Maintenance, Reliability and Asset Management

Introduction

Maintenance is no longer viewed merely as a repair activity. In advanced operations management, it is a coordinated system for preserving the capability, safety, quality and economic value of physical assets. Decisions about maintenance affect production capacity, delivery reliability, employee safety, energy consumption, environmental performance and total cost of ownership.

A modern maintenance function therefore asks four related questions:

1. What function must an asset perform?
2. How and why might it fail?
3. What maintenance strategy best manages the consequences of failure?
4. How should the asset be acquired, operated, improved and eventually replaced?

This chapter evaluates corrective, preventive, predictive and reliability-centred maintenance; develops key reliability measures; explains failure analysis, Overall Equipment Effectiveness (OEE) and Total Productive Maintenance (TPM); and applies asset lifecycle management to operational decision-making.

Learning objectives

By the end of this chapter, the learner should be able to:

• distinguish major maintenance strategies and select between them;
• calculate and interpret reliability, availability, maintainability, MTBF and MTTR;
• apply failure mode and effects analysis and basic root-cause analysis;
• calculate OEE and identify its operational losses;
• explain the principles and implementation requirements of TPM;
• use reliability-centred maintenance (RCM) to align tasks with failure consequences;
• evaluate asset decisions across the complete lifecycle; and
• assess the limitations, risks and governance requirements of maintenance systems.

1. Maintenance as an operations capability

An asset is any physical resource that provides a useful operational function. Examples include production machines, vehicles, buildings, utilities, information-technology infrastructure and medical equipment. Maintenance comprises the technical, administrative and managerial actions required to retain an asset in, or restore it to, a state in which it can perform its required function.

The purpose of maintenance is not to keep every component in perfect condition. It is to achieve the required level of performance at an acceptable combination of cost, risk and sustainability. This distinction is important because excessive maintenance can create unnecessary cost and introduce maintenance-induced failures, while insufficient maintenance can cause breakdowns, unsafe conditions and quality losses.

Maintenance decisions should be aligned with operations strategy. A high-volume process with little spare capacity may require high availability and planned redundancy. A low-utilisation facility may rationally accept occasional corrective maintenance. A safety-critical asset, however, may require stringent controls even when its failure probability is low.

Important terminology

• Failure: the termination of an asset’s ability to perform a required function.
• Fault: an abnormal condition that may lead to failure but does not necessarily mean total loss of function.
• Breakdown: a failure that interrupts operation.
• Maintenance task: an action intended to prevent, detect, control or correct failure.
• Availability: the proportion of required time for which an asset is capable of performing its function.
• Reliability: the probability that an asset performs its required function for a specified time under stated conditions.
• Maintainability: the probability that maintenance can be completed within a specified time under stated conditions.
• Asset criticality: the relative importance of an asset based on consequences such as safety, environment, production, quality and cost.

2. Maintenance strategies

2.1 Corrective maintenance

Corrective maintenance restores an asset after a fault or failure has been detected. It may be run-to-failure, where no intervention occurs until functional failure, or deferred corrective maintenance, where a known defect is repaired at a planned time.

Corrective maintenance is appropriate when failure consequences are minor, repair is rapid, spare parts are inexpensive and failure does not create unacceptable safety, quality or environmental risk. It is often reasonable for non-critical lighting, low-cost consumables and redundant equipment.

Its disadvantages include unplanned downtime, secondary damage, emergency procurement, overtime and disruption to production. Run-to-failure is not equivalent to neglect: it should be a deliberate, risk-based decision supported by a repair plan.

2.2 Preventive maintenance

Preventive maintenance is performed at predetermined intervals or according to prescribed criteria to reduce the probability of failure or degradation. Examples include lubrication every 500 operating hours, inspection each month and replacement after a specified number of cycles.

Time-based maintenance is effective when failure risk increases predictably with age or usage. However, many modern components do not exhibit a simple age-related failure pattern. Replacing a healthy component too early may waste useful life and may introduce installation errors.

Preventive maintenance should therefore be based on evidence, manufacturer information, operating conditions and failure history—not merely on a calendar.

