11. Maintenance, Reliability and Asset Management

Introduction and Context

Operations management is concerned not only with designing and running processes, but also with sustaining the physical assets on which those processes depend. Warehouses, fleets, cranes, conveyor systems, refrigeration units, information systems, material-handling equipment and production facilities all deteriorate through use, environmental exposure, poor operation and obsolescence. When an asset fails, the consequence may extend beyond repair cost: shipments may be delayed, inventory may be damaged, employees may be exposed to hazards, and customer service commitments may be breached.

Maintenance is therefore a strategic operations activity rather than a purely technical function. Reliability concerns the probability that an asset will perform its intended function for a specified period under stated conditions. Availability concerns whether the asset is capable of performing when required. Asset management extends the perspective across the whole life of the asset—from need identification and acquisition to operation, maintenance, renewal and disposal. In logistics, these ideas are especially important because operational continuity depends on interconnected assets and tightly timed flows.

A useful managerial principle is that maintenance decisions should be based on risk and life-cycle value, not simply on the immediate cost of a repair. A low-cost maintenance policy can produce high costs through downtime, emergency transport, lost sales or safety incidents. Conversely, excessive preventive work can waste resources and introduce maintenance-induced failures. Effective asset management balances performance, risk, cost and sustainability.

Learning Objectives

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

1. Distinguish maintenance, reliability, availability, maintainability and asset management.

2. Calculate and interpret important reliability and maintenance indicators.

3. Explain failure patterns and select suitable preventive or predictive strategies.

4. Apply total productive maintenance and risk-based maintenance principles.

5. Analyse replacement, spare-parts and life-cycle decisions.

6. Evaluate the role of digital technologies, analytics and governance in asset management.

7. Recommend maintenance policies for operational continuity in logistics environments.

The Language of Asset Performance

Reliability is commonly represented by R(t), the probability that an item survives without failure up to time t. If failures follow a constant exponential failure rate λ, then:

R(t) = e^(-λt)

The failure probability by time t is F(t) = 1 − R(t). Mean time between failures (MTBF) is widely used for repairable equipment, while mean time to failure (MTTF) is more suitable for non-repairable items. Mean time to repair (MTTR) measures the average time required to restore a failed asset, including diagnosis, access, repair, testing and return to service.

Availability links reliability with maintainability. For a repairable asset under stable operating conditions, inherent availability may be approximated as:

Availability = MTBF / (MTBF + MTTR)

For example, equipment with an MTBF of 900 hours and an MTTR of 6 hours has an approximate availability of 900/906, or 99.34 percent. This calculation does not include waiting for spare parts, technicians or production access. Operational availability is consequently often lower than inherent availability.

Overall equipment effectiveness (OEE) combines availability, performance and quality:

OEE = Availability × Performance × Quality

A machine may be available for most of a shift but still have low OEE because it operates below its designed speed or produces defective output. OEE is therefore a diagnostic measure, not a substitute for reliability analysis. It should be used to identify losses rather than to encourage manipulation of production figures.

Maintainability describes the ease and speed with which an asset can be restored. Design features such as modular components, diagnostic access, standard fasteners and safe isolation points improve maintainability. Dependability is a broader concept incorporating reliability, availability, maintainability and supportability. Supportability includes tools, documentation, training, spare parts and technical assistance.

Other useful measures include planned maintenance percentage, schedule compliance, emergency-work percentage, maintenance cost per operating hour, backlog age, mean time between stoppage and mean time to detect. No single indicator gives a complete picture. High MTBF accompanied by a large maintenance backlog may conceal deteriorating risk; high schedule compliance may simply indicate that an inappropriate schedule is being followed consistently.

Failure Behaviour and the Logic of Maintenance

A failure occurs when an asset cannot perform a required function. Functional failure is defined relative to the required standard. A conveyor that moves cartons at half the required rate may be functionally failed even if its motor is still running. Failure causes may be physical deterioration, incorrect operation, design weakness, contamination, software error, environmental stress or inadequate maintenance.

The bathtub curve is a useful, though simplified, model of failure behaviour. It contains three regions:

• Early-life failures, often associated with design defects, installation errors or commissioning problems.
• A useful-life period with approximately random failures.
• A wear-out period in which ageing, fatigue, corrosion or depletion increases failure probability.

