fails?
• Is there redundancy?
• How quickly can the part be obtained?
• Can it be repaired?
• Is there an acceptable substitute?
• Does it require special storage?
• Can suppliers guarantee availability?

Vendor claims such as “normally in stock” should not automatically be treated as a reliability strategy.

Lubrication Management

Lubrication problems can shorten equipment life significantly.

A lubrication program should consider:

• correct lubricant
• correct quantity
• correct interval
• contamination control
• storage
• labeling
• application method
• compatibility

More grease is not automatically better.

Overlubrication can damage equipment just as underlubrication can.

Lubrication tasks should be specific enough that technicians know exactly what is required.

Maintenance and Production Must Share Information

Maintenance cannot operate effectively as an isolated department.

Production systems know:

• whether equipment is running
• what job is being produced
• machine operating hours
• production counts
• downtime events
• current schedule

Maintenance systems know:

• equipment condition
• open work orders
• preventive maintenance due dates
• asset availability
• failure history
• repair estimates

Connecting those views improves both departments.

Read: How Maintenance Data Should Feed Production Operations

CMMS and MES Integration

A useful integration might work like this:

1. MES detects or records that Machine 17 has stopped.
2. The failure is classified as maintenance-related.
3. CMMS creates or receives a maintenance request.
4. Maintenance diagnoses the problem.
5. CMMS records repair information, labor, and parts.
6. Equipment is returned to service.
7. MES receives updated equipment availability.
8. Scheduling reacts to the production interruption if necessary.

The objective is to avoid requiring people to manually recreate the same event in several systems.

Machine Data and Maintenance Data Are Different

A PLC may report:

Motor current = 38.6 A

A maintenance system needs to know:

• which asset
• what component
• operating condition
• expected range
• historical trend
• work history
• associated failure modes

Raw telemetry becomes useful maintenance information only when it has context.

This is the same data-contextualization issue that appears throughout manufacturing systems architecture.

Maintenance Safety

Maintenance work often exposes employees to hazards that are not present during normal production.

Potential energy sources include:

• electrical
• mechanical
• hydraulic
• pneumatic
• thermal
• chemical
• gravity
• stored pressure

For U.S. general industry, OSHA’s Control of Hazardous Energy standard, commonly called lockout/tagout or LOTO, establishes requirements for protecting employees from unexpected equipment energization, startup, or release of stored energy during servicing and maintenance.

Maintenance planning therefore must include safety, not just technical repair instructions.

A work order should never be treated as authorization to bypass required energy-control procedures.

Electrical Maintenance

Electrical distribution equipment is itself a critical manufacturing asset.

Examples include:

• switchgear
• transformers
• switchboards
• panelboards
• circuit breakers
• motor-control centers
• busway
• UPS systems
• generators
• automatic transfer systems

Failure can affect many production assets at once.

NFPA 70B provides a recognized framework for electrical equipment maintenance.

Electrical maintenance programs may involve:

• inspections
• cleaning
• testing
• torque checks where appropriate
• breaker maintenance
• thermography
• condition assessment
• documentation

Electrical maintenance should be performed under applicable safety requirements and by appropriately qualified personnel.

Maintenance for a New Manufacturing Facility

Maintenance should be involved before production equipment is purchased and installed.

A new facility creates an opportunity that existing plants rarely have:

maintainability can be designed into the plant.

Important questions should be asked during equipment specification, layout, construction, installation, and commissioning.

Design for Maintenance Access

Equipment should provide enough space to:

• open electrical cabinets
• remove motors
• replace pumps
• remove spindles
• pull filters
• remove belts
• access lubrication points
• inspect components
• use lifting equipment
• perform required testing

A machine may physically fit into a location while being nearly impossible to maintain there.

The layout should consider the maintenance envelope, not just the operating footprint.

Design Equipment Removal Paths

Ask:

If the largest maintainable component fails, how will we remove it?

Consider:

• doors
• roof access
• overhead clearance
• cranes
• forklifts
• removable panels
• aisle widths
• structural obstacles

A large motor that can only be removed by dismantling half the production line creates a maintainability problem that will eventually become a production problem.

Plan Energy Isolation

Maintenance access should include practical methods for safely isolating energy sources.

Depending on the equipment, this can involve:

• electrical disconnects
• pneumatic isolation
• hydraulic isolation
• steam isolation
• gas isolation
• gravity control
• stored-energy discharge

Isolation points should be identifiable and accessible.

Lockout/tagout requirements should be considered during equipment acceptance rather than discovered after commissioning.

