Maintenance, Reliability and Asset Management

Introduction and Context

Maintenance is the coordinated combination of technical, administrative and managerial actions intended to keep an asset in, or restore it to, a condition in which it can perform its required function. In supply-chain operations, maintenance is not merely a workshop activity. Equipment failure can interrupt production, delay customer orders, increase inventory requirements, create safety risks and reduce the reliability of the entire supply network.

Modern operations therefore manage maintenance as part of an integrated system involving production, engineering, procurement, finance, quality, health and safety, information technology and suppliers. The central managerial question is not simply “How can equipment be repaired?” but “What level of asset performance is required, at what life-cycle cost and with what risk?”

This chapter examines maintenance strategies, reliability and availability measures, Total Productive Maintenance (TPM), Overall Equipment Effectiveness (OEE), and asset life-cycle management.

Learning Objectives

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

1. Compare reactive, preventive, predictive and condition-based maintenance.
2. Explain the relationships among failure, reliability, maintainability and availability.
3. Calculate basic reliability, mean time between failures and availability measures.
4. Apply TPM principles and calculate and interpret OEE.
5. Evaluate asset decisions across the life cycle, including acquisition, operation, renewal and disposal.
6. Recommend maintenance improvements while considering cost, safety, quality, sustainability and governance.

1. Maintenance as an Operations Capability

An asset may be a machine, vehicle, warehouse conveyor, refrigeration unit, information system, building service or production line. Its required function includes more than physical operation. It may need to produce a specified quantity at a specified quality, speed, accuracy, safety level and environmental condition.

Maintenance management seeks a balanced outcome:

• high availability when the asset is needed;
• dependable performance and product quality;
• controlled maintenance and energy costs;
• safe working conditions;
• compliance with legal and environmental requirements; and
• an economically justified asset life.

A maintenance system normally contains the following activities:

1. Identification: defining assets, locations, functions and criticality.
2. Planning: determining tasks, skills, tools, spare parts and permits.
3. Scheduling: assigning work at suitable times with available resources.
4. Execution: performing inspections, servicing, repair or replacement.
5. Verification: testing that the asset has returned to the required condition.
6. Learning: analysing failure data and improving equipment or procedures.

A useful distinction is between a failure and a fault. A fault is a condition that may reduce performance or indicate abnormality; failure occurs when the asset can no longer perform its required function. A bearing vibration increase may be a fault, while seizure of the bearing may be a failure.

2. Maintenance Strategies

2.1 Reactive or Corrective Maintenance

Reactive maintenance is performed after a failure has occurred. “Run to failure” is a deliberate form of reactive maintenance in which an item is allowed to operate until failure because the consequences are acceptable and restoration is economical.

Advantages include low routine intervention, simple planning for non-critical items and full use of component life. Disadvantages include unplanned downtime, secondary damage, urgent purchasing, safety exposure and production disruption. Reactive maintenance is inappropriate where failure threatens life, regulatory compliance, product integrity or major financial loss.

Reactive maintenance can be rational for inexpensive, non-critical components that are quick to replace and whose failure has little system impact. It should be a controlled decision, not the default result of poor planning.

2.2 Preventive Maintenance

Preventive maintenance is performed at predetermined intervals or according to prescribed criteria, regardless of the immediate condition of the asset. Examples include changing lubricant every 2,000 operating hours, inspecting a safety valve monthly or replacing a belt annually.

Time-based preventive maintenance is suitable when deterioration is age-related and the relationship between age and failure is reasonably understood. However, replacing components too early can waste useful life and introduce installation errors. Preventive tasks should therefore be justified by failure behaviour, safety requirements, manufacturer guidance, operating experience or regulatory rules.

2.3 Predictive Maintenance

Predictive maintenance uses measurements, historical data and analytical techniques to estimate when deterioration is likely to reach an unacceptable level. Typical technologies include vibration analysis, infrared thermography, oil analysis, ultrasonic inspection, motor-current analysis and performance trending.

A predictive programme requires competent interpretation. Collecting data without a decision rule is not predictive maintenance. The organisation must define the parameter, measurement frequency, alarm levels, response time and required intervention.

2.4 Condition-Based Maintenance

Condition-based maintenance initiates maintenance when evidence shows that an asset’s condition has crossed a defined threshold or is moving toward functional failure. It may use continuous sensors, periodic inspection or operator observation. Predictive maintenance is often considered a data-intensive form of condition-based maintenance, but the terms are not always used identically.

For example, a pump may be inspected weekly. If vibration exceeds a specified alarm level, a work order is created. If analysis predicts that the vibration will reach a critical level within two weeks, the repair is scheduled before failure. The first decision is condition-based; the forecast is predictive.

