1. Foundations of Maintenance and Reliability Management

Introduction and context

Production systems convert materials, energy, information, and human effort into products or services. Their performance depends not only on production methods and workforce capability but also on the condition of the physical assets that support production. Machines, utilities, buildings, control systems, transportation equipment, and information systems must perform their intended functions when required. Maintenance and reliability management provides the managerial and technical discipline for achieving this condition.

Maintenance is often treated narrowly as repair work. Such a view is inadequate for postgraduate study and for contemporary industrial practice. Maintenance includes the decisions, engineering activities, organisational arrangements, inspections, interventions, and improvement actions used to preserve or restore asset function. Reliability management extends this concern by asking whether an asset, system, or process can perform its required function consistently over a specified period and under specified conditions.

The subject therefore links engineering with operations management, finance, quality, safety, risk, supply chain management, and human decision-making. Poor maintenance can cause production interruptions, defective products, energy waste, environmental releases, injuries, and loss of reputation. Conversely, excessive or poorly designed maintenance can create unnecessary cost, introduce errors, and reduce equipment availability. The central managerial challenge is to obtain the required performance at an economically and socially acceptable level of risk.

Learning objectives

After studying this chapter, the student should be able to:

1. Define maintenance, reliability, maintainability, availability, failure, and related terminology.

2. Explain the historical development of maintenance and reliability practices.

3. Distinguish corrective, preventive, predictive, proactive, and opportunistic maintenance.

4. Explain the relationships among reliability, maintainability, availability, quality, and production performance.

5. Assess the consequences of equipment failure and apply the idea of asset criticality.

6. Interpret basic reliability and availability measures.

7. Explain managerial, ethical, technological, and organisational issues in maintenance decisions.

8. Apply foundational concepts to an industrial case involving asset performance and risk.

The changing meaning of maintenance

The word maintenance derives from the idea of preserving a desired condition. In an industrial context, an asset is maintained when it is kept capable of delivering its required function. This definition is more useful than defining maintenance as the repair of a broken machine because many maintenance tasks occur before failure.

Maintenance is the combination of all technical, administrative, and managerial actions during the life cycle of an item intended to retain it in, or restore it to, a state in which it can perform a required function. The phrase “required function” is important. An asset may still operate but fail to meet its required production rate, accuracy, safety standard, energy efficiency, or product specification.

A failure is the termination of an item’s ability to perform a required function. Failure need not mean total physical destruction. A pump that continues to run but cannot deliver the required flow, or a sensor that produces inaccurate readings, has functionally failed. A fault is an abnormal condition that may cause failure or reduce performance. A fault can sometimes be detected and corrected before functional failure occurs.

Maintenance management concerns planning, organising, directing, controlling, and improving maintenance resources. These resources include people, tools, spare parts, information, contractors, budgets, facilities, and time. Reliability management adds a broader focus on failure prevention, design decisions, operating conditions, data analysis, and risk reduction across the asset life cycle.

Core terminology

An asset is an item, system, or resource that has potential or actual value to an organisation. A physical asset may be a machine, vehicle, production line, building, pipeline, or utility system. An item is any part, component, device, subsystem, or system that can be individually considered.

Function describes what an asset is required to do, including performance standards and operating context. Functional failure occurs when the asset cannot perform a specified function to the required standard. Hidden failure is a failure that does not become evident during normal operation, often because the protective device is normally inactive. Examples include a standby generator that cannot start when needed or a safety relief device that is stuck.

Repairable items can be restored after failure, whereas non-repairable items are replaced when they fail. A degraded state is a condition in which an asset still performs but at a reduced level. Defect elimination is the removal of an underlying condition that produces repeated faults, such as poor alignment, contamination, inadequate lubrication, or an unsuitable design.

Mean time to failure (MTTF) is the average operating time to failure for non-repairable items. Mean time between failures (MTBF) is commonly used for repairable equipment to describe the average elapsed operating time between inherent failures, although organisations must define the term carefully because some reports use it to mean total calendar time between failures. Mean time to repair (MTTR) is the average active time required to restore a failed item, but practical studies should distinguish hands-on repair time from waiting, diagnosis, access, and testing time.