2.3 Predictive and condition-based maintenance

Predictive maintenance uses measurements of actual asset condition to determine when intervention is needed. Condition indicators include vibration, temperature, oil particles, electrical signatures, pressure, acoustics, corrosion and performance trends.

A predictive programme normally includes:

1. identifying failure modes that produce detectable degradation;
2. selecting a technically valid condition indicator;
3. establishing a baseline and alarm limits;
4. collecting reliable data at an appropriate frequency;
5. diagnosing the likely failure mechanism; and
6. scheduling corrective action before functional failure.

Predictive maintenance is valuable when the warning period is sufficient, the measurement is trustworthy and the cost of monitoring is justified. It is not automatically superior to preventive maintenance. A complex sensor system can generate false alarms, data overload and maintenance costs without improving decisions.

2.4 Reliability-centred maintenance

RCM is a structured process for determining what must be done to ensure that an asset continues to perform its required functions in its operating context. It focuses on functions, functional failures, failure modes, effects and consequences.

A typical RCM analysis asks:

• What are the asset’s required functions and performance standards?
• In what ways can the asset fail to meet those functions?
• What causes each functional failure?
• What happens when each failure occurs?
• Why does the failure matter: safety, environmental, operational or economic consequences?
• What proactive task, failure-finding task, redesign or deliberate run-to-failure policy is justified?

RCM does not mean maintaining everything more frequently. It may recommend condition monitoring, scheduled restoration, scheduled discard, functional testing, redesign or no scheduled maintenance. Its strength is the explicit connection between maintenance tasks and failure consequences.

2.5 Strategy comparison

3. Reliability engineering and measures

3.1 Reliability function

Reliability is a probability, not a guarantee. If T is the random time to failure, the reliability function is:

Here, R(t) is reliability at time t, P(T > t) is the probability that failure time exceeds t, and F(t) is the cumulative distribution function of failure time. Time may be measured in hours, cycles or kilometres, provided the unit is used consistently.

For an exponential failure distribution with a constant failure rate λ:

The exponential model assumes a constant hazard rate and is therefore most appropriate for the useful-life period of certain repairable items, not automatically for every asset. Under this model, the mean time to failure is:

MTTF applies to non-repairable items. For repairable assets, the more common measure is MTBF.

3.2 MTBF, MTTR and availability

For a repairable asset operating under broadly stable conditions:

MTBF is an average operating time between inherent failures. It should not be interpreted as a guaranteed life or as proof that failures occur at regular intervals.

Mean Time To Repair is:

MTTR includes the time required to restore the asset under the selected measurement definition. Organisations must state whether waiting for technicians, parts and permits is included.

A commonly used inherent or steady-state availability approximation is:

This assumes alternating operating and repair periods, stable averages and consistent definitions. Operational availability may additionally include logistics, administrative delays, planned maintenance and supply delays; it should not be confused with the simpler inherent-availability approximation.

Worked example: availability

A packaging machine operates for 1,800 hours during a year and experiences six corrective failures. Corrective repair time totals 30 hours.

MTBF = 1,800 ÷ 6 = 300 hours.

MTTR = 30 ÷ 6 = 5 hours.

Therefore:

A = 300 ÷ (300 + 5) = 0.9836, or approximately 98.36%.

The result is an average performance indicator. It does not show whether failures are concentrated in one period, whether planned downtime is excluded or whether production was actually available when required.

3.3 Failure rate and hazard patterns

The hazard rate describes the instantaneous risk of failure conditional on survival to a given time. Failure behaviour is often represented by the “bathtub curve”:

• early-life failures: high but decreasing hazard, often linked to design, installation or manufacturing defects;
• useful life: approximately random failures with relatively stable hazard; and
• wear-out: increasing hazard caused by fatigue, corrosion, erosion, ageing or wear.

The bathtub curve is a conceptual model, not a universal law. Different components may display different patterns, and some electronic components do not benefit from age-based replacement. Maintenance policy must be based on actual failure behaviour.

4. Failure analysis and maintenance decision-making

4.1 Failure Mode and Effects Analysis

FMEA identifies possible failure modes, their causes and their effects before or during operation. A conventional FMEA records the failure mode, local effect, system effect, cause, existing controls, recommended action and responsible owner.

