Early Equipment Management and Lifecycle Design

Introduction

Total Productive Maintenance (TPM) is often associated with autonomous maintenance, planned maintenance and equipment effectiveness. However, the most economical opportunity to influence equipment performance occurs before the equipment is purchased. Once a machine has been specified, installed and placed in service, many costs and risks have already been designed into it. Early Equipment Management (EEM) is the TPM discipline through which knowledge from operations and maintenance is deliberately incorporated into the specification, design, procurement, installation and commissioning of new or modified equipment.

EEM extends the maintenance function upstream. Its purpose is not simply to obtain a machine that works at start-up, but to create an asset that is safe, reliable, maintainable, operable, ergonomically sound, resource-efficient and economically viable throughout its life. In production management, this is strategically important because asset design influences capacity, quality, delivery performance, labour requirements, safety exposure, energy consumption and total cost of ownership.

A useful principle is that lifecycle performance is designed, not inspected into equipment. Acceptance testing can reveal defects, but it cannot fully compensate for inaccessible components, poor diagnostic architecture, unsuitable control logic or inadequate spare-parts support.

Learning Objectives

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

1. Explain the purpose, stages and organisational requirements of Early Equipment Management.
2. Integrate reliability, availability, maintainability, operability and ergonomics into equipment specifications.
3. Apply RAMS analysis and design-for-maintenance principles.
4. Evaluate procurement and commissioning decisions using lifecycle cost rather than purchase price alone.
5. Design feedback loops that convert operational experience into improved equipment standards.
6. Critically assess digital tools, sustainability issues, governance requirements and limitations of lifecycle design.

1. The Strategic Logic of Early Equipment Management

EEM is a cross-functional process involving production, maintenance, engineering, quality, health and safety, finance, procurement, suppliers and, where relevant, customers and regulators. It prevents the common situation in which a project team selects equipment primarily on capital price and nominal output, while the operating organisation later discovers high failure rates, unsafe access, excessive changeover time or unavailable spare parts.

The lifecycle can be viewed as a sequence of decision gates:

EEM should not be treated as a one-time project meeting. It is a controlled learning system. Each new project should use evidence from existing equipment: recurring failures, emergency work, quality losses, difficult inspections, operator complaints, safety incidents, energy use and spare-parts consumption. This evidence becomes a design requirement or a supplier-selection criterion.

Design maturity and front-loaded decision-making

Design decisions made early generally have greater influence over lifecycle cost than decisions made after construction. For example, specifying modular access panels and condition-monitoring points during design is less expensive than retrofitting them after installation. The exact saving depends on the asset, but the managerial principle is robust: delay reduces freedom of choice and increases the cost of change.

EEM therefore uses formal design reviews. At each review, the team asks:

• What can fail, and what will the consequence be?
• How will the failure be detected and diagnosed?
• Can the failed item be isolated and replaced safely?
• Can routine inspection and cleaning be performed without production disruption?
• Are the required skills, tools, documentation and spares available?
• Does the design support quality, ergonomics, energy efficiency and legal compliance?

2. Translating Production Needs into Equipment Requirements

A project begins with a business need, but suppliers require testable requirements. A weak statement such as “high-speed, reliable equipment” is open to interpretation. A strong URS defines the product, demand pattern, operating conditions, quality criteria, interfaces, safety requirements, maintainability expectations and evidence required for acceptance.

Requirements should be:

• Functional: what the equipment must do, such as forming, filling, conveying or inspecting.
• Performance-based: output rate, accuracy, cycle time, yield and changeover performance.
• Reliability-based: allowable failure frequency, mission duration and critical failure modes.
• Maintainability-based: access time, diagnosis, isolation, repair and restoration requirements.
• Operability-based: controls, abnormal-operation response, cleaning and changeover.
• Human-centred: reach, force, posture, visibility, noise, heat and training demands.
• Interface-based: utilities, information systems, upstream and downstream equipment.
• Compliance-based: legislation, standards, environmental requirements and internal policies.

Requirements must be measurable. “Easy to maintain” can become “routine inspection points shall be accessible without removal of guards; the design shall permit replacement of the drive module by two trained technicians within the specified target time.” Such a requirement is still subject to verification, but it is more useful than an adjective.

A requirements traceability matrix links each requirement to its design feature, verification method, responsible party and approval status. This prevents requirements from disappearing during procurement or installation.