Not all assets follow the bathtub curve. Research in reliability-centred maintenance has shown that many complex systems experience age-independent or mixed failure patterns. Consequently, replacing every item at a fixed age may not improve reliability and can create unnecessary cost or maintenance-induced damage.

Failure mode and effects analysis (FMEA) examines how an asset can fail, the effects of each failure, its causes, existing controls and the priority of corrective action. A traditional risk priority number is calculated by multiplying severity, occurrence and detectability ratings. This score is useful for discussion but should not be treated as an objective probability. A low-frequency failure with catastrophic consequences may deserve priority even if its numerical score is modest.

Fault-tree analysis works in the opposite direction. It begins with an undesirable top event, such as loss of cold-chain control, and logically decomposes the combinations of failures that could produce it. Event-tree analysis begins with an initiating event and examines possible consequence paths. These methods are particularly useful for safety-critical and highly interconnected systems.

Maintenance Policies and Their Appropriate Use

Corrective maintenance restores an asset after failure. Run-to-failure is rational where the item is inexpensive, non-critical, easily replaced and unlikely to create safety or operational consequences. It is inappropriate for assets whose failure interrupts a network, damages products or creates serious risk.

Preventive maintenance is performed at predetermined intervals or according to predetermined usage. Time-based inspection, lubrication, calibration and replacement can reduce wear-out failures. However, an interval that is too short raises labour and parts cost, while an interval that is too long leaves exposure to failure. Usage-based maintenance—such as servicing after operating hours, kilometres or cycles—can be more meaningful than calendar intervals.

Predictive or condition-based maintenance uses evidence about actual asset condition. Vibration analysis can identify bearing defects; infrared thermography can reveal abnormal heat; oil analysis can detect contamination and wear particles; ultrasonic inspection can identify leaks or electrical discharge; and motor-current analysis can indicate electrical or mechanical problems. Predictive maintenance is valuable when failure development can be detected in time for planned intervention and when the consequences of failure justify monitoring costs.

Prescriptive maintenance goes further by recommending an action based on predicted condition, operational context, risk and resource availability. It may advise whether to reduce load, inspect immediately, defer work until a planned shutdown or replace a component. Such recommendations require reliable data and clear accountability.

Reliability-centred maintenance (RCM) asks seven practical questions: what are the required functions; how can the asset fail; what causes each failure; what happens when it fails; why does the failure matter; what task can manage the failure; and what should be done if no suitable preventive task exists? The resulting policy may be condition-based, scheduled restoration, scheduled discard, failure-finding, redesign or deliberate run-to-failure.

Risk-based maintenance prioritises work according to the combination of probability and consequence. A minor failure on a non-critical office printer may wait, while a developing fault in a refrigerated pharmaceutical vehicle may require immediate action. Criticality assessment should consider safety, regulatory exposure, customer service, financial impact, environmental harm, recovery time and the existence of alternative capacity.

Total Productive Maintenance and Human Involvement

Total productive maintenance (TPM) treats equipment effectiveness as a shared responsibility of operators, maintenance personnel, engineers and managers. Its purpose is not merely to increase maintenance activity but to eliminate chronic losses and develop ownership of equipment condition. Operators may perform cleaning, inspection, tightening and basic lubrication, while specialists handle diagnosis, overhaul and engineering modifications.

TPM commonly includes autonomous maintenance, focused improvement, planned maintenance, quality maintenance, early equipment management, training, safety and office or administrative improvement. The five-S method—sort, set in order, shine, standardise and sustain—supports visual control and early detection of abnormalities. Small-group improvement activities can address repeated minor stops, changeover losses or contamination.

TPM fails when operators are given responsibility without time, training, authority or feedback. It also fails when management measures only short-term output and treats basic care as work that can be skipped during busy periods.

Spare-Parts Planning and Maintenance Work Management

Spare-parts policy is a balance between stockout risk and inventory carrying cost. Criticality, demand frequency, lead time, supplier reliability, repairability, commonality, shelf life and consequence of failure should influence stocking decisions. A low-demand item may still require local availability if its absence can stop a critical operation for several weeks.

