Asset Governance, Leadership and Future Practice

Introduction

Maintenance has traditionally been treated as a technical function concerned with repairing equipment and preserving availability. In contemporary production systems, this view is inadequate. Assets consume capital, energy, materials, labour and environmental capacity throughout their life cycles. Their performance affects product quality, worker safety, regulatory compliance, customer service and organisational reputation. Asset governance therefore concerns the direction, control and accountability of physical-asset decisions, while maintenance leadership converts those decisions into reliable operational practice.

This chapter examines how predictive and preventive maintenance can be aligned with ISO 55000 asset-management principles, safety and environmental obligations, organisational change, ethical responsibilities and future technologies. It also considers autonomous maintenance and self-optimising assets. The central argument is that maintenance excellence is not produced by sensors alone. It is produced by a coherent system of policy, leadership, competence, data, risk management and learning.

Learning objectives

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

1. Explain the relationship between asset governance, asset management and maintenance.
2. Apply ISO 55000 principles to maintenance decision-making.
3. Design governance arrangements that integrate safety, environment, cost and performance.
4. Select meaningful maintenance and asset-performance indicators.
5. Lead organisational change, autonomous maintenance and workforce development.
6. Evaluate ethical, sustainability and regulatory issues in predictive maintenance.
7. Critically assess autonomous and self-optimising asset systems.

1. Asset governance as a strategic discipline

Asset management is the coordinated activity of an organisation to realise value from assets. Value is not limited to financial return. It includes performance, risk reduction, compliance, safety, environmental stewardship, resilience and stakeholder outcomes. Governance is the system by which authority is exercised, decisions are justified, performance is monitored and accountability is enforced.

A useful distinction is:

• Corporate governance: direction and control of the organisation as a whole.
• Asset governance: policies, decision rights and controls governing asset-related decisions.
• Asset management: coordinated life-cycle activities that realise value from assets.
• Maintenance management: planning, execution and improvement of activities that preserve or restore required function.
• Maintenance execution: the technical work performed on equipment and systems.

Maintenance is consequently one component of asset management, not its synonym. Replacing a pump, changing an operating procedure, redesigning a process or retiring an unsafe asset may all be better asset-management decisions than repeatedly repairing the original pump.

The value–risk–cost balance

Asset decisions involve competing objectives. A low-cost maintenance regime may increase failure risk; excessive preventive work may introduce human error and unnecessary expenditure; operating equipment at maximum output may reduce its life. Governance makes these trade-offs explicit.

A simple decision model is:

where J is the decision burden or total relevant consequence, C_L is life-cycle cost, C_R is risk-related consequence, C_E is environmental consequence and C_Q is quality or service consequence. This is a conceptual additive model, not a universal accounting identity. The terms must be defined consistently, and monetary conversion of safety or ecological harm should not be used to suggest that human life or legal duties are merely exchangeable costs. In practice, legal and safety constraints may be treated as non-negotiable limits rather than weighted costs.

2. ISO 55000 principles and maintenance alignment

The ISO 55000 family provides an internationally recognised vocabulary and management-system perspective for asset management. ISO 55000 introduces concepts and principles; ISO 55001 specifies requirements for an asset-management system; ISO 55002 provides guidance for implementation. The standards do not prescribe a particular maintenance tactic or guarantee high performance. They require a systematic, auditable approach linked to organisational objectives.

Important principles include:

1. Value: assets exist to create value for the organisation and stakeholders.
2. Alignment: asset-management objectives should derive from organisational objectives.
3. Leadership: senior leaders establish policy, roles, resources and accountability.
4. Assurance: decisions and controls should provide confidence that objectives will be achieved.
5. Life-cycle thinking: acquisition, operation, maintenance, modification and disposal are interconnected.
6. Risk-based decision-making: uncertainty and consequences should influence priorities.
7. Systems thinking: assets interact with processes, people, technology and external conditions.

