Sustainable, Ethical and Resilient Lean Futures

Introduction

Lean manufacturing and lean service operations were developed to improve customer value while reducing waste, delay and unnecessary resource consumption. Contemporary operations, however, cannot define performance only through cost, throughput and quality. Climate change, resource scarcity, geopolitical disruption, labour-rights concerns, cyber risks and rapid automation have expanded the meaning of operational excellence. A process may be efficient yet environmentally damaging, socially unjust or dangerously fragile.

Sustainable, ethical and resilient lean operations therefore seek simultaneous improvement in five outcomes: customer value, operational performance, ecological integrity, human well-being and the capacity to withstand and recover from disruption. This chapter integrates these dimensions and examines how lean principles must evolve from narrow waste elimination toward responsible value creation.

Learning objectives

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

1. Explain how lean methods can support environmental sustainability and circular operations.
2. Evaluate the relationship between lean work design, employee well-being and ethical management.
3. Analyse resilience, responsible sourcing and disruption-response capabilities.
4. Identify ethical risks associated with work intensification, surveillance, algorithmic bias and automation.
5. Assess human-centric, adaptive and AI-enabled operating models.
6. Apply these ideas in a comprehensive lean transformation case.

1. From lean efficiency to responsible value

Traditional lean thinking defines waste as activity that consumes resources without creating value for the customer. Common categories include overproduction, waiting, transport, over-processing, inventory, motion, defects and unused human capability. This logic remains valuable, but it requires three extensions.

First, environmental externalities must be included. Carbon emissions, toxic discharge, water depletion and biodiversity damage may not appear on a factory’s income statement, but they are genuine consequences of operations. Second, social outcomes must be treated as operational outcomes. Excessive overtime, unsafe work, insecure employment and denial of employee voice can produce short-term output while weakening long-term capability. Third, resilience must be considered: the lowest-cost process is not necessarily the best process if a single failure causes prolonged interruption.

A useful broader definition is:

Responsible lean is the disciplined design and improvement of value streams to reduce operational waste while protecting people, ecosystems, rights and the continuity of essential service.

This definition changes the improvement question from “How can output be increased with fewer resources?” to “How can desirable value be created with fewer harmful inputs, fairer work and greater adaptive capacity?”

The triple-bottom-line perspective

The triple-bottom-line approach evaluates performance through economic, environmental and social dimensions. In operations, these dimensions should not be treated as independent projects. For example, reducing packaging may lower cost and emissions, but only if product damage does not increase. Automation may improve safety, but only if workers are retrained and not exposed to arbitrary displacement. Supplier consolidation may simplify control, but it may also increase vulnerability to disruption.

A balanced operational scorecard may include:

Lean must also distinguish between operational waste and necessary protective capacity. Some inventory, redundancy, inspection and spare capacity may appear inefficient in stable conditions but provide valuable resilience. The appropriate objective is not minimum resources at all times; it is an economically and ethically justified level of resources for the risk environment.

2. Environmental sustainability and lean operations

Lean and eco-efficiency

Eco-efficiency means creating more customer value with fewer environmental impacts. Lean and eco-efficiency overlap because waiting, defects, rework, excess transport and overproduction often consume energy and materials. A value-stream map can therefore be expanded to show electricity, fuel, water, material loss, packaging and emissions at each process step.

For a process with energy consumption E and output Q, energy intensity is:

where I_E is energy intensity, E is energy consumed in a stated period (for example, kWh), and Q is good output in the same period (for example, units). The measure is meaningful only when output quality and product mix are comparable. If products have different complexity, energy should be normalised using an agreed production-equivalent measure.

Similarly, material yield can be expressed as:

where M_good is material incorporated into conforming products and M_input is total material issued. Scrap reused internally should be reported transparently rather than automatically treated as elimination of environmental impact, because remelting or reprocessing also consumes energy.

Carbon accounting requires clear boundaries. Scope 1 emissions arise directly from owned or controlled sources; Scope 2 emissions arise from purchased electricity, heat or steam; Scope 3 emissions arise elsewhere in the value chain, including purchased materials, logistics, product use and end-of-life treatment. A lean project that reduces factory electricity but increases outsourced transport may improve one measure while worsening total impact.