Items can be classified as consumables, repairables, rotable components, insurance spares and strategic spares. Critical spares require accurate bills of material, preservation controls, traceability and periodic verification. Obsolete or duplicated stock ties up capital and can create disposal problems, but indiscriminate reduction of inventory may transfer cost to emergency procurement and downtime.

Maintenance work management normally includes identification of a defect, screening and prioritisation, planning, scheduling, execution, testing and close-out. A good work order records the asset, symptom, cause, work performed, parts used, labour, downtime and recommendations. Backlog should be reviewed by age, risk and craft requirement rather than by volume alone. Planned work improves safety and productivity because tools, permits, isolation and parts can be prepared in advance.

Replacement and Life-Cycle Decisions

An asset should not be replaced solely because it has reached a traditional accounting life or because repair has occurred several times. Replacement analysis considers future operating cost, reliability, energy use, capacity, technology, safety, environmental performance and the opportunity cost of downtime.

For alternatives with different purchase prices and service lives, equivalent annual cost (EAC) can support comparison. It converts capital cost and expected future costs into an annual equivalent, allowing alternatives to be compared on a consistent basis. Discounted cash-flow analysis should include acquisition, installation, training, maintenance, energy, downtime, disposal and residual value. The relevant comparison is incremental life-cycle cost, not only the purchase price.

The physical life of an asset may exceed its economic life. An old vehicle may remain operable but consume excessive fuel, require scarce parts and expose the organisation to unreliable service. Conversely, premature replacement may discard useful capacity and embed emissions associated with manufacture. Life-cycle assessment and total cost of ownership encourage more balanced decisions.

Asset management systems such as ISO 55000 emphasise alignment between organisational objectives, asset portfolios, risk and value. Asset strategies should connect corporate goals—such as delivery reliability or cold-chain compliance—to asset performance standards, investment plans and maintenance policies. Governance requires defined ownership, documented decision criteria and auditable records.

Digital Maintenance and Emerging Practice

Computerised maintenance management systems (CMMS) organise asset registers, work orders, preventive schedules, labour, parts and history. Enterprise asset management systems connect maintenance with procurement, finance, projects, inventory and compliance. Their value depends on data quality: an advanced platform cannot compensate for missing equipment hierarchies, inaccurate meter readings or inconsistent failure codes.

Internet-of-Things sensors enable continuous monitoring of temperature, vibration, location, pressure and energy consumption. Cloud platforms can aggregate data across depots and compare asset populations. Machine-learning models may identify abnormal patterns, but predictions should be validated against engineering knowledge and operational consequences. False alarms create alert fatigue, while missed alarms create false confidence.

Digital twins can represent an asset or network using operational and engineering data to simulate performance and maintenance alternatives. Mobile devices improve field access to procedures, drawings and work orders. Drones and robotic inspection can reduce exposure in hazardous or inaccessible locations. Cybersecurity is essential because connected assets can become pathways into operational technology. Access control, patch governance, network segmentation, secure data ownership and recovery planning are maintenance concerns as well as information-technology concerns.

Illustrative Case: Reliability Improvement in a Regional Distribution Centre

Background

The following is an illustrative case. A regional distribution centre processes consumer goods through a network of powered conveyors, scanners and automated sortation equipment. Customer orders are cut off at 20:00 for next-day delivery. The centre has experienced repeated evening stoppages on one conveyor line.

Situation and Analysis

The line recorded an apparent availability of 97 percent, but operational availability was lower because technicians often waited for a replacement motor or access to the line. Failure records were inconsistent: operators described symptoms such as “jam,” “motor trip” and “belt problem” without identifying root causes. A review found that minor belt misalignment caused increasing friction, which overheated bearings and eventually triggered motor protection. Emergency work represented 42 percent of maintenance hours, and a critical bearing was not stocked because its annual demand was low.

The team mapped the process and assessed asset criticality. The conveyor was classified as high criticality because there was no practical bypass during the evening dispatch window. FMEA identified misalignment, contamination, poor tension and bearing deterioration as significant failure modes. Vibration monitoring was judged worthwhile for drive-end bearings, while visual inspection and cleaning were assigned to operators.

Approach and Solution

The centre introduced a standard inspection route at shift start, with photographs showing acceptable belt tracking and tension. Operators received training in abnormality detection and escalation. Maintenance planners created a preventive task for alignment verification and a condition-based task using monthly vibration readings. The bearing was classified as an insurance spare and stored under controlled conditions. Work orders were revised to distinguish symptom, failure mode and verified cause.