3. RAMS: Reliability, Availability, Maintainability and Safety

RAMS is a structured way of evaluating whether an asset can perform its intended function safely and economically. The four elements are related but not interchangeable.

Reliability

Reliability is the probability that an item performs a required function, under stated conditions, for a stated time. For a constant failure-rate model:

where R(t) is reliability at time t, λ is the constant failure rate in failures per unit time, and e is the base of natural logarithms. This model assumes statistically independent failures and a constant hazard rate; it is not appropriate for every lifecycle phase. It is most defensible over a period in which the constant-rate assumption is reasonable.

If failures follow an exponential model, mean time to failure is:

For repairable equipment, mean time between failures (MTBF) is a measure of operating time between corrective failures. MTTF and MTBF should not be treated as identical in every context: MTTF is commonly used for non-repairable items, whereas MTBF is used for repairable systems and depends on a clearly defined failure and operating-time convention.

Maintainability

Maintainability is the probability that a failed item can be restored to a specified condition within a stated time, using prescribed procedures and resources. It depends on access, modularity, diagnostic information, standardisation, tools, skills, spare parts and safety isolation.

For an exponential repair-time model:

where M(t) is the probability of restoration within time t and μ is the repair rate. The model assumes a constant repair rate and independent repair processes. If mean time to repair is MTTR, then, under this model, μ = 1 ÷ MTTR. Actual repair-time distributions are often skewed, so planners should use observed data or suitable distributions when greater accuracy is required.

Availability

For a repairable system in steady state, inherent availability is commonly expressed as:

This formula considers corrective repair time and assumes stable operating conditions. It excludes or separately treats preventive maintenance, logistics delays, administrative delays and supply constraints. Operational or achieved availability must reflect the definition adopted by the organisation. The denominator and included time categories must always be stated.

For a series system, in which every subsystem is required for operation, a simplified reliability relationship is:

where R_i(t) is the reliability of subsystem i and n is the number of subsystems. This assumes a series structure and appropriate independence assumptions. Redundancy changes the system model and must not be evaluated using the series equation.

Safety

Safety is not merely another availability variable. A highly available machine may still be unacceptable if it exposes people to uncontrolled energy, hazardous substances, excessive noise or foreseeable misuse. Safety functions require hazard identification, risk reduction, validation and legal compliance. Production targets must never justify bypassing guards, interlocks, lockout procedures or emergency systems.

4. Reliability and Maintainability Design Methods

Failure Mode and Effects Analysis

Design FMEA identifies potential failure modes, their causes, local and system effects, existing controls and recommended actions. The team should prioritise actions by severity, occurrence and detectability or, preferably, by a risk-prioritisation method consistent with organisational policy. A numerical risk-priority score can support discussion, but it should not replace engineering judgement. A low score must not conceal a catastrophic but infrequent hazard.

The analysis should include functional failures, not only broken parts. For example, “inaccurate fill volume” may result from sensor drift, software logic, pressure variation, contamination or calibration error. The resulting actions may include redundancy, diagnostic checks, environmental protection, calibration access and mistake-proofing.

Fault-tree and event-tree analysis

Fault-tree analysis starts with an undesired top event and reasons downward through combinations of causes using logical AND and OR relationships. Event-tree analysis starts with an initiating event and examines possible consequences after protective functions succeed or fail. These methods are particularly valuable for safety-critical systems, major loss scenarios and common-cause failures.

Reliability-centred maintenance considerations

Although Reliability-Centred Maintenance (RCM) is usually applied to operating assets, its questions are valuable during design:

1. What are the functions and required performance standards?
2. How can the asset fail to perform those functions?
3. What causes each functional failure?
4. What happens when each failure occurs?
5. Why does the failure matter?
6. What proactive task or redesign is technically and economically justified?
7. What should be done if no suitable preventive task exists?

The answers influence sensor provision, access, redundancy, protective devices, spare modules and maintenance intervals.

Design for maintainability

Design-for-maintenance principles include:

• safe and direct access to inspection and service points;
• modular replacement rather than complex on-machine repair where justified;
• standard fasteners, components, connectors and lubrication systems;
• visible status indication and unambiguous fault codes;
• isolation points that support lockout and verification;
• lifting points and handling provisions for heavy components;
• separation of clean and contaminated maintenance work;
• drainage, cleaning and inspection features;
• physical and digital documentation available at the point of work;
• condition-monitoring provisions that are useful, accessible and maintainable.

