Digitalization, AI, Asset Management and Future Developments

Introduction

Solar and renewable-energy systems are becoming digitally connected, geographically distributed and increasingly dependent on software. A modern photovoltaic (PV) plant is not simply an array of modules connected to an inverter. It is a cyber-physical system containing sensors, communication networks, control algorithms, weather interfaces, market signals, maintenance databases and grid-support functions. Wind turbines, battery systems, electric vehicles and hybrid renewable plants have similar characteristics.

Digitalization enables operators to observe assets continuously, compare actual performance with expected performance, identify abnormal behaviour and coordinate maintenance. Artificial intelligence (AI) can assist with forecasting, fault diagnosis, predictive maintenance and operational optimization. However, digital tools do not automatically produce reliable decisions. Poorly calibrated sensors, missing data, biased training sets, insecure networks and unexplained algorithms can create operational and financial risks.

This chapter examines the architecture and practical use of supervisory control and data acquisition (SCADA), Internet of Things (IoT) sensors, digital twins, geographic information systems (GIS), data analytics and AI in renewable energy. It then considers emerging technologies, including perovskite and tandem PV, floating and agrivoltaic systems, advanced inverters, vehicle-to-grid operation and circular manufacturing.

Learning Objectives

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

1. Describe how SCADA, IoT, GIS and digital twins support renewable-energy operations.
2. Calculate and interpret key performance indicators such as availability, performance ratio and forecast error.
3. Explain AI applications in forecasting, fault diagnosis, predictive maintenance and optimization.
4. Evaluate data quality, model validation, explainability, cybersecurity and governance requirements.
5. Compare important future developments in PV, inverters, sector coupling and circular manufacturing.
6. Formulate a practical digital asset-management strategy for a renewable-energy plant.

1. From Monitoring to Digital Asset Management

Asset management is the coordinated management of physical assets, information, people, processes and risk throughout an asset’s life cycle. In a renewable plant, the life cycle includes resource assessment, design, procurement, construction, commissioning, operation, maintenance, repowering and decommissioning.

Traditional monitoring answers questions such as: “What is the present power output?” Digital asset management asks broader questions:

• Is the asset producing what it should produce under current weather conditions?
• Which component is degrading fastest?
• What maintenance action minimizes risk and life-cycle cost?
• How will a proposed control action affect grid stability, revenue and equipment life?
• Can evidence from operations support warranty, insurance and regulatory requirements?

A useful asset-management cycle is:

Measure → Communicate → Store → Analyse → Decide → Act → Verify.

The cycle is continuous. A decision is valuable only when it leads to an appropriate action and the result is verified.

1.1 The digital architecture

A typical renewable-energy digital architecture has five layers:

The architecture should distinguish operational technology (OT), which controls physical processes, from information technology (IT), which stores and processes information. A failure in an office database may be inconvenient; a failure in an inverter-control network may interrupt power production or compromise grid stability.

2. SCADA, IoT and Industrial Data

2.1 SCADA systems

SCADA systems supervise geographically distributed equipment. A plant-level SCADA system generally includes sensors, remote terminal units or programmable logic controllers, communication networks, a central historian and a human–machine interface (HMI).

SCADA commonly records:

• active and reactive power;
• voltage, current and frequency;
• inverter status and alarms;
• module, ambient and transformer temperatures;
• irradiance, wind speed, wind direction and humidity;
• breaker position and protection status; and
• energy exported to the grid.

SCADA is designed for operational visibility and control. Its data may be sampled every second, minute or several minutes, depending on the variable and the application. High-frequency electrical data may be needed for power-quality analysis, whereas five-minute data may be adequate for routine energy reporting.

2.2 IoT sensors

IoT extends monitoring beyond conventional plant instrumentation. Wireless sensors can be installed on bearings, tracker drives, module frames, battery racks or transformer enclosures. Examples include vibration sensors, acoustic sensors, strain gauges, infrared sensors and fibre-optic temperature sensors.

IoT deployments require attention to battery life, calibration, radio coverage, time synchronization, electromagnetic interference and physical security. A low-cost sensor that drifts significantly may be less useful than a more expensive, stable sensor.

2.3 Data acquisition and time consistency

Every measurement should have a value, timestamp, unit, quality flag and source identifier. Time synchronization is essential when comparing irradiance, power and equipment states. Network Time Protocol may be sufficient for routine applications, while higher-precision applications can require precision time protocols or GPS-referenced clocks.