Historical development: from repair to asset strategy

Industrial maintenance developed alongside mechanisation and mass production. In the early industrial period, equipment was often operated until it failed. Maintenance was predominantly reactive: a breakdown created an immediate repair requirement. This approach could be acceptable where equipment was simple, spare capacity was available, and failure consequences were modest. It became increasingly unsuitable as production systems grew more complex and interdependent.

During the era of mass production, equipment failure could stop an entire line. Organisations began to schedule inspections, lubrication, overhaul, and component replacement at predetermined intervals. This was the foundation of preventive maintenance. The underlying assumption was that age or operating time provided a useful basis for predicting failure. That assumption is valid for some wear-out mechanisms but not for all failure modes.

The expansion of aviation, defence, nuclear power, chemical processing, and other high-risk industries encouraged systematic reliability engineering. Engineers began to study failure modes, failure rates, redundancy, human factors, and the consequences of failure. Reliability became a design and system-performance concern rather than merely a maintenance department concern.

From the 1970s onward, condition monitoring, total productive maintenance, reliability-centred maintenance, quality management, and life-cycle costing broadened the field. Operators became more involved in routine care; maintenance moved closer to production; and organisations began to distinguish between technically possible maintenance and economically justified maintenance.

The contemporary stage is characterised by asset management, digital condition monitoring, mobile work management, remote operations, and data analytics. These developments do not eliminate the need for engineering judgement. Data may be incomplete, sensors may be poorly placed, algorithms may produce false alarms, and a statistically attractive intervention may be operationally impractical. Technology strengthens maintenance decisions only when it is connected to clear functions, failure mechanisms, competent people, and disciplined processes.

Maintenance approaches and their logic

Corrective maintenance is performed after a fault has been detected. It may be planned or unplanned. Planned corrective maintenance occurs when a known defect is deliberately allowed to remain in service until a suitable intervention opportunity. Unplanned corrective maintenance follows an unexpected breakdown. The distinction matters because planned corrective work can be controlled, whereas emergency breakdown work often disrupts production and increases risk.

Preventive maintenance is performed at predetermined intervals or according to prescribed criteria to reduce the probability of failure or degradation. Time-based replacement, scheduled lubrication, inspection, and calibration are examples. Preventive maintenance is useful when a clear relationship exists between age, usage, and failure probability. It can be wasteful when replacement intervals are chosen arbitrarily or when intrusive maintenance creates new defects.

Predictive maintenance uses measurement and analysis of equipment condition to determine when intervention is needed. Vibration analysis, thermography, oil analysis, motor-current analysis, ultrasonic inspection, and performance trending are common techniques. Predictive maintenance does not mean that failure is predicted with certainty. Rather, it identifies evidence of degradation early enough to support an informed decision.

Condition-based maintenance initiates action when a measurable condition crosses a defined limit or shows a meaningful deterioration trend. It requires an understanding of what is measured, how measurement relates to failure, and what response is appropriate. A warning limit should not be confused with a failure limit; the first should allow time for planning and risk control.

Proactive maintenance seeks to eliminate or control the causes of failure. Alignment improvement, design modification, contamination control, precision installation, root-cause analysis, and operator training are proactive actions. Their value is often greater than repeatedly repairing the same symptom.

Reliability-centred maintenance (RCM) begins with functions and performance standards, identifies functional failures and failure modes, evaluates their consequences, and selects technically feasible and economically justified tasks. It does not assume that every asset requires preventive maintenance. If no scheduled task is effective, the appropriate choice may be redesign, failure finding, run-to-failure, redundancy, or a change in operating practice.

Total productive maintenance (TPM) emphasises the participation of production operators, maintenance personnel, engineers, and managers in improving equipment effectiveness. Autonomous maintenance activities may include cleaning, inspection, lubrication, tightening, and early abnormality detection. TPM is not simply a housekeeping programme; its effectiveness depends on standards, training, measurement, and management support.

Reliability, maintainability, and availability

Reliability is the probability that an item will perform its required function without failure for a specified time under stated conditions. The conditions may include load, environment, duty cycle, speed, temperature, maintenance policy, and operator practice. Reliability is therefore not an intrinsic number independent of context.