2.5 Comparison of Strategies

A mature organisation uses a maintenance mix. The choice should reflect failure consequences, detectability, failure pattern, required availability, safety and total cost—not preference for a particular technology.

3. Reliability, Maintainability and Availability

3.1 Reliability

Reliability is the probability that an item performs its required function under stated conditions for a stated time interval. It is not the same as quality. Quality commonly concerns conformance at manufacture or delivery; reliability concerns continued performance over time.

If failures follow a constant failure rate λ and are statistically independent, the reliability of a non-repairable item over time t is:

Here, R(t) is reliability, λ is the constant failure rate per unit of time, t is operating time, and e is the base of natural logarithms. The formula assumes a constant hazard rate and is most appropriate for the useful-life period of an item whose failures are approximately random.

The corresponding probability of failure by time t is:

For a constant failure rate, the mean time to failure is:

MTTF applies to non-repairable items. For repairable assets, the related measure is usually Mean Time Between Failures (MTBF), provided the definition of “between failures” is used consistently.

For a series system, all components must operate for the system to operate. If component reliabilities are independent:

For a parallel redundant system, the system works if at least one component works. For two independent components:

These simple system formulas assume independence. Common-cause failures, shared utilities, human error and dependency can make them misleading.

Worked Example 1: Reliability

A component has a constant failure rate of 0.0002 failures per operating hour. Estimate its reliability over 1,000 hours.

The estimated reliability is approximately 81.87%. The probability of failure during the interval is:

Thus, the probability of failure is approximately 18.13%, under the constant-rate assumption.

3.2 Failure Rate and the Bathtub Curve

Many assets show three broad phases of failure behaviour:

1. Early-life failures: high initial failure rate caused by design defects, manufacturing problems, installation errors or inadequate commissioning.
2. Useful life: approximately stable random failure rate.
3. Wear-out: increasing failure rate caused by fatigue, corrosion, erosion, ageing or degradation.

The bathtub curve is a useful conceptual model, not a universal law. Some electronic components do not show pronounced wear-out, while some mechanical components have strongly age-related deterioration. Maintenance policy should be based on evidence for the specific asset and operating context.

3.3 Maintainability

Maintainability is the probability that a failed item can be restored to a specified operating condition within a stated time, under stated procedures and resources. Design features affecting maintainability include accessibility, modular replacement, diagnostic capability, standard fasteners, documentation and availability of spares.

Mean Time To Repair (MTTR) is commonly calculated as:

The measure must state what downtime includes. Some organisations include diagnosis, waiting for parts and testing; others include only hands-on repair time. Comparisons are valid only when definitions are consistent.

3.4 Availability

Availability is the probability that an asset is operational and able to perform its required function when required. In a simple steady-state model for a repairable asset:

MTBF and MTTR must use the same time unit. This formula assumes repeated operating and repair cycles, stable average values and that downtime is represented by the repair interval. It is often called inherent or basic operational availability in simplified applications, but organisations should state their precise definition.

For a period-based measure:

Planned shutdowns, waiting time, logistics delays and administrative delays may or may not be included, depending on the chosen availability definition.

Worked Example 2: Availability

A conveyor operates for 1,920 hours during a year and experiences 80 hours of downtime. Its period availability is:

Availability is therefore 96.0%.

If the same conveyor has MTBF of 240 hours and MTTR of 10 hours, the simplified steady-state estimate is:

The agreement between the two calculations is coincidental in this constructed example; in practice, data definitions and operating conditions must be examined carefully.

4. Total Productive Maintenance

Total Productive Maintenance is an organisation-wide approach that seeks maximum equipment effectiveness through the participation of operators, maintenance personnel, engineers, managers and suppliers. TPM extends responsibility beyond the maintenance department. Operators perform appropriate routine activities such as cleaning, inspection, lubrication and identification of abnormalities, while specialists handle complex technical work.

Commonly described TPM pillars include:

• Autonomous maintenance: operators maintain basic conditions and detect abnormalities.
• Planned maintenance: preventive, predictive and corrective work is planned systematically.
• Focused improvement: cross-functional teams remove chronic losses and recurring causes.
• Quality maintenance: equipment conditions are controlled to prevent defects.
• Early equipment management: maintainability and reliability are designed into new assets.
• Training and education: people develop technical, analytical and problem-solving competence.
• Safety, health and environment: maintenance improves protection and compliance.
• Office TPM: administrative processes such as purchasing, planning and data management are improved.

TPM is not equivalent to asking operators to perform all maintenance. It requires clear boundaries, training, safe procedures, escalation rules and respect for competence. Poorly designed autonomous maintenance can transfer risk to operators without giving them adequate authority or support.