Green value-stream mapping

Green value-stream mapping adds environmental information to the conventional map. It identifies:

• energy and water use by process;
• material inputs, scrap and hazardous substances;
• transport distance, mode and load utilisation;
• emissions and waste-treatment routes;
• environmental compliance risks;
• opportunities for prevention, reuse, remanufacture or substitution.

The method should follow the same discipline as ordinary lean analysis: define customer demand, establish process boundaries, measure current conditions and distinguish facts from assumptions. Environmental improvements should be tested for unintended consequences. For example, smaller delivery batches may reduce warehouse inventory but increase vehicle kilometres and packaging per unit.

Circular operations

A linear operating model follows “extract, make, use and dispose.” Circular operations aim to maintain products, components and materials at their highest practical value for as long as possible. Important strategies include repair, maintenance, refurbishment, remanufacturing, reuse, product-life extension, recycling and industrial symbiosis.

Circularity changes the design of the value stream. Forward flows deliver products to customers, while reverse flows collect returns, used products, packaging and recoverable components. These flows create new forms of uncertainty: returned items differ in condition, quantity and timing. Consequently, circular lean systems need inspection standards, modular product design, traceability and flexible disassembly processes.

A simple material circularity indicator can be used for internal comparison, but no single indicator captures all circular impacts. Managers should report the proportion of recycled or recovered input, the proportion of product value retained, product life extension and the quality of recovered material. Recycling a complex product into low-value material may be less circular than repairing or remanufacturing it.

Illustrative application. A medical-equipment producer maps a disposable packaging process and finds that 8% of packaging is damaged during internal handling. A redesign using reusable transport containers eliminates damage and reduces packaging purchases. However, the containers require reverse logistics and cleaning. The project is successful only if the additional cleaning energy, water and transport are lower than the environmental burden of single-use packaging. The correct lean decision is based on the complete life cycle, not on the local process alone.

3. Employee well-being and the ethics of lean work

Lean relies on employee knowledge, problem solving and standardised work. It becomes ethically problematic when “respect for people” is reduced to a slogan while targets are raised continuously. Work intensification occurs when employees must produce more quickly, with fewer pauses, less staffing or greater monitoring, without corresponding redesign, support or reward.

Potential consequences include fatigue, musculoskeletal injury, stress, errors, absenteeism and turnover. These outcomes are not merely human-resources concerns; they increase defects, instability and hidden cost. A sustainable lean system therefore treats safe pace, recovery time, ergonomic design and psychological safety as process requirements.

Human-centred work design

Human-centred lean work design includes:

1. Participatory improvement: employees help identify waste and design countermeasures.
2. Ergonomic analysis: tools, reach distances, force, posture and repetition are assessed.
3. Reasonable standard work: standards describe a safe and repeatable method, not an impossible maximum speed.
4. Protected problem escalation: workers can stop or escalate unsafe conditions without retaliation.
5. Skill development: cross-training increases capability without using flexibility to create insecurity.
6. Fair recognition: improvement ideas and additional responsibilities receive appropriate acknowledgement.
7. Psychological safety: employees can report mistakes, near misses and ethical concerns.

A takt-based system illustrates the need for care. If available production time is T and customer demand is D, takt time is:

T must exclude legitimately unavailable time such as planned breaks and scheduled maintenance, and D must represent the required quantity for the same period. Takt is a demand signal, not automatically a permissible individual work rate. It does not account by itself for fatigue, task variation, learning, ergonomic limits or unavoidable delays. Treating takt as a performance quota can create unsafe intensification.

Measuring well-being

Useful indicators include lost-time injury frequency, near misses, ergonomic risk scores, absenteeism, turnover, overtime, schedule predictability, training hours, employee voice and psychological-safety survey results. These should be interpreted together. A low injury rate may reflect under-reporting rather than safety. A high improvement count may reflect genuine engagement or coercive idea quotas.

4. Resilient lean operations

Resilience is the ability of an operating system to anticipate, absorb, adapt to and recover from disruption while maintaining essential functions. Disruptions may arise from natural hazards, pandemics, cyberattacks, equipment failure, labour disputes, supplier insolvency, geopolitical restrictions or sudden demand changes.

Lean systems can support resilience through visual control, short feedback loops, standard work, rapid problem solving and reduced complexity. Yet extreme just-in-time dependence, single sourcing and minimal buffers can amplify vulnerability. Resilience is therefore a design trade-off rather than the rejection of lean.