For a constant failure rate, the reliability of a non-repairable item can be represented by:

R(t) = e^(-λt)

where λ is the failure rate and t is operating time. The constant-rate assumption is a simplification. Many assets display changing failure behaviour, including early-life defects, relatively stable random failures, and wear-out failures. A bathtub curve is often used to illustrate these phases, but it should not be treated as a universal law or as proof that age-based replacement is always appropriate.

Maintainability is the probability that a failed item can be restored to a specified operating condition within a stated time, using prescribed procedures and resources. Maintainability is influenced by accessibility, modular design, diagnostic information, tools, spare parts, standardisation, skills, safety isolation, and work planning. A highly reliable machine may still be operationally unsatisfactory if it is extremely difficult to repair.

Availability is the ability of an item to be in a state to perform a required function at a given point in time or over a stated period, assuming external resources are available. For a simple steady-state repairable system:

Inherent availability = MTBF / (MTBF + MTTR)

This expression excludes logistics and administrative delays. Operational availability includes factors such as waiting for spares, permits, technicians, transport, and production access. Consequently, reducing waiting time may improve operational availability even when the physical repair itself remains unchanged.

Reliability and availability must not be confused. A machine may have high reliability but low availability if repairs are lengthy. Another machine may fail frequently but recover rapidly, producing acceptable availability in a low-consequence application. The appropriate performance target depends on the required function, risk, production economics, and customer expectations.

A system perspective

For components arranged in series, system reliability is generally limited because failure of any essential component can stop the system. If component failures are independent, series reliability may be expressed as:

R system = R1 × R2 × ... × Rn

Parallel or redundant arrangements can improve reliability or availability, but redundancy introduces its own issues: common-cause failures, maintenance of standby units, switching failures, testing requirements, and additional capital cost. A redundant pump system, for example, may fail if both pumps share a blocked suction line or a common electrical supply.

From equipment performance to production performance

Maintenance outcomes matter because equipment is part of a production system. A common measure of production effectiveness is overall equipment effectiveness (OEE), often expressed as:

OEE = Availability × Performance × Quality

Availability in this context usually reflects planned production time lost to breakdowns, setup, and adjustments. Performance reflects speed losses and minor stops, while quality reflects defective output and startup losses. OEE is useful as a diagnostic framework, but it should not be used as a simplistic ranking tool. Improving the numerator without understanding demand, product mix, constraints, or measurement rules can lead to misleading conclusions.

Reliability supports production by reducing unplanned stoppages and unstable operating conditions. Maintainability reduces restoration time. Availability captures whether capacity is ready when required. Quality is linked because equipment deterioration can cause variation, contamination, dimensional error, incorrect dosage, or loss of process control. Production performance is therefore the combined result of asset capability, operating method, material supply, labour, scheduling, quality control, and maintenance.

Maintenance decisions also affect cost. Direct costs include labour, materials, contractors, tools, and spare parts. Indirect costs may include lost production, overtime, customer penalties, energy consumption, environmental remediation, and reputational damage. A low-cost maintenance task is not necessarily economically attractive if it creates frequent interruptions; a costly inspection may be justified if it prevents a catastrophic event.

Failure consequences and asset criticality

Failure consequence is the effect produced when an asset loses a required function. Consequences may be safety-related, environmental, operational, quality-related, financial, legal, or reputational. The same failure can have different consequences in different operating contexts. Failure of a workshop fan is not equivalent to failure of a ventilation system in a hazardous process area.

Asset criticality is a structured assessment of the importance of an asset relative to defined objectives and risks. Criticality is not simply the purchase price of an asset. A low-cost valve can be more critical than an expensive machine if its failure can cause a major release or shut down an entire facility.

A practical criticality assessment commonly considers:

• consequence severity, including safety, environmental, quality, delivery, and financial effects;
• likelihood or frequency of failure;
• detectability and warning time;
• availability of standby capacity;
• restoration time and spare-part lead time;
• legal, regulatory, and customer requirements;
• interdependence with other assets.

Organisations may classify assets as high, medium, or low criticality, or use a numerical risk matrix. Such tools support consistency but do not replace engineering judgement. Risk matrices can hide uncertainty, encourage artificial precision, and produce different results when severity categories are poorly defined. Criticality should be reviewed when process conditions, production plans, regulations, technology, or failure history changes.