From policy to work order

A mature governance chain normally contains the following levels:

A maintenance department is aligned when a work order can be traced to an identified function, risk or performance requirement. This does not mean every task needs a complex business case. It means that task selection, frequency and method are technically defensible and connected to asset purpose.

Criticality and risk

Asset criticality should reflect consequence and likelihood, but a criticality ranking must not be confused with a precise probability of failure. A qualitative risk matrix can prioritise analysis, while high-consequence decisions may require quantitative reliability, consequence or safety assessment.

For an event with probability P and consequence magnitude C, expected loss is often represented as:

Here E[L] is expected loss, P is a probability between 0 and 1, and C is expressed in appropriate consequence units, often currency for economic analysis. The model assumes that probability and consequence are meaningfully estimated and that repeated outcomes can be represented by an expected value. It is not sufficient for catastrophic, irreversible or legally prohibited events, where low probability does not make the activity acceptable.

A governance process should therefore identify safety-critical and environmentally critical functions separately from ordinary production-critical functions. A machine that causes little production loss but can release toxic material requires a different control regime from a non-critical conveyor.

3. Safety, environmental and regulatory obligations

Maintenance changes the condition of equipment and therefore changes risk. Poorly controlled maintenance can defeat guards, disturb containment, introduce ignition sources, expose workers to stored energy or create latent defects. Safety must be designed into planning and execution rather than added during an audit.

Core controls include:

• isolation and lockout/tagout of hazardous energy;
• permit-to-work systems for confined space, hot work and line breaking;
• competence verification and supervision;
• management of temporary repairs and bypasses;
• inspection and testing of safety-critical devices;
• human-factors review of access, lifting, ergonomics and error likelihood;
• formal management of change after modifications;
• post-maintenance testing before return to service.

Environmental obligations include control of spills, emissions, waste, refrigerants, lubricants, contaminated components, noise and energy consumption. Preventive maintenance may reduce leakage and energy use, but maintenance itself can create waste and emissions. Life-cycle assessment should consider both effects.

A useful environmental performance indicator is maintenance-related waste intensity:

where W_I is waste intensity, W_M is maintenance waste mass, and Q is a clearly defined production or service denominator. Units may be kilograms per tonne of product or kilograms per operating hour. Comparisons are valid only when the denominator, boundary and waste categories remain consistent.

Compliance should be treated as a minimum condition. A site may achieve excellent availability while failing legally or ethically because it suppresses emissions, bypasses alarms or exposes contractors to unacceptable risk. Governance boards should therefore review safety and environmental leading indicators alongside production indicators.

4. Leadership and organisational change

Maintenance transformation is a socio-technical change. Introducing condition monitoring without changing planning discipline, data quality, authority or learning routines often produces expensive dashboards rather than better decisions.

Leadership responsibilities

Effective leaders:

• define why reliability matters to customers, workers and society;
• establish clear decision rights between operations, maintenance, engineering, safety and finance;
• protect long-term asset health from short-term production pressure;
• allocate resources for competence, analysis and improvement;
• reward reliable behaviour rather than heroic emergency repair;
• make failures discussable without confusing learning with blame avoidance;
• demonstrate that stop-work authority is real.

A practical governance structure may include an executive asset council, an asset-management leader, cross-functional reliability teams and frontline ownership. The structure must be proportionate: excessive committees delay decisions, while no cross-functional forum produces local optimisation.

Change process

A credible change programme commonly follows this sequence:

1. Establish the case for change using evidence of risk, cost and missed value.
2. Assess current maturity in strategy, data, work management, competence and culture.
3. Define a target operating model and measurable outcomes.
4. Pilot the model on a representative asset system.
5. Learn from the pilot and revise standards.
6. Scale through coaching, training and governance reviews.
7. Institutionalise the change in procedures, incentives and audits.