Four resilience capabilities

• Anticipation: horizon scanning, risk assessment, supplier monitoring and scenario planning.
• Absorption: buffers, cross-trained staff, alternative routes and protective capacity.
• Adaptation: flexible equipment, modular products, substitution rules and rapid reconfiguration.
• Recovery and learning: continuity plans, restoration priorities, root-cause analysis and institutional learning.

A useful risk priority model evaluates likelihood, consequence and detectability, but numerical scores should not create false precision. A qualitative risk register should accompany any scoring system. Critical risks deserve attention even when their probability is low if consequences are catastrophic.

For a disruption, recovery time objective (RTO) is the maximum acceptable time to restore a process or service, while recovery point objective (RPO) is the maximum acceptable loss of data or transactions measured in time. These concepts are especially relevant to digital service operations.

Resilience decisions should consider total cost. Holding additional inventory has a carrying cost, while stockout has a disruption cost. A simplified decision compares the expected annual disruption loss with the annual cost of protective capacity. This is not an exact optimisation unless probabilities, costs and dependencies are reliable. It is a structured managerial aid.

Responsible sourcing

Responsible sourcing integrates cost, quality and delivery with labour rights, environmental performance, anti-corruption and supply continuity. It requires more than a supplier questionnaire. Effective practice includes:

• a supplier code aligned with recognised labour and environmental principles;
• risk-based due diligence rather than identical audits for all suppliers;
• traceability for critical materials;
• worker grievance channels independent of management;
• corrective-action support and realistic remediation periods;
• purchasing practices that do not force unsafe overtime or impossible lead times;
• contingency qualification of alternative suppliers;
• contractual requirements for data security and ethical conduct.

Audits can fail when suppliers coach workers, hide subcontracting or treat compliance as a temporary performance. Buyer behaviour matters: severe price pressure, unstable forecasts and late engineering changes can transfer risk down the chain. Responsible lean procurement seeks stable collaboration and transparency rather than merely shifting waste and risk to weaker parties.

5. Ethical risks in digital and automated lean systems

Surveillance and data governance

Digital lean systems may collect scan times, location data, machine interactions, productivity rates, video records and biometric information. Such data can improve safety and flow, but continuous surveillance can undermine dignity and trust. Ethical governance requires a legitimate purpose, data minimisation, transparency, access controls, limited retention and a process for challenging inaccurate interpretations.

Managers should distinguish process information from personal judgement. A delay may be caused by a defective material, unclear instruction or unsafe workstation, not low effort. Individual rankings based on incomplete data can reward speed over quality and encourage concealment of problems.

Algorithmic bias

AI systems may allocate work, predict absenteeism, screen applicants, assess performance or prioritise maintenance. Bias can enter through historical data, proxy variables, unrepresentative samples or objectives that reward the wrong behaviour. For example, a scheduling algorithm trained on historical overtime may infer that employees with caring responsibilities are less available and allocate them fewer development opportunities.

Governance should include:

1. documented purpose and decision boundaries;
2. data-quality and representativeness checks;
3. fairness testing across relevant groups;
4. explainable outputs for affected employees;
5. human review and appeal;
6. monitoring after deployment;
7. prohibition of uses that violate law, rights or safety.

A fairness metric is not universally correct. Statistical parity, equal error rates and individual fairness may conflict, particularly when base rates differ. Therefore, technical testing must be combined with legal, ethical and employee-representative review.

Automation and employment

Automation can remove dangerous, monotonous or physically demanding work, but it can also displace workers, deskill jobs or concentrate control. A responsible automation decision assesses the whole work system: task content, safety, learning, job quality, transition support and distribution of benefits.

A human-centric principle is “automation with augmentation”: automate tasks where machines are more reliable or safer, while retaining human judgement for ambiguous, relational, ethical or exception-rich work. Before implementation, organisations should conduct a skills-impact assessment, consult affected employees, provide retraining and establish fair redeployment or compensation arrangements.

6. Human-centric, adaptive and AI-enabled operating models

Human-centric operations

Human-centric operations place human dignity, capability and meaningful participation at the centre of process design. This does not mean rejecting productivity technology. It means optimising technology around human flourishing and system reliability rather than treating people as variable machine capacity. Practices include team-based problem solving, worker representation in design, skill matrices, job rotation where appropriate and transparent performance dialogue.