A short daily meeting reviewed open abnormalities, parts constraints and the next dispatch risk. The maintenance team also changed the motor mounting arrangement to improve access and reduced repair time by preparing lifting equipment and isolation procedures in advance.

Outcome

Over the following operating period, repeated motor trips declined, emergency work fell and repair duration shortened. The most important improvement did not come from purchasing more sensors; it came from combining criticality analysis, operator involvement, data discipline and preparation. The centre still retained a run-to-failure policy for low-value, non-critical rollers, demonstrating that the policy was differentiated rather than uniformly intensive.

Learning Points

The case illustrates that reliability improvement is a system problem. A technically superior component cannot compensate for poor installation, weak work management or unavailable spares. It also shows why maintenance indicators must be interpreted with operational context and why low-demand spares may be economically justified when downtime consequences are high.

Challenges, Risks and Professional Considerations

Common mistakes include confusing preventive activity with reliability improvement, treating MTBF as a guarantee, using OEE as the only performance measure, replacing components on arbitrary calendars, ignoring human factors and failing to close work orders with meaningful failure information. Maintenance-induced failures may result from incorrect torque, contamination, wrong parts, poor isolation or inadequate testing.

Safety must take precedence over production urgency. Lockout/tagout or equivalent energy-isolation procedures, permit-to-work systems, confined-space controls and competency requirements protect workers and contractors. Maintenance records should support regulatory compliance, warranty claims and incident investigation. Organisations should avoid pressuring technicians to conceal failures or defer safety-critical work to meet superficial performance targets.

Ethical asset management includes honest reporting of condition, transparent investment assumptions, responsible disposal of hazardous materials and consideration of environmental impacts. Procurement decisions should be protected from conflicts of interest. Automated recommendations should remain reviewable, and accountability should not be transferred to an algorithm.

Best-Practice Principles

Effective organisations generally:

• define asset criticality in relation to business consequences;
• maintain accurate asset registers, hierarchies and failure codes;
• combine operator care, planned maintenance and condition monitoring appropriately;
• prioritise work by risk rather than by who complains most loudly;
• integrate spare-parts decisions with maintenance and supply-chain planning;
• measure both technical outcomes and business outcomes;
• investigate recurring failures through root-cause analysis;
• design maintainability and data requirements into new assets;
• review maintenance policies after changes in demand, technology or risk; and
• use digital tools to improve decisions rather than to automate poor processes.

Chapter Summary

Maintenance preserves the function of operational assets, while reliability expresses the likelihood of failure-free performance and availability reflects readiness for use. MTBF, MTTR, OEE and related indicators support diagnosis but must be interpreted in context. Failure behaviour varies, so preventive replacement should not be assumed to be universally effective. Corrective, preventive, predictive, prescriptive, reliability-centred and risk-based approaches should be selected according to failure behaviour and consequence.

TPM extends responsibility for equipment condition across the organisation. Spare-parts planning, work management and replacement analysis connect maintenance with inventory, finance and service performance. Life-cycle asset management aligns acquisition, operation, renewal and disposal with organisational objectives. CMMS, sensors, analytics and digital twins can improve visibility, but data quality, cybersecurity, human judgement and governance remain essential. The central managerial task is to achieve continuity and value while controlling risk, cost and resource use.

Key Terms

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

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

Maintainability: Ease and speed with which an asset can be restored after failure.

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

MTTR: Mean time required to repair and restore an asset.

MTTF: Mean time to failure, generally used for non-repairable items.

OEE: Product of availability, performance and quality.

Corrective maintenance: Maintenance performed after a failure or detected defect.

Preventive maintenance: Planned maintenance based on time, usage or predetermined intervals.

Predictive maintenance: Maintenance triggered by evidence of actual asset condition.

Condition monitoring: Measurement used to detect deterioration or abnormal behaviour.

Reliability-centred maintenance: A method that selects maintenance tasks according to functions, failure modes and consequences.

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

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

TPM: Organisation-wide approach to maximising equipment effectiveness through shared responsibility.

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

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

Insurance spare: Low-demand spare held because its unavailability would create unacceptable risk.