4.1 Overall Equipment Effectiveness

OEE measures how effectively a production asset uses its scheduled production time. It is the product of Availability, Performance and Quality:

The three factors are usually defined as:

Availability captures losses from breakdowns and setup or adjustment time. Performance captures speed losses and small stops. Quality captures rejects and rework, according to the organisation’s counting rules. Each factor is expressed as a decimal or percentage, and all three must refer to the same period and equipment boundary.

OEE is not a universal productivity score and should not be compared across sites without consistent definitions, product mix and planned-time rules. A high OEE can still coexist with an economically poor system if the asset is overproducing, consuming excessive energy or producing the wrong items.

Worked Example 3: OEE

A machine has 480 minutes of planned production time. It runs for 420 minutes, produces 800 units, has an ideal cycle time of 0.5 minutes per unit and produces 760 good units.

The largest percentage loss is not necessarily the largest financial loss. Management should examine the economic impact of each loss before selecting improvement projects.

5. Asset Life-Cycle Management

Asset life-cycle management coordinates decisions from conception to disposal. The stages commonly include:

1. Need and specification: define required capacity, operating environment, safety, quality and maintainability.
2. Design and acquisition: evaluate suppliers, whole-life cost, energy use, spares, data access and support.
3. Installation and commissioning: verify foundations, utilities, settings, documentation and acceptance criteria.
4. Operation and maintenance: control work, monitor condition, manage risks and improve performance.
5. Renewal or modification: decide whether to repair, upgrade, replace, redesign or change the operating model.
6. Decommissioning and disposal: remove safely, protect information, recover materials and meet environmental obligations.

The acquisition price is only one element of life-cycle cost. A simplified model is:

The terms represent acquisition, installation, operating, maintenance, downtime and disposal costs, less residual value. The equation is a cost framework rather than a complete investment appraisal. For long-lived assets, future cash flows should be discounted:

Here, CF_t is the net cash flow in period t, r is the discount rate per period and n is the analysis horizon. Costs may be entered as negative cash flows, or benefits and costs may be shown separately. Assumptions concerning inflation, taxation, energy prices, residual value and risk should be documented.

Criticality-Based Asset Management

Asset criticality is the consequence of failure combined with the likelihood and detectability of failure. A practical assessment considers:

• safety and environmental consequences;
• effect on customer service and production;
• quality and regulatory consequences;
• repair time and spare-part lead time;
• availability of alternative capacity;
• financial exposure; and
• vulnerability to common-cause events.

Criticality should determine the depth of analysis and control. It does not automatically mean that the most expensive predictive technology is justified. A critical asset may instead require redesign, redundancy, a stocked spare, improved procedures or a legally mandated inspection.

Spare Parts and Inventory Decisions

Maintenance and supply-chain decisions are closely connected. Holding no spare may increase downtime; holding too many ties up working capital and creates obsolescence. Decisions should consider failure rate, lead time, part criticality, interchangeability, storage conditions and supplier reliability.

A critical spare may be justified even when demand is infrequent because the cost of stockout is extreme. Conversely, a common low-value part may be sourced rapidly. Maintenance planning should integrate bills of material, equipment hierarchies, supplier information and work-order history in the enterprise asset management system.

6. Data, Digital Technologies and AI

Computerised Maintenance Management Systems (CMMS) and Enterprise Asset Management (EAM) platforms support asset registers, work orders, schedules, labour, materials, costs and compliance records. Good software cannot compensate for poor master data. Each asset should have a unique identity, functional location, relevant specifications and a consistent failure and remedy coding structure.

Sensors and industrial Internet of Things technologies can provide continuous measurements of temperature, pressure, vibration, energy and operating hours. Analytics can detect deviations from normal behaviour and estimate remaining useful life. Machine-learning models may assist with anomaly detection, but they require representative data, stable operating conditions and controls against false alarms.

AI-supported maintenance should be treated as decision support rather than an unquestioned authority. Important controls include:

• validation against known failure modes;
• human review of high-consequence decisions;
• cybersecurity and access control;
• protection of operational and personal data;
• monitoring for model drift;
• explainable escalation rules; and
• records showing why a maintenance decision was made.

7. Applied Case: Illustrative Food-Processing Packaging Line

Background

A fictional food manufacturer operates a packaging line containing a filler, capper, labeller and conveyor system. The line is scheduled for 16 hours per day. Unplanned stoppages cause missed dispatch windows and occasional product waste.