Adaptive operations

Adaptive operations combine stable standards with controlled flexibility. Standard work defines the current best-known method, while an improvement system permits revision when evidence changes. Adaptive systems use modular capacity, cross-training, postponement, scenario planning and rapid experimentation. They avoid two extremes: rigid bureaucracy and uncontrolled improvisation.

AI-enabled operations

AI can support demand forecasting, anomaly detection, predictive maintenance, computer vision, route optimisation, knowledge retrieval and decision support. Its value depends on data quality, process stability and human judgement. AI does not repair a badly defined process automatically; it may accelerate incorrect decisions.

A responsible AI deployment follows a lifecycle:

1. define the operational problem and affected stakeholders;
2. establish data ownership, quality and lawful use;
3. test the model against operational and fairness criteria;
4. run a limited pilot with human override;
5. monitor accuracy, drift, safety and unintended effects;
6. audit decisions and revise or withdraw the system when necessary.

The relevant question is not “Can AI optimise this?” but “What decision should be made, using what evidence, with what accountability and appeal?”

7. Capstone case: Meridian Medical Devices

The following is an illustrative fictional case.

Background

Meridian Medical Devices manufactures infusion pumps in three countries and supplies hospitals through regional distribution centres. Its lean programme has reduced assembly lead time by 35% and work-in-process by 40%. The company uses takt-based line balancing, supplier delivery windows and digital dashboards.

Problem

Several problems now threaten the sustainability of the transformation. Employees report increased wrist pain and pressure to avoid reporting minor stoppages. A low-cost electronic component is sourced from a single supplier in a politically unstable region. The supplier passes social audits but has unexplained subcontracting. Meridian’s packaging is largely single-use, and returned pumps are discarded because inspection and refurbishment are not standardised. A proposed AI system would rank operators by cycle-time adherence using scanner and video data.

A severe regional disruption then closes the component supplier for six weeks. At the same time, a dashboard incorrectly identifies a group of experienced employees as low performers because they conduct additional quality checks not captured by the system.

Analysis

The transformation has improved local flow but optimised too narrowly. Inventory reduction increased exposure to a single-source failure. The productivity dashboard created measurement bias by treating recorded cycle time as the full definition of performance. Work intensification reduced the reliability of problem reporting. Disposal of returned pumps created environmental waste and lost recovery value. The audit programme focused on supplier documentation rather than actual labour conditions and subcontracting visibility.

The leadership team uses a five-dimension review:

Approach and solution

Meridian forms a cross-functional transformation council including operators, engineers, procurement, quality, sustainability, information security and employee representatives. It adopts the following sequence:

1. Stabilise safety and voice. The company pauses individual rankings, protects stop-work escalation and conducts ergonomic assessments. Standard work is rewritten to include safe posture, planned recovery and quality checks.
2. Redesign resilience. Critical components are classified by supply risk. A second source is qualified, and a modest strategic buffer is established. The buffer is justified as protection against high-consequence disruption, not as uncontrolled inventory.
3. Strengthen sourcing. Contracts require subcontractor disclosure, worker grievance access, corrective action and responsible forecasting. Procurement performance is measured partly by supplier stability and ethical compliance.
4. Develop circular flow. Returned pumps are triaged into reuse, repair, remanufacture and responsible recycling. Product designers introduce modular components and serial-level traceability.
5. Govern AI. The company replaces operator ranking with a team-level decision-support tool that flags unusual process conditions. Data are used for coaching and system improvement, not automatic discipline. Employees can inspect and challenge records.
6. Use a balanced scorecard. Monthly reviews include first-pass yield, lead time, energy intensity, packaging recovery, ergonomic risk, near misses, supplier-risk closure, recovery readiness and employee trust.

Outcome

After twelve months, Meridian restores supply continuity during a second regional disruption because the alternate source and buffer are operational. Packaging waste falls, and a proportion of returned pumps is safely refurbished. Reported near misses initially rise because employees trust the reporting process; later, injury risk declines. Productivity does not increase as rapidly as in the original programme, but quality, retention and delivery reliability improve. The AI tool detects equipment anomalies without producing individual disciplinary scores.

The case demonstrates that responsible lean may reject a locally efficient action when it damages safety, rights or resilience. It also shows that employee participation is not an obstacle to advanced operations; it is a source of valid data, ethical legitimacy and better problem solving.