Plan Electrical Maintainability

Electrical infrastructure should consider:

• equipment accessibility
• working clearances
• labeling
• current single-line diagrams
• spare breaker capacity
• test access
• infrared inspection access where appropriate
• protective-device maintenance
• equipment condition monitoring
• expansion

Electrical rooms should not gradually become storage rooms.

They are operating infrastructure.

Consider Redundancy Strategically

Not every system requires redundancy.

Some do.

Examples can include:

• compressed air
• process cooling
• network infrastructure
• critical pumps
• UPS systems
• controls
• production bottlenecks
• environmental systems

Redundancy should be based on the consequence of failure.

Two machines do not create meaningful redundancy if both depend on the same single point of failure upstream.

Identify Single Points of Failure

During design, ask:

What single component could stop this entire process or facility?

Possible examples include:

• one transformer
• one compressor
• one chilled-water pump
• one network switch
• one server
• one control panel
• one specialty tool
• one inspection machine
• one exhaust system

Once identified, decide whether the risk should be addressed through:

• redundancy
• spare parts
• preventive maintenance
• condition monitoring
• rapid replacement
• contingency planning

Build the Asset Register During Construction

Do not wait until the facility is operating to determine what equipment exists.

Create asset records as equipment is purchased and commissioned.

Capture information such as:

• manufacturer
• model
• serial number
• asset number
• location
• installation date
• warranty
• supplier
• manuals
• drawings
• spare parts
• recommended maintenance
• electrical requirements
• utility requirements

The facility handover should populate the maintenance system rather than deliver several cabinets full of documentation that nobody enters later.

Load Preventive Maintenance Before Startup

Vendor maintenance requirements should be reviewed during commissioning.

Before production ramps, establish:

• required PM tasks
• frequencies
• responsible trades
• parts
• consumables
• procedures
• safety requirements

Maintenance should not discover six months later that an important warranty-required service interval was missed.

Capture Baseline Condition Data

Commissioning provides an excellent opportunity to record baseline information.

Depending on the equipment, this could include:

• vibration
• thermal images
• alignment
• electrical measurements
• oil samples
• pressures
• temperatures
• flow rates

Future measurements can then be compared with known-good operating conditions.

Establish Critical Spares Before Production

Equipment procurement should include spare parts analysis.

Ask suppliers:

• Which parts commonly fail?
• Which have long lead times?
• Which are proprietary?
• Which parts become obsolete quickly?
• What should be stocked locally?
• What can be sourced commercially?
• What programming or configuration files are required?
• What happens if the supplier disappears?

Do not wait for the first breakdown to learn that a proprietary controller has a 20-week lead time.

Preserve Software and Configuration

Modern manufacturing equipment contains software.

Maintenance documentation should therefore include:

• PLC programs
• HMI programs
• drive parameters
• robot programs
• controller configuration
• machine recipes
• firmware information
• network configuration
• vendor software requirements
• backup procedures

Appropriate backups should be maintained before equipment enters production.

A failed industrial computer should not require reconstructing the machine configuration from scratch.

Plan Maintenance Workspaces

A manufacturing facility may require:

• maintenance shop
• electrical bench
• mechanical repair area
• welding area
• tool storage
• parts storage
• lubricant storage
• battery charging
• calibration storage
• contractor staging

These spaces should support the work actually expected at the facility.

Commission Documentation as Seriously as Equipment

Facility turnover should include current:

• electrical drawings
• piping drawings
• network drawings
• control drawings
• equipment manuals
• spare parts lists
• certifications
• warranties
• software backups
• maintenance procedures
• inspection requirements

Documents should be searchable and controlled.

A PDF buried on someone’s laptop is not an effective maintenance system.

CMMS vs. Spreadsheet Maintenance Tracking

Spreadsheets can work for small maintenance programs.

They become increasingly difficult when the organization needs:

• multiple technicians
• recurring work orders
• asset history
• mobile access
• parts inventory
• approvals
• work requests
• labor tracking
• attachments
• notifications
• audit history
• equipment hierarchies
• integrations

At that point, CMMS software usually provides a stronger system of record.

Read: CMMS vs. Spreadsheet Maintenance Tracking

When Has a Plant Outgrown Excel for Maintenance?

Warning signs include:

• technicians maintain separate spreadsheets
• PMs are missed
• nobody knows which list is current
• equipment history is difficult to reconstruct
• maintenance requests arrive through email and text messages
• parts are frequently unavailable
• failure history cannot be analyzed
• management cannot distinguish planned from emergency work
• the same repair is repeatedly performed without recognizing the pattern

The trigger for CMMS should be operational complexity rather than company size.