Asset management: Coordinated activity for obtaining value from assets throughout their life cycle.

Review Questions

#### Short Answer Questions

1. Distinguish reliability from availability.

2. Why can operational availability be lower than inherent availability?

3. What is the purpose of FMEA?

4. State two conditions under which run-to-failure may be appropriate.

5. Identify three types of condition-monitoring technology.

#### Long Answer Questions

1. Explain the bathtub curve and discuss why it should not be used as an automatic justification for age-based replacement.

2. Compare corrective, preventive, predictive and prescriptive maintenance approaches.

3. Explain how reliability-centred maintenance supports maintenance-policy selection.

4. Discuss the contribution of TPM to operational continuity and its implementation risks.

5. Explain how total cost of ownership improves asset replacement decisions.

#### Analytical / Application Questions

1. A repairable vehicle component has an MTBF of 1,200 hours and an MTTR of 12 hours. Calculate approximate inherent availability and explain why actual service availability may differ.

2. A distribution centre must choose between holding a low-demand motor locally or accepting a ten-day supplier lead time. Develop a decision framework using criticality, consequence, cost and alternatives.

3. A conveyor shows rising vibration but has not yet failed. Recommend an action using risk-based and predictive-maintenance principles, including the data and operational conditions you would review.

22. Multiple Choice Questions with Answers

Question 1

A conveyor has an MTBF of 900 hours and an MTTR of 6 hours. Its approximate inherent availability is:

A. 90.00%

B. 96.00%

C. 99.34%

D. 99.99%

Correct Answer: C. 99.34%

Question 2

Which maintenance policy is most appropriate when failure development can be detected with sufficient warning and failure consequences justify monitoring costs?

A. Run-to-failure

B. Predictive maintenance

C. Unplanned corrective maintenance

D. Arbitrary annual replacement

Correct Answer: B. Predictive maintenance

Question 3

The primary purpose of reliability-centred maintenance is to:

A. Maximise the number of preventive tasks

B. Select tasks according to functions, failure modes and consequences

C. Replace every component at the same age

D. Eliminate the need for operators

Correct Answer: B. Select tasks according to functions, failure modes and consequences

Question 4

Which factor most directly explains why operational availability may be lower than inherent availability?

A. The asset has a high purchase price

B. Repair calculations exclude waiting for parts and technicians

C. The asset has a documented hierarchy

D. The asset is monitored by sensors

Correct Answer: B. Repair calculations exclude waiting for parts and technicians

Question 5

In the illustrative distribution-centre case, the most important initial cause of repeated motor trips was:

A. Excessive inventory accuracy

B. Belt misalignment increasing friction and bearing heat

C. An oversized spare-parts warehouse

D. A reduction in customer demand

Correct Answer: B. Belt misalignment increasing friction and bearing heat

Question 6

Which statement best describes an insurance spare?

A. A frequently consumed lubricant

B. A low-demand part held because stockout consequences are severe

C. A part discarded after every inspection

D. A component purchased only after failure

Correct Answer: B. A low-demand part held because stockout consequences are severe

Question 7

OEE is calculated as:

A. Reliability plus maintainability plus safety

B. MTBF divided by MTTR

C. Availability multiplied by performance and quality

D. Maintenance cost divided by inventory value

Correct Answer: C. Availability multiplied by performance and quality

Question 8

A key risk of using machine-learning predictions without engineering review is:

A. Elimination of all maintenance records

B. False alarms or missed alarms leading to poor decisions

C. Guaranteed reduction in spare-parts demand

D. Automatic improvement in asset design

Correct Answer: B. False alarms or missed alarms leading to poor decisions

Question 9

Which action best reflects total productive maintenance?

A. Restricting all equipment knowledge to the maintenance department

B. Having operators perform basic cleaning and abnormality checks with appropriate training

C. Replacing all assets annually

D. Measuring only emergency maintenance hours

Correct Answer: B. Having operators perform basic cleaning and abnormality checks with appropriate training

Question 10

When comparing assets with different service lives, equivalent annual cost is useful because it:

A. Ignores disposal costs

B. Compares alternatives on a consistent annual basis

C. Measures only purchase price

D. Eliminates the need for discounting

Correct Answer: B. Compares alternatives on a consistent annual basis