Resistance is not necessarily irrational. Operators may fear surveillance, technicians may distrust unreliable sensors, planners may anticipate additional administration and managers may fear transparency about deferred work. Leaders should involve affected workers in design, explain data use and remove redundant processes.

5. Autonomous maintenance and workforce capability

Autonomous maintenance, associated especially with total productive maintenance, enables operators to perform defined routine activities such as cleaning, inspection, lubrication, abnormality identification and basic tightening. It does not mean that operators independently perform hazardous specialist repairs. Boundaries, competence and escalation rules are essential.

A mature autonomous-maintenance arrangement includes:

• restored basic equipment condition;
• visual standards and abnormality tags;
• operator inspection routes;
• clear escalation thresholds;
• competent maintenance support;
• feedback on whether reported abnormalities are resolved;
• periodic audits of standards and behaviours.

The approach can improve early detection because operators observe equipment frequently. It can also fail if it transfers maintenance work without time, training or recognition. Workforce design should consider workload, physical demands, psychological safety and collective agreements.

Competence should be assessed by demonstrated capability rather than attendance at training alone. A competence matrix may distinguish awareness, supervised performance, independent performance and authority to approve work. Digital work instructions can support consistency, but they cannot replace judgement where conditions differ from the assumed task.

6. Measuring governance and maintenance performance

No single key performance indicator captures maintenance effectiveness. A balanced system should cover outcomes, leading controls and unintended consequences.

For a repairable asset observed over a defined period, a common operational availability approximation is:

where A_o is operational availability, U is uptime and D is downtime within the stated boundary. If uptime is 720 hours and downtime is 30 hours, then A_o = 720 ÷ 750 = 0.96, or 96%. The result is meaningful only if planned and unplanned downtime, logistics delays and organisational exclusions are defined consistently. Availability is not equivalent to reliability and should not be used alone to judge asset health.

Audit evidence should include records, field observations, interviews, competence assessments, work-quality checks and outcomes. A document audit can show that a procedure exists; it cannot prove that workers can safely execute the task or that the control is effective.

7. Ethics, data governance and workforce impacts

Predictive maintenance creates ethical questions beyond technical accuracy. Sensor and workforce data may reveal operator behaviour, productivity or location. If collected without transparency, monitoring can damage trust and encourage concealment. Organisations should define purpose, access, retention, consent where applicable and legitimate use of data.

Important ethical principles include:

• Safety first: algorithms must not recommend continued operation when credible evidence indicates unacceptable danger.
• Explainability: decision-makers should understand the basis and limitations of alerts.
• Accountability: responsibility remains with authorised people; it cannot be transferred to software.
• Fairness: automation should not systematically disadvantage particular workers, shifts or contractors.
• Privacy and security: operational and personal data require protection.
• Professional integrity: engineers must report uncertainty, weak evidence and conflicts of interest.
• Sustainability: technology should reduce total life-cycle harm, not merely shift impacts to data centres, batteries or electronic waste.

Cybersecurity is an asset-governance issue because connected sensors and remote access can affect physical safety. Segmentation, identity management, patching, secure vendor access, incident response and manual fallback arrangements should be integrated with maintenance planning.

8. Emerging self-optimising asset systems

A self-optimising asset system senses conditions, estimates state, predicts future behaviour and adjusts operating or maintenance decisions. A typical architecture contains:

1. sensors and data acquisition;
2. data validation and contextualisation;
3. diagnostic and prognostic models;
4. decision logic or optimisation;
5. human approval or automatic actuation;
6. feedback, monitoring and model revision.

Examples include variable-speed systems that adjust to demand, machine-learning models that estimate remaining useful life, and fleets that dynamically prioritise inspection resources. However, a prediction is not a decision. A remaining-useful-life estimate may be statistically accurate but operationally useless if the required spare part has a six-month lead time or if intervention creates greater risk than continued operation.