Problem

During three months, the line recorded repeated capper failures. The organisation responded reactively, replacing bearings after seizure. Average downtime was increasing, and maintenance planners frequently requested emergency parts. Operators also reported unusual vibration but had no formal inspection standard.

Analysis

The engineering team classified the capper as high criticality because failure stopped the complete line and could create hygiene and quality risks. Failure records showed that bearing seizure was preceded by increasing vibration and temperature. A simple availability review showed 36 hours of downtime against 1,800 scheduled hours:

Availability was 98%, but the figure concealed clustered failures, overtime and missed deliveries. The team therefore examined downtime distribution, failure causes, spare-part lead times and quality consequences rather than relying on one KPI.

Approach and Solution

The organisation implemented the following measures:

1. Operators received training in cleaning, lubrication checks and abnormal-noise reporting.
2. A weekly vibration and temperature inspection was introduced.
3. Alarm and action limits were defined with engineering approval.
4. Bearings were replaced according to condition and installation quality, rather than automatically at short intervals.
5. A critical spare bearing and seal kit was stocked.
6. Root-cause analysis identified misalignment during previous installations.
7. A standard work instruction and torque specification were introduced.
8. The CMMS captured failure mode, cause, remedy, downtime and parts used.
9. OEE losses were separated into breakdown, speed, setup and quality categories.

Outcome

In the next quarter, repeated seizure events declined, emergency purchases were reduced and planned interventions increased. The principal improvement did not come from sensors alone. It resulted from combining condition information, competent installation, operator involvement, critical-spare planning and disciplined data capture.

Learning Points

The case illustrates that maintenance improvement is socio-technical. A technically accurate condition signal has little value if nobody owns the response. It also shows why availability alone is insufficient: quality, speed, cost, safety and customer impact must be considered together.

8. Governance, Safety and Limitations

Maintenance work may involve isolation of energy, confined spaces, lifting, chemicals, electrical systems, pressure and work at height. Permit-to-work systems, lockout/tagout or equivalent isolation controls, competency verification and post-maintenance testing are essential. Production urgency must never override a safe isolation requirement.

Records should demonstrate that inspections, statutory tests, calibration and safety-critical maintenance were completed. Contractors require qualification, induction, supervision and clear responsibility boundaries. Environmental responsibilities may include refrigerant control, oil disposal, hazardous waste, noise and energy efficiency.

Important limitations include:

• reliability statistics can be distorted by incomplete failure records;
• MTBF can hide multiple failure modes and does not predict the next failure for every asset;
• OEE can encourage local optimisation or misleading data classification;
• predictive models can generate false positives and false negatives;
• manufacturer maintenance intervals may not reflect actual operating conditions; and
• maintenance cost reduction can increase long-term risk if critical work is deferred.

Best practice is to define measures precisely, segment assets by function and criticality, verify data quality, review failure modes, involve the people who operate equipment and connect maintenance decisions to business and safety outcomes.

Chapter Summary

Maintenance management protects operational continuity, quality, safety and supply-chain performance. Reactive maintenance responds after failure; preventive maintenance uses predetermined intervals; predictive maintenance forecasts deterioration using data; and condition-based maintenance acts when measured condition indicates a need. No strategy is universally superior: the appropriate mix depends on failure behaviour, consequences, detectability, cost and risk.

Reliability expresses the probability of successful performance over time. Maintainability concerns restoration after failure, while availability combines operating and repair behaviour. Under a constant failure-rate assumption, reliability is R(t) = e^(−λt). For a simplified repairable system, availability is A = MTBF ÷ (MTBF + MTTR).

TPM creates shared responsibility for equipment effectiveness through autonomous maintenance, planned maintenance, focused improvement, training, quality maintenance and safe work. OEE decomposes effectiveness into Availability, Performance and Quality, but it must be interpreted with consistent definitions and economic context.

Asset life-cycle management begins with need and specification and continues through acquisition, commissioning, operation, renewal and disposal. Whole-life decisions should consider capital, operating, maintenance, downtime, risk, sustainability and residual value. CMMS, EAM, sensors and AI can improve decisions, but only when supported by reliable data, competent people, cybersecurity and human governance.

Key Terms

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

Failure: Inability of an asset to perform its required function.

Fault: An abnormal condition that may precede or cause failure.

Reactive maintenance: Maintenance performed after a failure or defect has occurred.

Preventive maintenance: Maintenance performed at predetermined intervals or criteria.

Predictive maintenance: Maintenance supported by analysis that forecasts future deterioration or failure.

Condition-based maintenance: Maintenance triggered by evidence of actual asset condition.

Maintainability: Probability of restoring an item within a stated time under stated conditions.

MTBF: Mean Time Between Failures for a repairable asset.