Data should not be silently overwritten. Corrections should preserve the original value, the corrected value, the reason for correction and the person or system making the change. This creates an audit trail.

3. Performance Assessment and Useful Metrics

3.1 Energy, power and availability

Power is the instantaneous rate of energy transfer and is measured in watts or multiples such as kW and MW. Energy is power integrated over time and is measured in Wh, kWh or MWh.

For a time interval, energy is:

where E is energy, P(t) is power as a function of time, and t is time. For uniformly sampled data, a practical approximation is:

where P_i is the measured power during interval i and ∆t_i is the interval duration. The approximation is appropriate when sampling intervals are sufficiently short and power does not vary rapidly within an interval.

Operational availability may be defined as:

where T_available is the time during which the plant or component can perform its required function, and T_scheduled is the time during which operation is expected. Contractual definitions vary. Some exclude grid curtailment, force majeure, planned maintenance or periods without sufficient resource.

3.2 PV performance ratio

A common PV indicator is performance ratio (PR):

where Y_f is final yield and Y_r is reference yield. For a PV plant:

and

Here E_AC is AC energy delivered, P_rated is installed DC power under the stated rating convention, H_POA is plane-of-array irradiation, and G_ref is reference irradiance, commonly 1 kW/m². Y_f and Y_r have units of hours, so PR is dimensionless. PR includes losses from temperature, soiling, mismatch, wiring, inverter conversion, availability and other mechanisms. It should not be interpreted as module efficiency.

Worked Example: PV performance ratio

A 10 MWp PV plant exports 42 MWh during a day. The measured plane-of-array irradiation is 6.0 kWh/m², and the reference irradiance is 1 kW/m².

The final yield is:

The reference yield is:

Therefore:

The result is plausible for a daily value, but interpretation requires information about temperature, curtailment, soiling, outages and sensor accuracy.

3.3 Forecast-error metrics

Forecasting supports dispatch, balancing and market participation. For observations y_t and forecasts ŷ_t over n periods, mean absolute error is:

Root mean square error is:

MAE expresses the average absolute error in the same units as the target. RMSE penalizes large errors more strongly. A normalized metric may divide by installed capacity or mean observed output, but the denominator must be stated because different normalizations are not interchangeable.

4. GIS and Spatial Intelligence

GIS links asset records to geographic coordinates and spatial layers. In renewable-energy planning and operations, GIS can combine solar or wind resource, terrain, roads, transmission lines, land ownership, protected areas, flood risk, population density and environmental constraints.

GIS supports:

• site screening and yield estimation;
• cable-route and access-road planning;
• identification of vegetation or shading risks;
• inspection planning using spatial clusters;
• wildfire, flood and storm-risk assessment;
• community and environmental consultation; and
• prioritization of spare parts and field teams.

A GIS database should distinguish geometric accuracy from attribute accuracy. A transformer may be located correctly on a map while its capacity, commissioning date or maintenance status is wrong. Coordinate systems, metadata and update responsibility should therefore be documented.

5. Digital Twins

A digital twin is a continuously updated digital representation of a physical asset or system, connected to operational data and capable of supporting analysis or decisions. It is more than a static three-dimensional model.

A useful twin may contain:

1. Identity: asset number, manufacturer, model and location.
2. Structure: electrical topology and component relationships.
3. State: current measurements and operating status.
4. Physics or empirical behaviour: expected response under conditions.
5. History: alarms, interventions, degradation and performance.
6. Decision interface: simulations, alerts and recommended actions.

Digital twins can be physics-based, data-driven or hybrid. A physics-based PV model may estimate expected power from irradiance, temperature and electrical characteristics. A data-driven model may learn normal behaviour from historical measurements. A hybrid model can use physical constraints to prevent unreasonable predictions.

The twin’s credibility depends on calibration and validation. A model that reproduces historical data but fails during unusual weather or sensor replacement is not reliable for operational decisions. Model versions, calibration data and uncertainty should be recorded.

6. Artificial Intelligence in Renewable Energy

AI includes statistical learning, machine learning, deep learning, optimization and knowledge-based methods. The appropriate method depends on the problem, data volume, required response time and consequences of error.