Applied case: reliability improvement in a food-processing line

Background

This illustrative case concerns a high-speed packaging line in a food-processing plant. The line contains a conveyor system, filling unit, sealing machine, vision inspection system, and case packer. The line operates two shifts per day and supplies products to several retail customers.

Situation and problem

The sealing machine experienced repeated stoppages. Maintenance records showed frequent replacement of heater cartridges and sporadic failures of temperature sensors. The maintenance department responded with emergency repairs and increased the frequency of cartridge replacement. Despite the additional work, monthly output remained below plan and customer complaints about seal integrity increased.

The initial temptation was to classify the problem as poor component reliability. However, the production team reported that operators sometimes cleaned the sealing jaws with compressed air while the machine was hot, and that product dust accumulated around the sensor mounting. The machine also experienced short voltage disturbances that were not recorded in the existing maintenance system.

Analysis

The team first defined the required functions: maintain sealing temperature within a specified range, create a hermetic seal at the required line speed, and provide accurate temperature feedback to the control system. Functional failures included inability to reach temperature, unstable temperature control, and defective seals despite apparent normal operation.

A review of failure modes identified several contributors: heater burnout, sensor drift, loose connections, contamination, incorrect cleaning practice, and voltage disturbance. The team separated failure frequency from failure consequence. A failed heater caused a stoppage, but sensor drift could allow production to continue while generating defective product, making its quality consequence significant.

The team then compared maintenance data with quality records. The apparent improvement from more frequent heater replacement was limited because the replacement task did not address contamination, electrical disturbance, or installation quality. Moreover, repeated hot cleaning created thermal shock and increased the risk of damage.

Approach and solution

The plant adopted a combined strategy. Sensors were inspected through a condition-based calibration programme; mounting points were redesigned to reduce contamination; electrical protection and disturbance logging were installed; cleaning procedures were revised; and technicians received training in connection torque and precision installation. The line’s criticality assessment was updated because defective seals created both customer and food-safety risks.

A planned corrective window was established for trend-based intervention rather than waiting for a breakdown. Operators were trained to identify abnormal temperature trends and seal appearance. The work-order system was modified to record failure mode, operating condition, and consequence rather than merely recording “heater failed.”

Outcome

The illustrative plant experienced fewer emergency stoppages, more reliable temperature control, and a reduction in seal-related defects. The important improvement did not arise from one additional preventive task. It resulted from linking function definition, failure-mode analysis, operator practice, condition monitoring, electrical reliability, quality data, and work-management discipline.

Learning points

The case demonstrates that:

1. A component replacement programme can treat symptoms without controlling causes.

2. Reliability analysis must include failures that produce defective output without stopping production.

3. Maintenance and quality data should be interpreted together.

4. Human practices and operating conditions are part of the reliability system.

5. Criticality should reflect consequence, not merely asset cost or failure frequency.

Managerial, ethical, and governance considerations

Maintenance managers allocate scarce resources under uncertainty. They must decide which assets receive engineering attention, which tasks are deferred, and which risks are accepted. These decisions should be transparent and traceable. Deferring maintenance may be reasonable when risk is assessed and controlled; concealing a known defect or manipulating availability figures is professionally unacceptable.

Safety-critical and environmentally significant assets require compliance with applicable laws, standards, permits, inspection regimes, and management-of-change procedures. Maintenance personnel must be protected through isolation, permit-to-work systems, lockout/tagout, confined-space controls, appropriate personal protective equipment, and competent supervision. Production pressure does not justify bypassing a safety interlock without formal risk assessment and authorisation.

Data governance is increasingly important. Sensor data, work histories, operator reports, and failure classifications influence investment and risk decisions. Incorrect timestamps, inconsistent coding, missing failure modes, and deliberate closure of work orders without effective restoration can make analytics misleading. Access controls and cybersecurity are also relevant where connected assets can be remotely monitored or controlled.

Digital maintenance and emerging directions

Industrial Internet of Things devices, wireless sensors, cloud platforms, digital work orders, and machine-learning models are changing how condition information is collected and used. Analytics can identify abnormal vibration, energy consumption, temperature, or process behaviour. Digital twins may combine engineering models with operating data to support diagnosis and scenario analysis.