Read: When Has Your Plant Outgrown Excel for Maintenance?

Choosing a Maintenance Strategy

A practical decision process can begin with five questions.

What Happens If the Asset Fails?

Consider:

• safety
• environment
• production
• quality
• customer impact
• repair cost

Can the Failure Be Detected in Advance?

If a measurable warning condition exists, condition monitoring may be useful.

Does Failure Probability Increase With Age or Usage?

If so, scheduled preventive replacement may be effective.

Is the Failure Random?

Time-based replacement may provide little value for random failure modes.

Is Run-to-Failure Acceptable?

If the consequence is small and restoration is easy, reactive maintenance may be appropriate.

Maintenance strategy should follow failure behavior and consequence.

A Practical Maintenance Improvement Process

Step 1: Build the Asset Register

Know what equipment exists.

Step 2: Establish Asset Criticality

Determine what matters most.

Step 3: Review Existing Maintenance

Identify:

• current PMs
• emergency work
• recurring failures
• maintenance costs
• downtime

Step 4: Identify Important Failure Modes

Understand how critical assets can lose their required functions.

Step 5: Select Maintenance Strategies

Choose appropriate combinations of:

• reactive
• preventive
• condition-based
• predictive

Step 6: Establish Work Management

Define:

• work requests
• priorities
• planning
• scheduling
• execution
• closure

Step 7: Improve Spare Parts Management

Identify critical parts and supplier risks.

Step 8: Capture Useful Failure Data

Improve work-order descriptions and failure coding.

Step 9: Measure Reliability

Track failure, recovery, downtime, and recurrence.

Step 10: Eliminate Chronic Problems

Use root cause analysis where justified.

Step 11: Expand Condition Monitoring

Apply predictive techniques where the economics support them.

Step 12: Continuously Review the Strategy

Equipment ages.

Production requirements change.

Maintenance strategies should change with them.

Common Maintenance Mistakes

Treating Every Asset the Same

Maintenance effort should reflect equipment criticality and failure behavior.

Creating PM Tasks Without Reviewing Them

Old preventive maintenance tasks can survive for decades simply because nobody questions them.

Measuring PM Completion Instead of Reliability

Completing every PM does not prove the equipment is reliable.

Performing Too Much Preventive Maintenance

Unnecessary intervention consumes labor and can introduce failures.

Buying Predictive Technology Before Defining the Problem

Sensors do not create reliability by themselves.

Determine which failure mode you are trying to detect.

Collecting Condition Data Without Acting on It

A vibration alarm nobody reviews has little value.

Ignoring Spare Parts

Diagnostic excellence does not shorten downtime if the replacement component takes three weeks to arrive.

Recording Poor Work-Order History

Useful analysis depends on useful records.

Repeatedly Repairing Without Finding the Cause

A recurring failure deserves engineering attention.

Ignoring Maintainability During Equipment Purchase

The cheapest machine to purchase may be expensive to own if it is difficult to service.

Separating Maintenance From Production

Maintenance needs production information.

Production needs equipment condition and availability.

Treating Safety as a Maintenance Procedure Step

Safety requirements need to be built into work planning, equipment design, and energy-isolation practices.

Frequently Asked Questions

What is preventive maintenance?

Preventive maintenance performs defined maintenance tasks according to time, operating hours, cycles, or another predetermined interval to reduce the probability of failure or degradation.

What is predictive maintenance?

Predictive maintenance uses equipment condition data and analysis to identify developing problems and help determine when maintenance intervention should occur.

What is condition-based maintenance?

Condition-based maintenance triggers work according to the observed condition of the equipment rather than only a calendar or usage interval.

What is reliability-centered maintenance?

Reliability-centered maintenance is a structured approach for selecting maintenance strategies according to equipment function, failure modes, failure consequences, and the effectiveness of available maintenance actions.

What does CMMS stand for?

CMMS stands for Computerized Maintenance Management System.

It manages assets, work orders, preventive maintenance, labor, parts, documentation, and maintenance history.

What is MTBF?

MTBF means Mean Time Between Failures.

It measures the average operating time between defined failures of repairable equipment.

What is MTTR?

MTTR commonly means Mean Time to Repair or Mean Time to Restore.

It measures how long restoration typically takes after a failure.

Organizations should define exactly how they calculate it.

What is the difference between MTBF and MTTR?

MTBF measures how often equipment fails.

MTTR measures how quickly it can be restored.