Designers should distinguish technical possibility from maintainability in practice. A component may be physically replaceable but still have poor maintainability if diagnosis is uncertain, access requires prolonged production shutdown, or the replacement module is unavailable.

5. Operability, Ergonomics and Human Performance

Operability concerns how effectively people can run, adjust, clean, change over and respond to abnormal conditions. It is shaped by control-system design, information presentation, standard work, training and the physical arrangement of the workplace.

Ergonomic design should consider anthropometric variation, posture, reach distance, manual force, repetition, vibration, temperature, illumination, noise and cognitive workload. A design that reduces a nominal repair time but requires awkward overhead work may increase injuries, errors and long-term absence.

Human factors should be examined during normal, start-up, shutdown, cleaning, jam-clearing, fault recovery and emergency conditions. Controls should have consistent meanings, clear status indication and protection against foreseeable incorrect actions. Alarm systems should prioritise actionable information rather than generate excessive alarms that operators learn to ignore.

Changeover is a major EEM concern in production environments. Equipment should support external preparation, quick-release tooling, recipe control, visual confirmation and error-proof connections. A reduction in changeover duration increases available capacity without necessarily increasing machine speed.

6. Procurement, Supplier Evaluation and Acceptance

Procurement should evaluate value over the asset life, not only acquisition price. Tender documents should require evidence concerning reference installations, failure history, service response, spare-parts availability, cybersecurity, software ownership, training, documentation and change control.

Supplier evaluation can use weighted criteria, but weights should reflect strategic risk. For a bottleneck or safety-critical asset, maintainability and support may appropriately outweigh a small difference in purchase price. Commercial contracts should define measurable acceptance criteria and responsibilities for defects discovered during ramp-up.

Factory and site acceptance

Factory Acceptance Testing (FAT) verifies equipment at the supplier’s site before shipment. Site Acceptance Testing (SAT) verifies installation, integration and performance under the customer’s conditions. Tests may cover:

• functional sequences and abnormal conditions;
• output, cycle time and quality capability;
• safety circuits and fail-safe behaviour;
• alarm and diagnostic functions;
• changeover and cleaning procedures;
• maintainability demonstrations;
• data collection and interface performance;
• documentation, training and spare-parts completeness.

A machine should not be declared operational merely because it runs for a short demonstration. Acceptance should include a defined performance run, representative materials, realistic staffing and documented treatment of open items.

7. Commissioning, Ramp-Up and Lifecycle Learning

Commissioning is a controlled transition from construction to stable operation. A sound process includes installation verification, dry runs, wet or product trials, safety validation, capability assessment, operator training and progressive release of responsibility.

The project team should establish a baseline containing:

• intended operating modes and standard speeds;
• quality parameters and critical process settings;
• failure codes and initial failure distributions;
• planned maintenance tasks and intervals;
• energy and utility consumption;
• changeover, cleaning and minor-stop performance;
• training and competency status.

Early failures may arise from design defects, installation errors, immature procedures or inadequate training. They should be analysed without prematurely blaming operators. A useful escalation route is: contain the risk, verify the failure, identify the physical and organisational causes, correct the design or standard, and update the knowledge repository.

Lessons learned should be codified into standard specifications, approved component lists, design checklists and supplier requirements. Otherwise, organisations repeatedly purchase similar problems.

8. Lifecycle Costing and Economic Evaluation

Lifecycle costing estimates all significant costs attributable to an asset over a defined study period. A general present-value model is:

where LCC is lifecycle cost, C₀ is initial acquisition and installation cost, C_t is cost incurred in period t, N is the study horizon, r is the discount rate per period, and S_N is residual or salvage value at the end. Costs must be expressed in consistent currency and price basis. The model requires assumptions about inflation, energy prices, utilisation, maintenance, downtime, training, disposal and residual value.

A more complete analysis separates cost categories: capital, installation, commissioning, labour, energy, consumables, planned maintenance, corrective maintenance, quality losses, production loss, compliance, software support and decommissioning. Downtime cost should be estimated carefully. It may represent lost contribution margin rather than lost sales revenue, and it may be partly recoverable through overtime, inventory or alternative capacity.