6.1 Forecasting

Forecasts may concern solar irradiance, PV power, wind power, electricity demand, prices or equipment temperature. Inputs can include numerical weather prediction, satellite imagery, sky cameras, recent plant output, calendar variables and nearby weather stations.

A forecasting workflow normally includes:

1. Define the forecast horizon and time resolution.
2. Establish a simple baseline, such as persistence.
3. Collect and align historical data.
4. Separate training, validation and test periods chronologically.
5. Train and tune the model without leakage from future information.
6. Evaluate using MAE, RMSE and operationally relevant measures.
7. Monitor performance after deployment.

Persistence forecasting assumes that the next value equals the latest value. It is a useful baseline, not necessarily a good final model. A complex neural network should not be adopted unless it provides meaningful improvement over simpler alternatives.

6.2 Fault diagnosis

Fault diagnosis may use alarm sequences, electrical signatures, thermal images, IV-curve measurements, vibration and weather-normalized power. Classification models can identify categories such as inverter trip, string outage, tracker malfunction or communication failure. Anomaly detection is useful when labelled fault data are scarce.

False positives create unnecessary inspections; false negatives allow damage or lost production to continue. The relative costs of both errors should be incorporated into threshold selection.

6.3 Predictive maintenance

Predictive maintenance estimates the probability or timing of failure from condition data. A simplified survival representation uses a hazard rate λ(t), describing the instantaneous failure intensity conditional on survival to time t. In practice, maintenance models may use survival analysis, gradient-boosting models, recurrent networks or remaining-useful-life estimation.

Predictions should lead to a work decision. An alert is not automatically a maintenance order. The operator should consider safety, production loss, spare-part availability, weather windows, technician access and warranty conditions.

6.4 Optimization and control

Optimization can schedule battery charging, select curtailment levels, coordinate reactive power and voltage control, or prioritize maintenance activities. A generic constrained optimization problem is:

Here x represents system states, u represents decisions or control inputs, J is the objective function, g represents equality constraints such as power balance, and h represents inequality constraints such as voltage, current, state-of-charge and thermal limits. The formulation is only valid when the objective and constraints correctly represent physical and operational requirements.

An optimizer must not be allowed to issue unrestricted commands to critical equipment. Supervisory limits, interlocks, human authorization and fail-safe states remain necessary.

7. Data Quality, Explainability and Governance

7.1 Data quality

Important dimensions of data quality include accuracy, completeness, consistency, timeliness, validity, uniqueness and traceability. A practical data-quality process should:

• identify impossible values, such as negative irradiance during daylight when the sensor cannot produce them;
• detect flat-line signals and repeated timestamps;
• compare redundant sensors;
• mark outages rather than replacing them silently;
• distinguish zero production from missing production;
• document imputation methods; and
• retain uncertainty and quality flags.

Data cleaning can improve a model, but excessive cleaning can remove genuine rare events. Every transformation should be reproducible.

7.2 Explainability

Operators need to understand why an AI system issued an alert or recommendation. Explainability can be global, describing overall model behaviour, or local, explaining one prediction. Feature importance, partial-dependence analysis, counterfactual explanations and interpretable models are useful tools.

Explainability does not prove that a model is correct. A model can provide a convincing explanation for a spurious correlation. Human review, physical plausibility checks and independent validation remain necessary.

7.3 Governance

AI governance assigns responsibility for data, models, decisions and consequences. A governance framework should define:

• who owns each data set;
• who approves model deployment;
• acceptable error and uncertainty limits;
• change-control and versioning procedures;
• audit and reporting requirements;
• human override authority;
• incident response; and
• retirement criteria for obsolete models.

Models should be monitored for drift. Drift occurs when the statistical relationship between inputs and outputs changes because of new equipment, sensor replacement, climate conditions, operating policy or market behaviour.

8. Cybersecurity and Resilience

Renewable plants are increasingly exposed to cyber risks because inverters, controllers and cloud services are networked. Threats include unauthorized access, ransomware, manipulation of measurements, denial of service, supply-chain compromise and unsafe control commands.

A robust cybersecurity programme includes:

1. Asset inventory and network segmentation.
2. Least-privilege access and multifactor authentication.
3. Secure configuration and removal of unnecessary services.
4. Encryption and certificate management where appropriate.
5. Patch and vulnerability-management procedures.
6. Backups and tested restoration.
7. Continuous logging and anomaly detection.
8. Incident-response exercises.
9. Vendor-security requirements.
10. Safe manual or local fallback operation.