Improving either can increase equipment availability.

Is higher MTBF always better?

Generally, higher MTBF means failures occur less frequently.

The metric still needs context.

The impact of each failure, repair duration, asset criticality, and consistency of the failure definition also matter.

What is the best maintenance strategy?

There is no single best strategy.

Critical equipment may justify preventive or predictive maintenance.

Low-consequence equipment may appropriately run to failure.

The strategy should match the failure mode, consequence, detectability, and economics.

Does every machine need predictive maintenance?

No.

Predictive monitoring should be applied where developing failures can be detected and where avoiding those failures creates enough value to justify the monitoring effort.

Can a manufacturer manage maintenance in Excel?

Yes, for a small and relatively simple maintenance operation.

A CMMS becomes increasingly useful as the number of assets, technicians, recurring tasks, parts, records, and coordination requirements grows.

How should maintenance information connect with production?

Production systems should communicate information such as machine status, operating hours, production counts, and downtime.

Maintenance systems should provide equipment availability, planned outages, work-order status, and equipment condition.

The systems should exchange the information required by the process without duplicating unnecessary records.

Practical Tools and Templates

Custom Industrial Solutions will provide practical resources including:

• Maintenance Strategy Selection Worksheet
• Asset Criticality Matrix
• Preventive Maintenance Schedule Template
• MTBF Calculator
• MTTR Calculator
• Equipment Availability Calculator
• Manufacturing Downtime Cost Calculator
• Maintenance Backlog Prioritization Worksheet
• Predictive Maintenance Technology Selection Guide
• Critical Spare Parts Analysis
• CMMS Requirements Checklist
• Maintenance Work Order Template
• Equipment Commissioning Data Sheet
• New Facility Maintainability Checklist
• Asset Register Template

Authoritative Frameworks and References

ISO 55000

ISO 55000 provides principles and terminology for managing physical and other assets throughout their life cycles.

Asset management is broader than maintenance and considers how assets create value for the organization.

OSHA Control of Hazardous Energy

OSHA 29 CFR 1910.147 establishes requirements for controlling hazardous energy during applicable servicing and maintenance activities in U.S. general industry.

NFPA 70B

NFPA 70B provides requirements for establishing and carrying out electrical equipment maintenance programs.

Reliability-Centered Maintenance

RCM provides a structured method for selecting failure-management strategies according to asset function and failure consequences.

Condition Monitoring

Technologies such as vibration analysis, thermography, oil analysis, ultrasound, and electrical monitoring provide methods for identifying developing equipment conditions.

Standards and technologies provide tools.

A good reliability program still depends on understanding the equipment, its operating context, and the consequences of failure.

The Bottom Line

Manufacturing maintenance is not simply the department that fixes machines after production calls.

It is part of the manufacturing system.

Reactive maintenance restores failed equipment.

Preventive maintenance attempts to reduce predictable failures.

Condition-based maintenance responds to measured equipment condition.

Predictive maintenance attempts to identify developing failures early enough to plan intervention.

Reliability engineering asks a larger question:

What combination of design, operation, maintenance, monitoring, spare parts, and redundancy will allow this asset to perform its required function when production needs it?

The best maintenance program does not maximize maintenance activity.

It provides the required equipment reliability at an acceptable level of cost and risk.

For a new facility, that work starts before the first machine is installed.

Maintainability, isolation, access, spare parts, asset data, documentation, monitoring, utility redundancy, and equipment recovery should be designed into the facility while those decisions are still inexpensive to change.

Explore Manufacturing Maintenance & Reliability

Manufacturing Maintenance Management: From Reactive to Predictive

How reactive, preventive, condition-based, predictive, and reliability-centered maintenance strategies fit together and how manufacturers can move toward more proactive maintenance.

CMMS for Manufacturing: Complete Buyer’s and Implementation Guide

How CMMS systems work, what information they should contain, how to structure assets and work orders, and how to avoid automating a weak maintenance process.

Manufacturing Downtime & Reliability Metrics: MTBF, MTTR, Availability and Cost

How to measure equipment reliability, define failures, calculate availability, quantify downtime, and use maintenance metrics without letting the metrics become the objective.

Need Help With Maintenance Systems or Reliability Data?

Custom Industrial Solutions helps manufacturers map maintenance processes, define CMMS requirements, connect maintenance and production information, structure asset data, and develop practical applications where existing systems do not fit the operation.

Good maintenance keeps the equipment running.

Good reliability engineering makes failures less frequent, less disruptive, and easier to recover from.