Some organisations use a Risk Priority Number:

where S is severity, O is occurrence and D is detectability, each scored according to an organisational scale. The product is only a prioritisation aid. It is ordinal-score arithmetic, not a true probability of risk. A low RPN must not override a catastrophic safety consequence, and two risks with the same RPN may require very different controls.

A stronger approach supplements the score with explicit consequence categories, legal requirements, safeguards and expert judgement.

4.2 Root-cause analysis

Root-cause analysis investigates why a failure occurred and why it was not prevented or detected. Useful methods include:

• five whys: progressive questioning, provided each answer is evidence-based;
• fishbone diagrams: classification of possible causes under people, methods, machines, materials, measurement and environment;
• fault-tree analysis: deductive analysis from an unwanted top event to combinations of lower-level causes; and
• event-tree analysis: forward analysis of possible consequences after an initiating event.

A symptom is not necessarily a root cause. For example, “bearing overheated” describes an effect; “incorrect lubrication quantity caused excessive friction” is closer to a causal explanation. Effective corrective action changes the physical, procedural or organisational conditions that permitted recurrence.

4.3 Criticality and risk

A practical criticality assessment considers:

• personnel safety and health;
• environmental release;
• product quality and regulatory compliance;
• production loss and customer service;
• repair cost and spare-part lead time;
• availability of redundancy; and
• detectability and potential secondary damage.

Risk-based maintenance allocates analytical effort and resources according to consequence and likelihood. It should be reviewed after process changes, repeated failures, incidents or changes in operating context.

5. Overall Equipment Effectiveness

OEE combines three dimensions of equipment performance:

where:

The time boundaries and treatment of planned stops must be defined consistently. OEE is intended for a particular equipment or process and period; it should not be used as a simplistic ranking tool across incomparable operations.

Worked example: OEE

A shift has 480 minutes. Planned breaks total 30 minutes, so planned production time is 450 minutes. A machine records 60 minutes of downtime, produces 1,500 units, has an ideal cycle time of 0.24 minutes per unit, and produces 1,440 good units.

Availability:

Availability = (450 − 60) ÷ 450 = 0.8667, or 86.67%.

Performance:

Performance = (0.24 × 1,500) ÷ 390 = 0.9231, or 92.31%.

Quality:

Quality = 1,440 ÷ 1,500 = 0.96, or 96.00%.

Thus:

OEE = 0.8667 × 0.9231 × 0.96 = 0.7679, or approximately 76.79%.

The loss structure is more useful than the single percentage. Availability losses include breakdowns and setup delays; performance losses include speed reduction and minor stops; quality losses include defects and start-up rejects. OEE should lead to targeted improvement, not pressure to manipulate data.

6. Total Productive Maintenance

TPM is a company-wide approach in which operators, maintenance personnel, engineers, managers and support functions jointly improve equipment effectiveness. Its central idea is that maintenance is not solely the responsibility of a specialist department.

Common TPM principles include:

• operator involvement in cleaning, inspection, lubrication and simple adjustments;
• planned maintenance based on equipment condition and failure history;
• focused improvement of chronic losses;
• early equipment management so maintainability is designed in;
• quality maintenance that prevents equipment conditions from creating defects;
• training and development of technical competence;
• safety, health and environmental integration; and
• administrative TPM for reliable support processes.

Autonomous maintenance does not mean transferring specialist repairs to untrained operators. It means defining safe, standardised front-line activities and clear escalation rules. TPM fails when it becomes a checklist campaign without time, training, management support or measurable problem-solving.

7. Asset lifecycle management

Asset lifecycle management coordinates decisions from need identification through disposal. A useful lifecycle is:

1. Strategic need and specification: define required capacity, function, quality, safety and environmental performance.
2. Design and selection: assess reliability, maintainability, energy use, interoperability, spares, skills and total cost.
3. Acquisition and commissioning: verify specifications, installation quality, documentation and baseline condition.
4. Operation and maintenance: control work, monitor condition, manage risks and improve performance.
5. Renewal or modification: compare continued operation, refurbishment, redesign and replacement.
6. Decommissioning and disposal: manage residual value, data, hazardous materials, environmental obligations and learning capture.