Model performance must be evaluated against a baseline and relevant decision costs. False negatives may be more dangerous than false positives for safety-critical equipment. Data drift, changing operating regimes, sensor degradation and previously unseen failure modes can invalidate models. Human override should be designed, logged and reviewed rather than treated as an informal exception.

A responsible deployment process includes:

• defining the decision and consequence before selecting an algorithm;
• establishing data quality and labelling requirements;
• testing on time-separated and operationally representative data;
• validating alerts in field trials;
• specifying confidence limits and escalation rules;
• monitoring drift and false-alert burden;
• retaining a safe manual mode;
• periodically reassessing whether the system creates net value.

9. Sustainability and future practice

Future maintenance practice will be shaped by circularity, decarbonisation, digital twins, robotics, additive manufacturing, advanced materials and industrial artificial intelligence. The strategic question is not whether a technology is novel, but whether it improves the required function with acceptable risk and life-cycle impact.

Circular asset strategies may prioritise repair, remanufacture, modular replacement and recovery of materials. These strategies require design information, spare-part traceability and quality assurance. Predictive maintenance can extend life, but extending the life of an inefficient or hazardous asset is not automatically sustainable. Whole-system analysis is required.

Digital twins can support scenario analysis by linking asset models, operating data and maintenance history. Their credibility depends on model validation, data lineage and explicit treatment of uncertainty. A visually impressive digital twin with inaccurate failure logic can produce worse decisions than a simple, well-maintained inspection programme.

Applied case: illustrative chemical-processing plant

Background

An illustrative chemical-processing plant operated three parallel cooling-water pumps. Production valued availability highly, while the safety team was concerned about loss of cooling during high-temperature operation. The plant had vibration sensors, but alerts were frequently ignored. Operators performed informal checks, planners carried a large backlog and the environmental register did not clearly identify pump-seal leakage as a reportable risk.

Problem

The organisation proposed purchasing more sensors. A cross-functional review found that the main problem was governance rather than sensing. Criticality was based only on production loss; safety and environmental consequences were missing. Alert ownership was unclear, work orders lacked failure codes and operators received no feedback after reporting abnormalities.

Analysis and approach

The asset council classified cooling as a safety- and production-critical function. It established a performance standard requiring one pump available and another capable of starting within a defined emergency response time. The maintenance strategy combined operator rounds, weekly vibration review, seal-leak inspection, periodic functional testing and condition-based replacement when evidence justified intervention.

The plant assigned alert ownership to the reliability engineer, created escalation limits for temperature, vibration and leakage, and required a documented decision for every high-severity alert. Operators were trained to identify abnormal noise, leakage and temperature, but specialist diagnosis remained with maintenance and engineering. A management-of-change review addressed sensor software, remote access and alarm priorities.

The team introduced a balanced dashboard: emergency work, overdue safety tasks, repeat failures, alert-to-action time, seal leakage, planned-work percentage and operator abnormality closure. It also audited field conditions rather than relying exclusively on digital records.

Outcome

After the pilot, nuisance alerts fell because thresholds were contextualised by operating state. Early leakage detection improved, and repeated seal failures led to a design change rather than repeated replacement. The plant did not claim that predictive analytics had eliminated failure; instead, it demonstrated stronger traceability from risk to strategy, work and assurance.

Learning points

1. Technology cannot compensate for unclear ownership.
2. Criticality must include safety and environmental consequences.
3. Autonomous inspection is effective when escalation and feedback are reliable.
4. A governance dashboard should reveal unwanted effects, not only production gains.
5. A design modification may be superior to an increasingly sophisticated maintenance task.

Challenges and limitations

Asset governance can become bureaucratic when approvals exceed the risk of the decision. Risk matrices can create false precision, and expected-value calculations can hide catastrophic outcomes. KPIs can drive dysfunctional behaviour when targets become incentives: technicians may close work orders prematurely, managers may defer difficult work, and teams may avoid reporting problems.