The Purdue-style separation of enterprise IT, supervisory systems and field-control networks can help limit the consequences of compromise, although modern cloud architectures require careful adaptation. Cybersecurity should be included during procurement and design rather than added after commissioning. Operators must also comply with applicable critical-infrastructure, privacy, electrical-safety and grid-code requirements.

9. Applied Case: Illustrative Digital Management of a Utility PV Plant

Background

Consider an illustrative 50 MWp PV plant with central inverters, single-axis trackers, a weather station and a grid connection. The plant experiences unexplained production losses during hot afternoons. Operators receive many inverter alarms, but maintenance records are inconsistent.

Problem

Monthly performance reports show that availability appears acceptable, yet energy yield is below the engineering estimate. The plant owner needs to distinguish resource variability, thermal losses, tracker faults, curtailment and inverter limitations.

Analysis

The team first establishes a common asset register linking each inverter, tracker block, combiner box and weather sensor to a GIS location. SCADA data are synchronized and quality-flagged. Plane-of-array irradiance, module temperature, ambient temperature, wind speed, inverter output and grid commands are compared at five-minute resolution.

The team calculates expected power using a validated baseline model and compares it with measured power. It finds that several temperature sensors intermittently report constant values. These faulty signals had caused the expected-power model to overestimate output and generated misleading anomaly alerts.

A second analysis groups alarms by inverter and time. The highest-loss blocks show repeated high-temperature derating during low-wind afternoons. Thermal inspection identifies restricted ventilation at inverter enclosures and dust accumulation near air filters.

Approach and solution

The operator implements the following measures:

• redundant temperature sensing at selected reference locations;
• automatic data-quality flags for flat-line and out-of-range signals;
• a digital twin combining weather-normalized expected output with inverter thermal limits;
• a fault-priority score based on estimated energy loss, safety risk and repair time;
• work orders linked to GIS coordinates and photographs;
• a cybersecurity review of remote access; and
• monthly model-performance and data-quality reviews.

The AI component is deliberately limited. An anomaly-detection model ranks unusual inverter behaviour, but technicians retain decision authority. The model is evaluated against confirmed work orders rather than alarm counts alone.

Outcome

The plant reduces repeated nuisance alarms and identifies thermal derating earlier. Maintenance crews travel less because work orders are spatially grouped and prepared with relevant component histories. Management gains a more credible separation between true equipment loss and sensor error.

Learning points

The case demonstrates that digital value comes from integration, not from AI alone. A reliable asset register, synchronized data, clear maintenance processes and cybersecurity controls are prerequisites for useful analytics. Model accuracy should be judged by improved decisions and reduced risk, not only by statistical scores.

10. Future Developments in Solar and Renewable Energy

10.1 Perovskite and tandem PV

Perovskite materials can be tuned to absorb selected parts of the solar spectrum. In a tandem cell, two or more subcells with different band gaps convert different portions of sunlight, potentially reducing thermalization losses compared with a single-junction cell.

The principal challenges include long-term stability, moisture and heat sensitivity, large-area manufacturing, encapsulation, lead management and bankability. Laboratory efficiency does not by itself establish field reliability or life-cycle sustainability. Tandem modules must be evaluated through accelerated testing, outdoor performance, manufacturing yield and end-of-life planning.

10.2 Floating PV

Floating PV systems are installed on reservoirs, quarry lakes or other water bodies. Potential advantages include reduced land competition, lower evaporation in some settings and possible cooling effects. Risks include wind loading, wave motion, anchoring failure, biofouling, corrosion, difficult access and ecological effects. Electrical insulation, earthing and maintenance procedures must account for the aquatic environment.

10.3 Agrivoltaics

Agrivoltaics co-locates agriculture and PV generation. Elevated or spaced modules can provide partial shading while allowing cultivation or grazing. Outcomes depend on crop species, latitude, module geometry, irrigation, local climate and farm operations. Agrivoltaics should be assessed using both electricity metrics and agricultural indicators such as yield, water use and farm income. It is not automatically beneficial for every crop or site.

10.4 Advanced inverters

Modern grid-forming and grid-following inverters provide functions beyond simple DC-to-AC conversion. These may include reactive-power control, voltage support, frequency response, fault ride-through, ramp-rate control, synthetic inertia-like response and black-start support in suitable systems.