MTTR: Mean Time To Repair or restore an asset, according to a defined measurement boundary.

Availability: Probability or proportion of time that an asset is able to perform its required function.

Failure rate: Frequency or hazard of failure, commonly represented by λ.

TPM: Total Productive Maintenance, an organisation-wide approach to maximising equipment effectiveness.

OEE: Overall Equipment Effectiveness, calculated from Availability, Performance and Quality.

Criticality: Relative importance of an asset based on failure consequences and risk.

CMMS: Computerised Maintenance Management System used to manage maintenance information and work.

EAM: Enterprise Asset Management system supporting whole-life asset decisions.

Life-cycle cost: Total cost associated with acquiring, operating, maintaining and disposing of an asset.

Common-cause failure: Failure affecting multiple supposedly independent components through a shared cause.

Review Questions

5 Short Answer Questions

1. What is the principal difference between preventive and condition-based maintenance?
2. Define reliability and distinguish it from maintainability.
3. What does MTTR measure?
4. Name the three components of OEE.
5. Why should asset criticality influence maintenance policy?

5 Long Answer Questions

1. Compare reactive, preventive, predictive and condition-based maintenance, including advantages, limitations and suitable applications.
2. Explain the assumptions behind the constant-failure-rate reliability model and discuss when those assumptions may be inappropriate.
3. Describe the major pillars and organisational implications of TPM.
4. Explain how life-cycle cost analysis supports asset acquisition and replacement decisions.
5. Discuss how CMMS, sensors and AI can improve maintenance while creating new governance risks.

3 Analytical / Application Questions

1. A repairable machine has MTBF of 360 hours and MTTR of 15 hours. Calculate its simplified availability and explain two practical limitations of the result.
2. A production line has 600 minutes of planned time, 540 minutes of run time, an ideal cycle time of 0.4 minutes, total output of 1,200 units and 1,140 good units. Calculate Availability, Performance, Quality and OEE.
3. A warehouse conveyor is inexpensive but causes major customer delays when it fails. Recommend a maintenance strategy and justify your recommendation using criticality, condition monitoring, spares and life-cycle cost considerations.

Multiple Choice Questions with Answers

Question 1

Which maintenance strategy is triggered primarily by evidence of actual equipment condition?

A. Reactive maintenance

B. Condition-based maintenance

C. Calendar-only maintenance

D. Emergency replacement

Correct Answer: B. Condition-based maintenance

Question 2

A component has a constant failure rate of 0.001 failures per hour. Under the exponential reliability model, what is its reliability after 100 hours?

A. Approximately 0.368

B. Approximately 0.500

C. Approximately 0.905

D. Approximately 0.999

Correct Answer: C. Approximately 0.905

Question 3

A machine has MTBF of 180 hours and MTTR of 20 hours. What is its simplified availability?

A. 80.0%

B. 85.0%

C. 90.0%

D. 95.0%

Correct Answer: C. 90.0%

Question 4

Which OEE factor captures rejects and rework?

A. Availability

B. Performance

C. Quality

D. Maintainability

Correct Answer: C. Quality

Question 5

A machine has 500 minutes of planned production time and runs for 450 minutes. What is its Availability component of OEE?

A. 90%

B. ใน?

C. 10%

D. 111.1%

Correct Answer: A. 90%

Question 6

Which statement best describes Total Productive Maintenance?

A. It assigns every maintenance task exclusively to engineers.

B. It focuses only on purchasing replacement machines.

C. It involves the organisation in maximising equipment effectiveness.

D. It eliminates the need for preventive maintenance.

Correct Answer: C. It involves the organisation in maximising equipment effectiveness.

Question 7

Which cost is most directly associated with whole-life asset management but often omitted from purchase-price comparisons?

A. Operating energy cost

B. Logo design cost

C. Employee social events

D. Unrelated office rent

Correct Answer: A. Operating energy cost

Question 8

For a series system of independent components, system reliability is generally:

A. The sum of component reliabilities

B. The product of component reliabilities

C. Always greater than every component reliability

D. Equal to one minus the sum of failure rates in all cases

Correct Answer: B. The product of component reliabilities

Question 9

Which action most directly improves maintainability?

A. Reducing access to the failed component

B. Using modular replacement and clear diagnostic procedures

C. Removing maintenance records

D. Increasing unplanned production interruptions

Correct Answer: B. Using modular replacement and clear diagnostic procedures

Question 10

A line has Availability of 0.80, Performance of 0.90 and Quality of 0.95. What is its OEE?

A. 68.4%

B. 72.0%

C. 76.0%

D. 85.5%

Correct Answer: A. 68.4%