8. Implementation framework and limitations

A practical implementation sequence is:

Diagnose

Map the value stream, environmental flows, work conditions, stakeholder impacts and disruption dependencies. Establish baselines and data boundaries.

Prioritise

Rank opportunities according to customer value, risk, environmental impact, human consequence and feasibility. Do not allow financial return alone to determine priority where safety or rights are involved.

Design

Create countermeasures with affected employees, suppliers and relevant specialists. Test trade-offs across the complete life cycle and value network.

Pilot

Use controlled experiments, clear safeguards and defined stop criteria. Compare results with a baseline and examine distributional effects across groups.

Govern and scale

Assign accountability, document decisions, audit data and suppliers, train managers, and provide grievance and appeal mechanisms. Review whether local gains create harm elsewhere.

Important limitations must be recognised. Sustainability metrics can be difficult to compare because boundaries and data quality vary. Carbon reductions may be overstated through offsets or favourable accounting assumptions. Resilience investments can appear wasteful until disruption occurs. Employee surveys can be distorted by fear. AI fairness cannot be reduced to one numerical test. Finally, no management framework substitutes for law, collective representation, professional ethics or democratic accountability.

Contemporary developments and best practices

Contemporary lean futures are shaped by decarbonisation, digital traceability, product-as-a-service models, industrial symbiosis, distributed manufacturing, cyber-physical systems and generative AI. These developments create opportunities but also new dependencies. A connected factory may reduce waste while increasing cyber vulnerability; product-as-a-service may encourage durable design but intensify data collection; generative AI may improve standard-work documentation but reproduce unsafe or biased instructions.

Best practice is therefore characterised by:

• lifecycle rather than local optimisation;
• measurable but not manipulated performance indicators;
• transparent assumptions and data provenance;
• employee participation and collective voice;
• risk-based redundancy for critical processes;
• supplier partnership combined with independent accountability;
• human oversight of consequential automation;
• continuous learning after near misses and disruptions;
• explicit protection of safety, dignity and rights.

Chapter Summary

Sustainable, ethical and resilient lean operations extend conventional waste reduction to include ecological limits, human well-being, responsible sourcing and continuity under disruption. Green value-stream mapping connects process improvement with energy, materials, emissions and waste. Circular operations add reverse flows and seek to preserve product and material value through repair, reuse and remanufacture.

Lean work must be human-centred. Takt and standard work are useful planning concepts but must not become instruments of unsafe work intensification. Psychological safety, ergonomics, fair recognition and employee participation are essential to reliable improvement. Resilience requires anticipation, absorption, adaptation and recovery; carefully justified buffers and alternate capacity can be compatible with lean.

Digitalisation and AI create benefits in prediction, anomaly detection and decision support, but they introduce surveillance, bias, privacy and accountability risks. Responsible systems use limited data, explainable decisions, human review and appeal. The Meridian case shows that an integrated transformation can improve sustainability and resilience while protecting people, provided that performance is evaluated across the entire value system rather than through narrow local efficiency.

Key Terms

Responsible lean: Lean improvement that protects customer value, people, ecosystems, rights and continuity.

Eco-efficiency: Creating value with reduced environmental impact.

Green value-stream mapping: Mapping process flow together with energy, material, water, emissions and waste information.

Circular operations: Operations that retain products, components and materials at high value through reuse, repair, refurbishment, remanufacture and recycling.

Material yield: The proportion of input material incorporated into conforming output.

Scope 1 emissions: Direct emissions from owned or controlled sources.

Scope 2 emissions: Indirect emissions from purchased energy.

Scope 3 emissions: Other value-chain emissions.

Work intensification: Increased work demands, pace or monitoring without adequate resources or protection.

Psychological safety: A climate in which people can speak up, report errors and raise concerns without fear of humiliation or retaliation.

Resilience: The capacity to anticipate, absorb, adapt to and recover from disruption.

Recovery time objective: The maximum acceptable time for restoring a process or service.

Responsible sourcing: Procurement that considers cost and performance together with rights, environment, integrity and continuity.

Surveillance: Systematic collection or observation of information about people or activities.

Algorithmic bias: Systematic unfairness produced or amplified by data, model design or deployment.

Human-in-the-loop: Human review or intervention within an automated decision process.

Automation with augmentation: Using technology to remove harmful or routine tasks while strengthening human judgement and capability.