The relevant financial perspective is total cost of ownership rather than purchase price alone. A simplified lifecycle-cost model is:

where C₀ is initial acquisition and installation cost, Cₜ is cost in period t, r is the discount rate per period, and the summation covers the asset’s evaluation horizon. Costs may include energy, labour, spares, planned maintenance, failure consequences, compliance, disposal and residual value. If residual value S is received at the end of period n, it is commonly subtracted in present-value terms:

The calculation assumes costs and timing can be estimated and that the discount rate is appropriate. Sensitivity analysis is essential because failure costs, energy prices and useful life are uncertain.

Make-or-buy and replacement decisions

Replacement should not be triggered solely by age. Managers should compare the expected future cost and risk of retaining an asset with the cost and performance of alternatives. Relevant factors include declining reliability, obsolete controls, unavailable parts, energy inefficiency, safety exposure, capacity requirements and technology compatibility.

A new asset with a lower purchase price may have higher energy, training or maintenance costs. Conversely, retaining an old asset may be rational if it is reliable, maintainable and well supported. Lifecycle decisions require cross-functional participation from operations, maintenance, finance, engineering, procurement, safety and sustainability personnel.

8. Digital maintenance, data and artificial intelligence

Computerised maintenance management systems (CMMS) support asset registers, work orders, preventive schedules, labour records, materials and history. Enterprise asset management systems extend this view to finance, procurement, projects and lifecycle governance.

The value of digital maintenance depends on data quality. Asset identifiers, failure codes, labour time, downtime definitions and work descriptions must be standardised. Poorly coded history can produce apparently sophisticated but unreliable analytics.

Sensors and industrial Internet of Things platforms can support condition monitoring. Machine-learning models may detect unusual patterns, estimate remaining useful life or prioritise work. However, predictions are only decision support. They require representative data, validation, cybersecurity, explainability and human review. Important risks include false negatives, false alarms, model drift, biased training data and unsafe automated recommendations.

Technology should be introduced where it improves a defined decision. A well-designed inspection route may be more effective than an expensive digital platform, particularly when failure modes are not measurable or the asset population is small.

9. Governance, safety and ethics

Maintenance management has direct ethical and regulatory implications. Organisations must provide safe systems of work, competent personnel, isolation and lockout procedures, permits, protective equipment and accurate records. Production targets must never justify bypassing protective devices or deferring legally required inspections.

Governance requires:

• clear accountability for asset risks;
• controlled maintenance procedures and change management;
• audit trails for safety-critical work;
• competent contractor selection and supervision;
• management of obsolete parts and hazardous waste;
• protection of operational data and cyber-physical systems; and
• honest reporting of downtime, defects and near misses.

Maintenance backlogs should be risk-ranked rather than hidden through altered status codes. A transparent backlog enables informed decisions about staffing, capital expenditure and operating risk.

10. Illustrative case: reliability improvement in a filling line

Background

A beverage plant operates a high-speed filling line. The line repeatedly stops because of conveyor motor failures, sensor contamination and changeover delays. The plant reports an OEE of 68%, missed delivery targets and increasing overtime.

Problem analysis

The maintenance department initially replaces motors after failure. Review of 12 months of records reveals that three motors account for most breakdown hours. However, failure codes are inconsistent, and planned maintenance includes calendar-based replacement of several sensors that rarely fail. A cross-functional team conducts a criticality review, FMEA and loss analysis.

The analysis identifies:

• motor overheating caused by dust accumulation and restricted cooling;
• loose electrical connections caused by vibration and inadequate torque control;
• sensor faults caused by contamination and poor mounting;
• long recovery time because critical motors are not held as spares; and
• changeover losses caused by inconsistent setup methods.

Approach

The team implements the following measures:

1. operators perform standard cleaning and visual checks at defined intervals;
2. technicians use thermal inspection and vibration monitoring on critical motors;
3. electrical connection checks are added to a planned task based on observed failure mechanisms;
4. a critical spare motor is stocked using a consequence and lead-time analysis;
5. failure codes are standardised in the CMMS;
6. changeover work is analysed using loss data and standard work; and
7. weekly reviews examine OEE components, repeat failures and overdue high-risk work.