Worked example: availability and lifecycle choice

Two alternative machines have the following steady-state data:

Using the simplified inherent-availability equation:

Machine A:

A_i = 600 ÷ (600 + 6) = 0.9901, or approximately 99.01%.

Machine B:

A_i = 900 ÷ (900 + 12) = 0.9868, or approximately 98.68%.

Machine B has the higher MTBF but lower inherent availability because its repair time is twice that of Machine A. This illustrates why reliability alone is insufficient.

Suppose the organisation values each unavailable production hour at $180 and uses the simplified unavailable fraction 1 − A_i. Estimated annual unavailability cost is:

• Machine A: 4,800 × (1 − 0.9901) × 180 ≈ $8,554.
• Machine B: 4,800 × (1 − 0.9868) × 180 ≈ $11,405.

The approximate annual operating cost becomes $60,554 for A and $54,405 for B. The annual operating advantage of B is therefore about $6,149, while its additional initial cost is $50,000. A simple payback estimate is approximately 8.1 years, before discounting and before considering quality, capacity, risk and residual value.

This result does not prove that A is preferable. It demonstrates the need to include maintainability, production consequences and uncertainty. If B’s modular design reduces future overhaul cost or supports expansion, its economic position may improve. Conversely, if downtime cost is much higher than assumed, the faster repairability of A may dominate. Decisions should use sensitivity analysis rather than a single forecast.

9. Applied Case: Illustrative Packaging-Line Replacement

Background

A food manufacturer replaced an ageing packaging line whose nominal speed exceeded current demand. In practice, the line experienced frequent minor stops, lengthy cleaning, difficult film-threading and repeated sensor contamination. Maintenance personnel were often called to resolve problems that operators could not diagnose.

Problem

The initial project proposal selected a replacement primarily on rated speed and purchase price. The maintenance manager identified risks: inaccessible sensors, proprietary drives with long lead times, limited diagnostic information and no clear method for validating cleanability.

Analysis

A cross-functional EEM team reviewed two years of failure and loss data. Pareto analysis showed that a small number of recurring issues—sensor contamination, film misalignment, changeover adjustment and drive overheating—accounted for most lost time. An FMEA linked these issues to design features rather than isolated component failures. Operators also reported awkward reaches and poor visibility during threading.

The team converted findings into requirements: tool-less access to sensor covers, positive film-guiding features, remote diagnostic codes, standard drives, external changeover preparation, safe access platforms, washdown-compatible components and a demonstrated cleaning procedure. The supplier had to perform a maintainability demonstration during FAT and provide critical spares before SAT.

Approach and solution

The project used design reviews at concept, preliminary design, FAT and SAT stages. A digital equipment model was used to check access envelopes and maintenance clearance. A simulated fault library tested diagnostic messages. Operators participated in trials rather than receiving the equipment only after technical decisions had been completed.

During commissioning, the team separated installation defects from design weaknesses and recorded both in the lessons-learned system. The supplier corrected alarm descriptions, relocated two inspection points and supplied a standardised drive module.

Outcome

In this illustrative case, the line reached its planned operating rate more quickly than the previous project baseline. Changeover work became more predictable, first-line operators resolved a larger share of minor stops, and maintenance tasks could be completed with fewer confined or awkward postures. The most important outcome was not a single availability number; it was the institutionalisation of a repeatable specification and acceptance process.

Learning points

1. Nominal speed is not equivalent to effective production capacity.
2. Operator and maintainer knowledge is a design input, not merely post-installation feedback.
3. Maintainability must be demonstrated, not promised.
4. Standard components and diagnostic clarity reduce dependence on scarce expertise.
5. Lessons learned have value only when embedded in future standards and procurement controls.

10. Digital Technology, AI and Contemporary Developments

Modern EEM increasingly uses digital twins, 3D access simulation, sensor-ready architectures, cloud or edge analytics and integrated asset information models. These tools can identify collision risks, estimate access time, support virtual commissioning and improve handover data.

AI and machine-learning methods may assist anomaly detection, image-based inspection, predictive maintenance and failure-mode discovery. They should not be treated as substitutes for engineering analysis. Models require representative data, stable measurement definitions and controls against false alarms, bias and model drift. A prediction that cannot be acted upon safely and economically has limited maintenance value.