Grid-following inverters use an existing voltage waveform as a reference, often through a phase-locked loop. Grid-forming inverters regulate a voltage-source-like output and can contribute to stability in weak grids or islanded microgrids. Their deployment requires coordinated protection, validated controls, communications and grid-code compliance.

10.5 Vehicle-to-grid systems

Electric vehicles can act as flexible loads and, where bidirectional hardware and suitable agreements exist, export power to buildings or the grid. A simplified battery-energy balance is:

where E_t is stored energy, P_c is charging power, P_d is discharging power, η_c and η_d are charging and discharging efficiencies, and ∆t is the time interval. The equation assumes consistent power and efficiency conventions during the interval and excludes additional effects such as self-discharge and temperature dependence.

Vehicle-to-grid value is constrained by mobility requirements, battery degradation, user acceptance, charger availability, tariffs, communication standards and distribution-network limits. Optimization must preserve a minimum state of charge for the driver.

10.6 Circular manufacturing

Circular manufacturing seeks to reduce virgin-material use, extend product life, enable repair and reuse, recover valuable materials and minimize waste. For PV, relevant practices include design for disassembly, recyclable encapsulants, module refurbishment, material traceability and recovery of glass, aluminium, silicon and selected metals.

A life-cycle perspective should include manufacturing energy, transport, operation, replacement, recycling and residual value. Recycling is not automatically circular if recovered materials are downcycled or if collection is impractical. Product passports and producer-responsibility systems can improve traceability and accountability.

11. Professional Best Practices

A competent digital-renewable-energy project should:

• define operational questions before selecting technology;
• create a verified asset hierarchy and data dictionary;
• specify units, sampling rates, timestamps and quality flags;
• establish baseline models before deploying AI;
• validate models using chronological and independent data;
• combine statistical evidence with engineering judgement;
• quantify uncertainty and consequences of error;
• separate advisory AI from autonomous control unless safety is demonstrated;
• protect OT networks throughout the supply chain;
• maintain human accountability and audit trails;
• review environmental, social and privacy impacts; and
• plan decommissioning, data retention and hardware disposal.

Chapter Summary

Digitalization transforms renewable-energy plants into connected cyber-physical systems. SCADA provides supervisory monitoring and control, while IoT sensors extend condition monitoring to detailed component and environmental measurements. GIS adds spatial intelligence, and digital twins integrate asset identity, physical behaviour, operating state and history.

Performance assessment requires clear definitions. Energy is the time integral of power; availability compares functional time with scheduled time; PV performance ratio compares final yield with reference yield; and forecasting quality can be evaluated with MAE and RMSE. These metrics are meaningful only when units, data quality and contractual boundaries are stated.

AI can improve renewable-energy forecasting, fault diagnosis, predictive maintenance and optimization. Its effectiveness depends on representative data, sound validation, explainability, drift monitoring and appropriate human oversight. Cybersecurity, governance and auditability are essential because digital decisions can affect physical equipment and grid reliability.

Perovskite and tandem PV may increase conversion efficiency, while floating PV and agrivoltaics address land and water-use constraints in particular contexts. Advanced inverters support increasingly inverter-dominated grids, vehicle-to-grid systems connect mobility with electricity flexibility, and circular manufacturing addresses material and end-of-life challenges. Future engineers must therefore combine electrical knowledge with data engineering, cybersecurity, environmental assessment and responsible decision-making.

Key Terms

SCADA: Supervisory system used to monitor and control geographically distributed equipment.

IoT sensor: Connected sensing device that measures and communicates physical or operational data.

Operational technology: Hardware and software that monitors or controls physical processes.

Historian: Time-series database used to store operational measurements and events.

Availability: Proportion of scheduled time during which an asset can perform its required function.

Performance ratio: PV indicator comparing actual final yield with reference yield.

Final yield: AC energy output divided by rated PV capacity.

Reference yield: Plane-of-array irradiation divided by reference irradiance.

GIS: Geographic information system used to manage and analyse spatial data.

Digital twin: Connected digital representation of a physical asset or system.

Anomaly detection: Identification of observations that differ significantly from expected behaviour.

Predictive maintenance: Maintenance planning based on estimated condition or failure risk.

Model drift: Change in the relationship between model inputs and outputs after deployment.

Explainability: Ability to communicate how a model produces a result or recommendation.