The plant does not install sensors on every motor. Monitoring is limited to assets where the warning period and consequence justify the cost.

Outcome

After implementation, breakdown hours fall, recovery time decreases because of the critical spare, and minor stops decline through improved cleaning and sensor mounting. OEE rises from 68% to 79% over the evaluation period. The improvement is not attributed to one technology: it results from better failure definition, operator involvement, targeted condition monitoring, spares policy and management of the largest losses.

Learning points

• Failure history must be coded consistently before advanced analysis is attempted.
• The appropriate strategy can differ between components of the same machine.
• OEE identifies loss categories but does not itself explain root causes.
• Operator involvement is effective when activities are safe, standardised and supported.
• Maintenance improvement requires operations, engineering, stores and management alignment.

11. Implementation framework and limitations

A practical maintenance improvement programme can follow these stages:

1. establish an accurate asset register and criticality ranking;
2. define required functions and performance standards;
3. clean and classify failure and work-order data;
4. analyse the highest-consequence or highest-loss assets;
5. select corrective, preventive, predictive or RCM tasks according to evidence;
6. define measures such as availability, MTTR, schedule compliance, repeat failures and OEE;
7. pilot changes on a representative asset group;
8. review safety, cost, quality and operational effects; and
9. standardise successful practices and continue improvement.

Limitations must be recognised. Reliability estimates can be unstable with small samples. MTBF can improve because minor failures are not recorded, not because the asset improved. OEE can be distorted by inconsistent definitions. FMEA scores can create false precision. Predictive models can fail when operating conditions change. Lifecycle costs depend on uncertain forecasts. Consequently, metrics should be triangulated with engineering judgement, observations, audit findings and operational outcomes.

Chapter Summary

Maintenance is a strategic operations capability concerned with preserving asset function, safety, quality and economic value. Corrective maintenance is suitable for selected low-consequence failures; preventive maintenance uses predetermined intervals; predictive maintenance uses measured condition; and RCM selects tasks by function, failure mode and consequence.

Reliability is the probability of successful performance for a stated time and condition. MTBF, MTTR and availability are useful but depend on clear definitions and assumptions. Failure analysis methods such as FMEA, fault-tree analysis and root-cause analysis help connect observed problems to effective actions.

OEE combines availability, performance and quality and is most useful when its component losses are investigated. TPM develops shared responsibility for equipment effectiveness through operator care, planned maintenance, focused improvement, training and early equipment management.

Asset lifecycle management extends beyond maintenance to specification, acquisition, operation, renewal and disposal. Total cost of ownership, risk, sustainability, safety, data quality and cybersecurity should inform decisions. Digital tools and artificial intelligence can improve condition-based decisions, but they cannot substitute for sound failure logic, competent people and ethical governance.

Key Terms

Reliability: Probability that an asset performs its required function for a specified time under stated conditions.

Maintainability: Probability that an asset can be restored within a specified time under stated conditions.

Availability: Proportion of required time during which an asset is capable of performing its function.

Corrective maintenance: Maintenance performed to restore an asset after a fault or failure.

Preventive maintenance: Maintenance performed at predetermined intervals or criteria to reduce failure probability.

Predictive maintenance: Maintenance planned from measured asset condition or performance trends.

Reliability-centred maintenance: A function- and consequence-based method for selecting maintenance policies.

Failure mode: The specific manner in which an asset fails to perform a function.

Failure effect: The consequence of a failure mode at a local, system or operational level.

FMEA: A structured analysis of failure modes, causes, effects and controls.

Root-cause analysis: Investigation of underlying physical, human or organisational causes of failure.

MTBF: Mean operating time between failures of a repairable asset.

MTTR: Mean time required to restore an asset after corrective failure.

OEE: Product of availability, performance and quality.

TPM: Company-wide approach involving all functions in improving equipment effectiveness.

Asset criticality: Relative importance of an asset based on consequences and risk.

Lifecycle cost: Total discounted cost of acquiring, operating, maintaining and disposing of an asset.

CMMS: Computerised system for managing assets, work orders, schedules, labour and maintenance history.

Condition monitoring: Measurement and interpretation of indicators of asset health.