Connected equipment also creates cybersecurity and governance obligations. Access rights, software updates, network segmentation, backup procedures, supplier responsibilities and incident response should be specified before commissioning. Data ownership and retention must be contractually clear. Digital dependency can create a new single point of failure if equipment cannot operate safely when connectivity or cloud services are unavailable.

Sustainability is another lifecycle concern. Design decisions affect energy, water, consumables, emissions, repairability, remanufacturing, material recovery and hazardous waste. A lower purchase price may conflict with long-term environmental performance. Lifecycle assessment should be proportionate to the decision and should avoid unsupported claims based only on the use phase.

11. Challenges and Limitations

EEM faces several practical difficulties:

• project schedules may reward installation dates rather than stable performance;
• capital budgets may be separated from operating and downtime costs;
• suppliers may resist sharing reliability evidence;
• future operating conditions may be uncertain;
• reliability data may be sparse or inconsistently defined;
• teams may overuse numerical risk scores;
• new technology may introduce vendor lock-in and cybersecurity exposure;
• standardisation may suppress legitimate innovation;
• human factors may be considered too late;
• lessons learned may remain informal and inaccessible.

These limitations require governance, not abandonment of the method. Assumptions should be documented, uncertainty should be tested through sensitivity analysis, and acceptance criteria should be agreed before commercial pressure intensifies. Independent review is appropriate for high-consequence assets.

Best-Practice Framework

A robust EEM system normally includes:

1. A cross-functional governance team with defined decision rights.
2. A controlled URS and requirements traceability matrix.
3. Historical loss and failure data used as design evidence.
4. RAMS, FMEA and safety analysis proportionate to risk.
5. Explicit ergonomic, cleanability, changeover and access requirements.
6. Lifecycle-cost evaluation with sensitivity analysis.
7. FAT and SAT criteria that test real operating and maintenance conditions.
8. Complete documentation, training, spares and digital records at handover.
9. A controlled ramp-up plan with baseline performance measures.
10. A formal lessons-learned loop that updates future standards.

Chapter Summary

Early Equipment Management moves TPM upstream by embedding production, maintenance, reliability, safety and human factors knowledge into equipment decisions before purchase and installation. Its lifecycle approach begins with a measurable business need and continues through specification, design, procurement, acceptance, commissioning, operation and renewal.

RAMS provides a disciplined vocabulary for reliability, maintainability, availability and safety, but each measure requires a clear definition and appropriate assumptions. Reliability and availability are not interchangeable; high MTBF can be offset by poor repairability. Design-for-maintenance principles address access, diagnosis, modularity, isolation, tools, spares and information. Operability and ergonomics protect both performance and human wellbeing.

Lifecycle costing exposes the economic consequences of energy, maintenance, downtime, quality loss, support and disposal. FAT, SAT and ramp-up controls verify that the delivered asset meets more than a nominal speed claim. Digital tools and AI can strengthen design and diagnosis, but they introduce data, cybersecurity, governance and model-validity risks. The strategic success of EEM is achieved when experience from current assets becomes a formal requirement for future assets.

Key Terms

Early Equipment Management: A TPM process for incorporating operational and maintenance learning into new or modified equipment.

Lifecycle design: Designing an asset with consideration of performance, cost, safety and environmental effects over its entire life.

RAMS: Reliability, Availability, Maintainability and Safety analysis.

Reliability: Probability that an item performs a required function for a stated time and condition.

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

Availability: Proportion or probability of time that equipment is capable of performing its function under a stated definition.

MTBF: Mean time between failures for a repairable item.

MTTR: Mean time to repair or restore a failed item, according to the organisation’s definition.

FMEA: Failure Mode and Effects Analysis.

Fault-tree analysis: Top-down logical analysis of combinations of causes leading to an undesired event.

User Requirement Specification: Controlled document defining what the equipment must do and how requirements will be verified.

Requirements traceability: Linkage between requirements, design features, tests, responsibilities and approvals.

FAT: Factory Acceptance Test conducted at the supplier’s facility.

SAT: Site Acceptance Test conducted after installation at the user’s site.

Lifecycle cost: Present value of acquisition, operation, maintenance, downtime, disposal and related costs.

Design for maintainability: Designing equipment to simplify safe inspection, diagnosis, repair and restoration.

Operability: Ease and effectiveness with which people can operate, adjust, clean and recover equipment.

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

Lessons learned: Explicit knowledge captured from experience and applied to future decisions.