Predictive maintenance also has limits. Not every failure has a detectable precursor; intermittent faults can be difficult to model; sensors require calibration; and false alarms consume scarce expert attention. Preventive replacement can be wasteful when failure is not age-related, while run-to-failure may be unacceptable for high-consequence functions. The appropriate strategy depends on failure behaviour, detectability, consequence, intervention effectiveness and total life-cycle impact.

The mature organisation therefore uses a portfolio of tactics: design improvement, lubrication, inspection, functional testing, condition monitoring, preventive replacement, redundancy, operator care and controlled run-to-failure. Governance determines when each tactic is justified.

Best-practice principles

• Begin with required function and consequence, not with a technology purchase.
• Connect corporate objectives to asset strategies and executable work.
• Treat safety, environmental protection and legal compliance as design constraints.
• Use life-cycle cost and risk evidence, while recognising non-monetary limits.
• Make roles, escalation and decision rights explicit.
• Develop operators and maintainers as knowledgeable participants in the system.
• Measure outcomes, leading controls and unintended consequences together.
• Audit field reality, not only documents and dashboards.
• Validate AI and autonomous systems under changing operating conditions.
• Preserve human accountability, cyber resilience and safe manual fallback.
• Learn from weak signals, repeat failures and successful recoveries.

Chapter Summary

Asset governance aligns maintenance with organisational purpose, stakeholder value and life-cycle responsibility. ISO 55000 principles provide a management-system framework based on value, alignment, leadership, assurance, risk and systems thinking. Effective governance connects policy to asset strategy, work execution and evidence-based improvement.

Safety and environmental duties are not secondary maintenance objectives. They establish constraints and obligations that must shape criticality, work planning, competence, change management and assurance. Leadership is responsible for balancing production pressure with long-term asset health and for creating a culture in which abnormal conditions are reported and acted upon.

Autonomous maintenance can strengthen early detection when operators receive training, time, clear boundaries and feedback. Performance measurement should combine reliability, availability, work management, safety, sustainability and learning indicators. Ethical predictive maintenance requires transparency, privacy, cybersecurity, explainability and human accountability.

Self-optimising assets offer significant potential but introduce model uncertainty, data drift, cyber risk and new workforce requirements. Their value depends on decision quality, not algorithmic sophistication. Sustainable future practice will combine digital capability with life-cycle thinking, circularity, competent people and disciplined governance.

Key Terms

Asset governance: The structures and controls through which asset decisions are directed, monitored and held accountable.

Asset management: Coordinated activities that realise value from assets.

Life-cycle thinking: Considering acquisition, operation, maintenance, modification and disposal as connected decisions.

Asset-management policy: Principles and commitments governing asset management.

Strategic asset-management plan: A plan linking organisational objectives with asset-management actions.

Criticality: The relative importance of an asset or function based on consequences and risk.

Expected loss: Probability multiplied by consequence under stated assumptions.

Assurance: Evidence that objectives, controls and processes are functioning as intended.

Operational availability: The proportion of defined observation time that an asset is available for operation.

Autonomous maintenance: Operator-performed routine care, inspection and abnormality detection within defined competence boundaries.

Condition monitoring: Measurement and interpretation of asset condition to support decisions.

Predictive maintenance: Maintenance informed by evidence about future degradation or failure.

Management of change: A formal process for evaluating risks created by modifications.

Leading indicator: A measure of preventive action or control health that may precede outcomes.

Digital twin: A digital representation linked to an asset or process for analysis and decision support.

Model drift: Decline in model validity as data, equipment or operating conditions change.

Human accountability: The principle that authorised people remain responsible for consequential decisions.

Circularity: Retaining product, component and material value through reuse, repair, remanufacture or recovery.

Review Questions

Short Answer Questions

1. Distinguish asset governance, asset management and maintenance management.
2. State six principles associated with ISO 55000-aligned asset management.
3. Why should safety-critical functions be treated differently from ordinary production-critical functions?
4. What are the essential boundaries of autonomous maintenance?
5. Identify three limitations of predictive-maintenance algorithms.