Run-to-failure: Deliberate policy of allowing an asset to operate until functional failure, when consequences are acceptable.

Review Questions

Short Answer Questions

1. Distinguish corrective, preventive and predictive maintenance.
2. Define reliability and explain why it is a probability rather than a guarantee.
3. What are the three components of OEE?
4. State two purposes of FMEA.
5. Explain the difference between MTBF and MTTR.

Long Answer Questions

1. Evaluate the advantages and limitations of preventive and predictive maintenance for a critical production asset.
2. Explain the RCM decision process from required function to selected maintenance task.
3. Discuss how TPM changes the roles of operators, maintenance personnel and managers.
4. Explain why lifecycle cost is more useful than purchase price in asset decisions.
5. Discuss the governance, safety and data-quality requirements of a digital maintenance programme.

Analytical / Application Questions

1. A machine operates for 2,400 hours and has eight failures requiring 48 hours of corrective repair. Calculate MTBF, MTTR and the MTBF-based availability approximation. State one limitation of your interpretation.
2. A process has planned production time of 420 minutes, 50 minutes of downtime, an ideal cycle time of 0.2 minutes, total output of 1,700 units and 1,615 good units. Calculate availability, performance, quality and OEE.
3. An inspection system detects occasional overheating in a safety-related pump, but the alarm threshold produces many false positives. Recommend a structured approach using criticality analysis, condition-monitoring validation and RCM principles.

Multiple Choice Questions with Answers

Question 1

Which maintenance strategy uses measured equipment condition to determine when intervention is required?

A. Corrective maintenance

B. Predictive maintenance

C. Run-to-failure maintenance

D. Emergency procurement

Correct Answer: B. Predictive maintenance

Question 2

An asset operates for 1,000 hours and fails five times. What is its MTBF?

A. 50 hours

B. 100 hours

C. 200 hours

D. 5,000 hours

Correct Answer: C. 200 hours

Question 3

Which assumption is required for the simple formula A = MTBF ÷ (MTBF + MTTR)?

A. The asset is non-repairable

B. Failure and repair averages are used under broadly stable conditions

C. All failures occur at exactly equal intervals

D. Planned downtime must always be included

Correct Answer: B. Failure and repair averages are used under broadly stable conditions

Question 4

Which OEE component measures the proportion of produced units that meet quality requirements?

A. Availability

B. Performance

C. Quality

D. Maintainability

Correct Answer: C. Quality

Question 5

What is the primary focus of reliability-centred maintenance?

A. Replacing every component at the same calendar interval

B. Maximising maintenance department activity

C. Linking maintenance tasks to functions, failure modes and consequences

D. Eliminating all corrective maintenance

Correct Answer: C. Linking maintenance tasks to functions, failure modes and consequences

Question 6

A machine has availability of 0.90, performance of 0.80 and quality of 0.95. What is its OEE?

A. 0.684

B. 0.855

C. 0.950

D. 2.650

Correct Answer: A. 0.684

Question 7

Which action is most characteristic of autonomous maintenance in TPM?

A. Operators perform defined cleaning, inspection and lubrication tasks

B. Operators bypass safety guards to reduce stoppages

C. All maintenance work is outsourced

D. Managers ignore minor abnormalities

Correct Answer: A. Operators perform defined cleaning, inspection and lubrication tasks

Question 8

Why should an RPN score in FMEA not be treated as a complete risk measure?

A. It cannot include any numerical values

B. It is based on ordinal ratings and may obscure catastrophic consequences

C. It applies only to financial risks

D. It always produces a probability of failure

Correct Answer: B. It is based on ordinal ratings and may obscure catastrophic consequences

Question 9

Which cost is normally included in lifecycle cost analysis?

A. Only the purchase price

B. Only annual maintenance labour

C. Acquisition, operation, maintenance and disposal-related costs

D. Only the asset’s resale price

Correct Answer: C. Acquisition, operation, maintenance and disposal-related costs

Question 10

A planned production period is 400 minutes, downtime is 40 minutes, total output is 1,800 units, ideal cycle time is 0.2 minutes, and good output is 1,710 units. What is the performance rate?

A. 0.80

B. 0.90

C. 1.00

D. 1.11

Correct Answer: C. 1.00