Review Questions

Short Answer Questions

1. What is the primary purpose of Early Equipment Management?
2. Distinguish reliability from maintainability.
3. Why should a User Requirement Specification contain measurable requirements?
4. What is the difference between FAT and SAT?
5. State two ergonomic factors that should be considered in equipment design.

Long Answer Questions

1. Explain how RAMS analysis can influence equipment specification, procurement and commissioning.
2. Discuss the relationship between MTBF, MTTR and availability, including the limitations of the simplified availability equation.
3. Explain how lifecycle costing can change a decision based solely on purchase price.
4. Describe a design-for-maintenance process for a new production machine.
5. Evaluate the role of operators, maintainers, suppliers and management in lifecycle learning.

Analytical / Application Questions

1. A machine has an MTBF of 750 hours and an MTTR of 10 hours. Calculate simplified inherent availability and explain which assumptions must be stated before using the result.
2. A supplier offers a cheaper machine with proprietary components and limited diagnostic information, while a more expensive alternative uses standard modules and has better access. Develop a decision framework using RAMS, lifecycle cost, risk and maintainability criteria.
3. Design an EEM gate-review plan for a safety-critical automated cell, identifying the evidence required at concept, design, FAT, SAT and handover stages.

Multiple Choice Questions with Answers

Question 1

Which statement best describes Early Equipment Management?

A. A process that begins only after repeated equipment failures

B. A process that embeds operational and maintenance learning into equipment lifecycle decisions

C. A method for increasing machine speed without changing design

D. A financial system for approving maintenance invoices

Correct Answer: B. A process that embeds operational and maintenance learning into equipment lifecycle decisions

Question 2

Under the constant failure-rate model, which expression represents reliability at time t?

A. R(t) = 1 − e^(−λt)

B. R(t) = λt

C. R(t) = e^(−λt)

D. R(t) = MTTR ÷ MTBF

Correct Answer: C. R(t) = e^(−λt)

Question 3

A repairable machine has MTBF = 900 h and MTTR = 12 h. Its simplified inherent availability is closest to:

A. 0.879

B. 0.987

C. 1.013

D. 0.133

Correct Answer: B. 0.987

Question 4

Which item is normally included in a measurable maintainability requirement?

A. The machine should be modern

B. The machine should be attractive

C. A drive module must be replaceable by two trained technicians within a specified time

D. The supplier should provide a competitive price

Correct Answer: C. A drive module must be replaceable by two trained technicians within a specified time

Question 5

What is the principal purpose of a requirements traceability matrix?

A. To record employee attendance

B. To link requirements to design features, verification and approval

C. To calculate depreciation only

D. To replace all equipment testing

Correct Answer: B. To link requirements to design features, verification and approval

Question 6

Which statement correctly distinguishes FAT from SAT?

A. FAT occurs at the supplier site; SAT occurs after installation at the user site

B. FAT is a financial audit; SAT is a safety audit

C. FAT occurs after disposal; SAT occurs before procurement

D. FAT and SAT are identical activities with different names

Correct Answer: A. FAT occurs at the supplier site; SAT occurs after installation at the user site

Question 7

Why can a machine with higher MTBF still have lower availability than another machine?

A. Availability never depends on repair time

B. Higher MTBF always causes lower reliability

C. Its MTTR may be sufficiently longer to offset the reliability advantage

D. Availability is determined only by purchase price

Correct Answer: C. Its MTTR may be sufficiently longer to offset the reliability advantage

Question 8

Which activity best supports lifecycle learning?

A. Discarding commissioning records after handover

B. Recording recurring failures and converting them into future design standards

C. Allowing every project to use unrelated specifications

D. Measuring only nominal machine speed

Correct Answer: B. Recording recurring failures and converting them into future design standards

Question 9

In lifecycle costing, future costs are commonly converted to present value using:

A. A discount rate

B. The machine’s colour

C. The number of operators only

D. The nominal speed rating

Correct Answer: A. A discount rate

Question 10

Which is an appropriate governance concern for connected equipment using AI-based diagnostics?

A. Eliminating all human review of predictions

B. Ignoring cybersecurity because the equipment is physical

C. Defining data ownership, access, model validity and safe response to false alarms

D. Using cloud connectivity even when safe local operation is impossible

Correct Answer: C. Defining data ownership, access, model validity and safe response to false alarms