Grid-forming inverter: Inverter capable of establishing or regulating a voltage waveform.

Agrivoltaics: Co-location of photovoltaic generation and agricultural activity.

Vehicle-to-grid: Bidirectional transfer of electricity between vehicles and the grid.

Circular manufacturing: Production approach emphasizing durability, reuse, repair and material recovery.

Review Questions

5 Short Answer Questions

1. What are the main functions of a SCADA system in a renewable-energy plant?
2. Distinguish power from energy.
3. Define performance ratio and identify two factors that influence it.
4. Why are timestamps and quality flags important in operational data?
5. State two cybersecurity measures suitable for renewable-energy OT networks.

5 Long Answer Questions

1. Explain the five-layer digital architecture of a renewable-energy asset and discuss the relationship between IT and OT.
2. Describe how a digital twin can support commissioning, operation and maintenance of a PV plant.
3. Compare MAE and RMSE as forecasting metrics, including the circumstances in which each is useful.
4. Discuss the benefits and limitations of AI for predictive maintenance.
5. Compare perovskite/tandem PV, floating PV and agrivoltaics in terms of opportunities and risks.

3 Analytical / Application Questions

1. A 20 MWp PV plant exports 72 MWh in a day when plane-of-array irradiation is 5.4 kWh/m². Calculate its final yield, reference yield and performance ratio, using 1 kW/m² as reference irradiance. Interpret the result and identify information needed before judging plant performance.
2. Design a data and cybersecurity architecture for a 10 MW PV plant with smart inverters, battery storage and remote maintenance access. Identify data flows, security boundaries, failure modes and fallback procedures.
3. An AI model reports frequent inverter faults, but technicians confirm that many alerts are caused by a drifting temperature sensor. Propose a revised workflow covering data quality, model validation, explainability, human approval and maintenance prioritization.

Multiple Choice Questions with Answers

Question 1

Which metric compares a PV plant’s final yield with its reference yield?

A. Availability

B. Performance ratio

C. Mean absolute error

D. Capacity factor

Correct Answer: B. Performance ratio

Question 2

What is the primary purpose of a SCADA historian?

A. Store time-series operational data

B. Convert DC power to AC power

C. Provide mechanical cooling

D. Replace all field sensors

Correct Answer: A. Store time-series operational data

Question 3

Which forecasting metric penalizes large errors more strongly because errors are squared before averaging?

A. MAE

B. Median error

C. RMSE

D. Availability

Correct Answer: C. RMSE

Question 4

A digital twin is best described as:

A. A static photograph of a plant

B. A connected digital representation of a physical asset

C. A replacement for all physical inspections

D. A geographic coordinate system only

Correct Answer: B. A connected digital representation of a physical asset

Question 5

Which practice best supports reliable AI deployment?

A. Training only on randomly selected future data

B. Ignoring missing-data flags

C. Comparing the model with a simple baseline and independent test data

D. Allowing unrestricted automatic control commands

Correct Answer: C. Comparing the model with a simple baseline and independent test data

Question 6

Which is an example of OT cybersecurity?

A. Network segmentation between control and enterprise systems

B. Increasing module tilt angle

C. Cleaning inverter filters

D. Changing the PV band gap

Correct Answer: A. Network segmentation between control and enterprise systems

Question 7

What is a characteristic objective of agrivoltaics?

A. Eliminating all agricultural activity

B. Co-locating PV generation with agricultural production

C. Installing PV only on ocean platforms

D. Replacing grid protection with AI

Correct Answer: B. Co-locating PV generation with agricultural production

Question 8

In a vehicle battery balance equation, charging efficiency represents:

A. The fraction of charging input stored as battery energy

B. The vehicle’s travel speed

C. The grid frequency

D. The PV module temperature

Correct Answer: A. The fraction of charging input stored as battery energy

Question 9

What distinguishes a grid-forming inverter from a basic grid-following inverter?

A. It can help establish or regulate a voltage waveform

B. It never requires protection

C. It operates only with DC loads

D. It cannot provide reactive power

Correct Answer: A. It can help establish or regulate a voltage waveform

Question 10

Which action is most consistent with circular manufacturing?

A. Designing products that cannot be disassembled

B. Maximizing single-use materials

C. Designing for repair, reuse and material recovery

D. Discarding modules after minor faults

Correct Answer: C. Designing for repair, reuse and material recovery