However, digitalisation creates several limitations. A sensor may measure a symptom rather than the failure mechanism. Algorithms trained on normal operating data may perform poorly after a product change or process modification. False alarms can cause alert fatigue, while missed alarms can produce misplaced confidence. Cybersecurity vulnerabilities may affect both production continuity and safety.

The emerging direction is not the replacement of maintenance professionals by artificial intelligence. It is the integration of human expertise, reliable data, engineering models, and decision-support tools. Explainable analytics, edge computing, remote assistance, augmented-reality work instructions, and risk-based inspection are likely to expand, but adoption should be governed by business value, competence, data quality, and operational risk.

Common mistakes and good practice

Common mistakes include using MTBF as a target without understanding failure definitions; assuming that all failures are age-related; measuring maintenance activity rather than restored function; treating preventive maintenance as automatically beneficial; ignoring hidden failures; optimising departmental cost while increasing production loss; and collecting data without a decision process.

Good practice begins with clear asset functions and performance standards. Organisations should maintain accurate equipment hierarchies, standardise failure and work-order codes, prioritise assets by consequence and risk, and select tasks according to credible failure mechanisms. Maintenance plans should be reviewed using evidence from condition monitoring, quality records, incidents, near misses, and production performance. Operators and maintainers should jointly own basic equipment care, while engineering should address design weaknesses and recurring causes.

At the organisational level, maintenance performance should be evaluated through a balanced set of measures: unplanned downtime, schedule compliance, planned-to-unplanned work, repeat failures, maintenance cost, backlog quality, safety performance, quality losses, energy use, and availability of critical assets. No single indicator represents maintenance effectiveness.

Chapter Summary

Maintenance is the life-cycle discipline of preserving or restoring an asset’s ability to perform its required function. Reliability concerns the probability of failure-free performance under stated conditions; maintainability concerns the speed and ease of restoration; and availability combines operating readiness with restoration capability. Corrective, preventive, predictive, condition-based, proactive, RCM, and TPM approaches are selected according to failure behaviour, consequence, risk, and economics.

Asset failure can affect safety, environment, quality, delivery, cost, and reputation. Criticality assessment helps organisations focus resources on assets whose failure matters most. Reliability and maintenance influence production performance through availability, speed stability, product quality, and process control. Effective practice requires sound definitions, credible data, cross-functional cooperation, ethical governance, competent people, and thoughtful use of digital technologies. The foundational principle is that maintenance should be designed around required function and risk, not around repair activity alone.

Key Terms

Asset: An item or resource that has actual or potential value to an organisation.

Availability: The ability of an item to perform a required function when required.

Condition-based maintenance: Maintenance initiated by evidence of an asset’s measured condition.

Corrective maintenance: Maintenance performed to restore an item after a fault or failure has been identified.

Failure: Termination of an item’s ability to perform a required function.

Fault: An abnormal condition that may cause failure or reduce performance.

Functional failure: Inability to perform a specified function to the required standard.

Maintainability: The probability that an item can be restored within a specified time using stated resources.

Maintenance: Technical, administrative, and managerial actions intended to retain or restore required function.

Mean time between failures: Average operating time between failures of a repairable item, subject to a defined measurement convention.

Mean time to failure: Average operating time until failure for a non-repairable item.

Mean time to repair: Average active time required to restore a failed item.

Predictive maintenance: Maintenance supported by measurement and analysis of equipment condition.

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

Proactive maintenance: Action directed at eliminating or controlling the causes of failure.

Reliability: Probability of performing a required function without failure for a stated time and condition.

Reliability-centred maintenance: A structured method for selecting maintenance policies from functions, failure modes, consequences, and task effectiveness.

Asset criticality: Relative importance of an asset based on failure consequences, likelihood, and related risk.

Total productive maintenance: A company-wide approach involving operators and maintainers in equipment effectiveness improvement.

Overall equipment effectiveness: A composite measure commonly expressed as availability multiplied by performance and quality.