Long Answer Questions

1. Explain how an organisation can trace a decision from corporate strategy to a maintenance work order.
2. Discuss the leadership behaviours required to implement a maintenance transformation successfully.
3. Evaluate the role of safety, environmental protection and management of change in maintenance governance.
4. Compare preventive, predictive, operator-led and run-to-failure strategies, including their limitations.
5. Discuss the ethical and cybersecurity implications of self-optimising asset systems.

Analytical / Application Questions

1. A repairable asset records 1,140 hours of uptime and 60 hours of downtime during a study period. Calculate operational availability using the formula in this chapter and explain two boundary conditions that must be defined before comparing it with another asset.
2. Design a governance arrangement for a plant in which production, maintenance, safety and finance disagree about replacing an ageing compressor. Specify decision rights, evidence required and safeguards against short-term production bias.
3. An AI system predicts bearing failure but generates many false alarms after a process change. Analyse possible causes and propose a validation, escalation and model-governance plan.

Multiple Choice Questions with Answers

Question 1

Which statement best describes asset governance?

A. The physical repair of failed equipment

B. The direction, control and accountability of asset-related decisions

C. The purchase of sensors for condition monitoring

D. The calculation of equipment availability

Correct Answer: B. The direction, control and accountability of asset-related decisions

Question 2

Which ISO 55000 principle requires asset decisions to be connected with organisational objectives?

A. Alignment

B. Redundancy

C. Isolation

D. Automation

Correct Answer: A. Alignment

Question 3

A function has an event probability of 0.02 and a consequence magnitude of 50,000 monetary units. Under the expected-loss model, what is the expected loss?

A. 100 monetary units

B. 1,000 monetary units

C. 2,500 monetary units

D. 1,000,000 monetary units

Correct Answer: B. 1,000 monetary units

Question 4

Which activity is most appropriate for autonomous maintenance?

A. Redesigning a pressure vessel

B. Approving a safety-case change

C. Routine cleaning, inspection and abnormality reporting

D. Overriding a safety interlock

Correct Answer: C. Routine cleaning, inspection and abnormality reporting

Question 5

If an asset has 720 hours of uptime and 30 hours of downtime, its operational availability is:

A. 90%

B. 93%

C. 96%

D. 97%

Correct Answer: C. 96%

Question 6

Why should availability not be used as the sole measure of maintenance performance?

A. Availability cannot be calculated for repairable assets

B. High availability can coexist with unsafe, environmentally harmful or unsustainable practices

C. Availability is always lower than reliability

D. Availability excludes all operating time

Correct Answer: B. High availability can coexist with unsafe, environmentally harmful or unsustainable practices

Question 7

Which is an ethical requirement for predictive-maintenance systems?

A. Removing all human review

B. Concealing uncertainty from operators

C. Maintaining human accountability for consequential decisions

D. Using personal data without defined purpose

Correct Answer: C. Maintaining human accountability for consequential decisions

Question 8

What is model drift in a self-optimising asset system?

A. A planned change of maintenance shift

B. A decline in model validity as data or operating conditions change

C. A mechanical rotation of a shaft

D. A reduction in spare-parts inventory

Correct Answer: B. A decline in model validity as data or operating conditions change

Question 9

Which control is most directly associated with hazardous-energy maintenance work?

A. Lockout/tagout

B. Marketing segmentation

C. Product redesign for aesthetics

D. Inventory obsolescence analysis

Correct Answer: A. Lockout/tagout

Question 10

Why should a predictive-maintenance project begin by defining the decision and its consequence?

A. To ensure that algorithmic novelty is prioritised

B. To determine whether the prediction will support a meaningful and proportionate action

C. To eliminate the need for data validation

D. To guarantee that every failure becomes predictable

Correct Answer: B. To determine whether the prediction will support a meaningful and proportionate action