Strategic buffer: Deliberately held capacity, inventory or time justified by risk and continuity requirements.

Lifecycle optimisation: Evaluating impacts across sourcing, production, use and end-of-life rather than at one process step.

Review Questions

Short Answer Questions

1. What is meant by responsible lean?
2. Distinguish between linear and circular operations.
3. Why can excessive just-in-time reduction weaken resilience?
4. Define work intensification and identify two possible consequences.
5. What is the purpose of human review in an AI-enabled operating system?

Long Answer Questions

1. Explain how green value-stream mapping extends conventional value-stream mapping. Include suitable measures and implementation cautions.
2. Discuss the relationship between employee well-being and operational performance in a lean system.
3. Evaluate the role of responsible sourcing in both ethical performance and supply resilience.
4. Explain the principal ethical risks of digital surveillance and algorithmic bias in operations.
5. Compare human-centric, adaptive and AI-enabled operating models.

Analytical / Application Questions

1. A plant reduces inventory by 50% but experiences repeated shutdowns after a single supplier fails. Analyse the decision using lean and resilience principles and recommend a revised policy.
2. An algorithm ranks employees by scanner-measured cycle time, but quality inspectors appear unproductive because inspection work is not captured. Diagnose the measurement problem and design an ethical alternative.
3. A manufacturer is considering reusable packaging. Identify the operational, environmental and social data required to determine whether the change is genuinely sustainable.

Multiple Choice Questions with Answers

Question 1

Which statement best defines responsible lean?

A. Minimising inventory regardless of risk

B. Maximising employee speed through continuous monitoring

C. Reducing waste while protecting people, ecosystems, rights and continuity

D. Replacing all human work with automation

Correct Answer: C. Reducing waste while protecting people, ecosystems, rights and continuity

Question 2

If a process consumes 500 kWh to produce 1,000 good units, what is its energy intensity?

A. 0.5 kWh per good unit

B. 2 kWh per good unit

C. 500 kWh per good unit

D. 1,500 kWh per good unit

Correct Answer: A. 0.5 kWh per good unit

Question 3

What is a major limitation of treating takt time as an individual work quota?

A. It eliminates customer demand information

B. It automatically accounts for fatigue and ergonomic limits

C. It may ignore variation, recovery needs and safe work pace

D. It can only be used in service operations

Correct Answer: C. It may ignore variation, recovery needs and safe work pace

Question 4

Which activity is most characteristic of circular operations?

A. Disposing of all returned products

B. Repairing and remanufacturing products to retain value

C. Increasing single-use packaging

D. Eliminating reverse logistics

Correct Answer: B. Repairing and remanufacturing products to retain value

Question 5

Which capability involves restoring operations and learning after a disruption?

A. Recovery and learning

B. Demand concealment

C. Local optimisation

D. Supplier dependency

Correct Answer: A. Recovery and learning

Question 6

Why can supplier audits alone fail to ensure responsible sourcing?

A. Audits always measure too many variables

B. Suppliers may conceal conditions or subcontracting

C. Audits eliminate the need for traceability

D. Audits measure only customer satisfaction

Correct Answer: B. Suppliers may conceal conditions or subcontracting

Question 7

Which is an example of algorithmic bias?

A. A model using representative data and human review

B. A system explaining its recommendations to affected workers

C. A scheduling model penalising workers because historical data reflect unequal opportunities

D. A maintenance model detecting machine anomalies

Correct Answer: C. A scheduling model penalising workers because historical data reflect unequal opportunities

Question 8

What is a central principle of automation with augmentation?

A. Automate every judgement task

B. Use technology to strengthen human capability while retaining appropriate human judgement

C. Remove employee consultation from technology projects

D. Use automation primarily for disciplinary ranking

Correct Answer: B. Use technology to strengthen human capability while retaining appropriate human judgement

Question 9

Which measure is most directly associated with Scope 2 emissions?

A. Emissions from an organisation’s own boiler

B. Emissions from customer use of a product

C. Emissions from purchased electricity

D. Emissions from supplier transport only

Correct Answer: C. Emissions from purchased electricity

Question 10

In the Meridian case, why was the individual AI ranking suspended?

A. It required no data

B. It failed to capture quality work and created unfair performance interpretations

C. It improved transparency too much

D. It increased the number of qualified suppliers

Correct Answer: B. It failed to capture quality work and created unfair performance interpretations