Review Questions

Short Answer Questions

1. Define maintenance in terms of required function rather than repair activity.

2. Distinguish reliability from availability.

3. What is meant by a hidden failure?

4. State two factors that influence maintainability.

5. Why is asset purchase price insufficient for determining criticality?

Long Answer Questions

1. Trace the historical development of maintenance from breakdown repair to contemporary asset management.

2. Compare corrective, preventive, predictive, condition-based, and proactive maintenance approaches.

3. Explain the relationship among reliability, maintainability, and availability, including the role of MTBF and MTTR.

4. Discuss how equipment condition can influence product quality and production performance.

5. Explain the principles of reliability-centred maintenance and the circumstances in which run-to-failure may be appropriate.

Analytical / Application Questions

1. A repairable machine has an MTBF of 900 hours and an MTTR of 9 hours. Calculate its inherent availability and explain why its operational availability might be lower.

2. A standby safety pump is rarely used and has no recent test record. Analyse why it may be a critical asset despite its low operating hours, and propose an appropriate maintenance strategy.

3. A plant reports that preventive-maintenance hours increased by 30 percent while unplanned downtime and product defects also increased. Diagnose possible reasons and recommend a structured investigation.

22. Multiple Choice Questions with Answers

Question 1

Which statement best distinguishes reliability from availability?

A. Reliability concerns repair cost, while availability concerns purchase cost.

B. Reliability concerns failure-free operation for a period, while availability concerns readiness to perform when required.

C. Reliability applies only to non-repairable items, while availability applies only to repairable items.

D. Reliability measures product quality, while availability measures worker productivity.

Correct Answer: B. Reliability concerns failure-free operation for a period, while availability concerns readiness to perform when required.

Question 2

A machine continues running but produces seals outside the required specification. What has most appropriately occurred?

A. No failure, because the machine is still rotating

B. A hidden failure only

C. A functional failure

D. A successful preventive-maintenance event

Correct Answer: C. A functional failure

Question 3

Which maintenance approach is most directly intended to eliminate recurring causes such as misalignment and contamination?

A. Reactive corrective maintenance

B. Proactive maintenance

C. Run-to-failure maintenance

D. Unplanned replacement

Correct Answer: B. Proactive maintenance

Question 4

For a repairable asset, which change most directly improves inherent availability, assuming MTBF remains constant?

A. Increasing MTTR

B. Reducing MTTR

C. Increasing product variety

D. Reducing inspection accuracy

Correct Answer: B. Reducing MTTR

Question 5

Why might a low-cost valve receive a higher criticality rating than an expensive workshop machine?

A. Criticality is based only on purchase price

B. The valve may have more attractive resale value

C. Its failure may create severe safety, environmental, or production consequences

D. Low-cost assets always fail more frequently

Correct Answer: C. Its failure may create severe safety, environmental, or production consequences

Question 6

A plant replaces bearings every six months although data show no relationship between age and bearing failure. What is the principal concern?

A. Preventive maintenance may be unnecessary and may introduce defects

B. Reliability cannot be measured for bearings

C. Condition monitoring always increases failure frequency

D. Corrective maintenance is prohibited in industry

Correct Answer: A. Preventive maintenance may be unnecessary and may introduce defects

Question 7

In the packaging-line case, why was replacing heater cartridges alone inadequate?

A. Heater cartridges are never replaceable

B. The underlying contributors included contamination, electrical disturbance, installation, and operating practices

C. Quality data cannot be used in maintenance decisions

D. The sealing machine had no required function

Correct Answer: B. The underlying contributors included contamination, electrical disturbance, installation, and operating practices

Question 8

Which situation is most characteristic of a hidden failure?

A. A conveyor visibly stops during production

B. A standby generator fails to start when demanded

C. An operator reports excessive vibration

D. A pump produces reduced flow during normal operation

Correct Answer: B. A standby generator fails to start when demanded

Question 9

If a system has essential components in series, what generally happens to system reliability as more independent components are added?

A. It necessarily increases

B. It remains exactly constant

C. It generally decreases because each component adds a possible failure point

D. It becomes unrelated to component reliability

Correct Answer: C. It generally decreases because each component adds a possible failure point

Question 10

Which is the strongest reason to avoid using a single OEE value as the sole maintenance-performance measure?

A. OEE has no relationship to production

B. A single composite value can conceal the causes of availability, performance, and quality losses

C. OEE measures only spare-parts cost

D. OEE can be calculated only for non-repairable assets

Correct Answer: B. A single composite value can conceal the causes of availability, performance, and quality losses