11. Digital Transformation, AI and Sustainable Logistics

Introduction and Context

Logistics has always depended on information: where goods are located, when they will arrive, what condition they are in, and which resources are required to move them. Digital transformation changes the scale, speed, accuracy and strategic value of this information. Sensors can report the temperature of a pharmaceutical shipment, telematics can record the fuel use of a vehicle, artificial intelligence can revise a delivery plan, and a digital twin can test the environmental consequences of alternative network designs before a physical decision is made.

Digitalisation, however, is not automatically sustainable. A poorly designed digital system may increase energy consumption, create electronic waste, intensify surveillance, or produce decisions that are efficient only because environmental and social costs have been ignored. Sustainable logistics therefore requires more than adopting advanced technologies. It requires the purposeful integration of data, analytical capability, governance, operational processes and sustainability objectives.

This chapter examines the principal digital technologies used in sustainable logistics: the Internet of Things (IoT), telematics, cloud platforms, control towers, digital twins, blockchain and artificial intelligence. It considers how these technologies support demand forecasting, routing, emissions visibility, traceability and predictive maintenance. It also evaluates data quality, cybersecurity, algorithmic bias and the technology footprint itself. The central argument is that digital transformation creates value when it improves the quality of decisions while preserving accountability, resilience and environmental integrity.

Learning Objectives

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

1. Explain the role of digital technologies in sustainable logistics and green supply chain management.

2. Distinguish among IoT, telematics, cloud platforms, control towers, digital twins, blockchain and AI.

3. Assess how digital tools improve forecasting, routing, emissions measurement, traceability and maintenance.

4. Apply a framework for evaluating data quality, interoperability, cybersecurity and technology footprints.

5. Recognise algorithmic bias, governance risks and ethical issues in logistics automation.

6. Evaluate digital transformation choices from operational, environmental, managerial and strategic perspectives.

From Visibility to Intelligent Orchestration

The digital maturity of a logistics organisation can be understood as a progression. At the first level, information is recorded manually and decisions are largely reactive. At the second, transactions are digitised through enterprise resource planning, warehouse management and transport management systems. At the third, physical assets generate continuous data through sensors and telematics. At the fourth, information from multiple organisations is integrated through platforms and control towers. At the fifth, advanced analytics and AI support prediction, optimisation and autonomous or semi-autonomous action.

This progression should not be mistaken for a simple technology ladder. A company may own sophisticated software but still make poor decisions because data is incomplete, responsibilities are unclear or users do not trust the system. Digital maturity is consequently a socio-technical capability involving people, processes, technology, governance and organisational learning.

A useful distinction is between visibility and intelligence. Visibility answers questions such as: Where is the shipment? What is the current inventory level? How much fuel has been consumed? Intelligence goes further by asking: What is likely to happen? Which action will achieve the lowest total impact? What trade-off exists between service, cost, resilience and emissions? Sustainable logistics depends on both. Visibility makes impacts measurable; intelligence helps managers act on that information.

The Digital Architecture of Sustainable Logistics

IoT and sensor-enabled operations

The Internet of Things refers to connected physical objects that collect, transmit and sometimes receive data. In logistics, these objects include vehicles, pallets, containers, forklifts, refrigeration units, loading docks and warehouse equipment. Sensors may capture location, temperature, humidity, vibration, pressure, door opening, energy use or engine performance.

IoT supports sustainability in several ways. Continuous temperature monitoring can prevent product spoilage and the emissions associated with producing and transporting replacement goods. Fill-level sensors can trigger waste collection only when required. Occupancy and energy sensors can reduce unnecessary lighting, heating and cooling in warehouses. Condition monitoring can identify abnormal vibration before a machine fails.

The value of an IoT device depends on the decision linked to its data. Installing sensors without defining ownership, response procedures and performance indicators produces data accumulation rather than transformation. Sensor calibration, battery life, connectivity and physical security are also important. A sensor that systematically reports temperatures lower than the true temperature may create false confidence and operational risk.

Telematics and fleet sustainability

Telematics combines telecommunications, vehicle data and information processing. Fleet telematics commonly records position, speed, idling, acceleration, braking, fuel or electricity consumption, engine status and driving hours. It provides a factual basis for route analysis, driver coaching, maintenance planning and emissions estimation.

For conventional vehicles, telematics can identify excessive idling, inefficient routes, harsh acceleration and underutilised capacity. For electric fleets, it can support charging schedules, battery monitoring and range planning. Telematics can also help compare planned routes with actual routes and identify recurring causes of delay.

The managerial challenge is to use telematics for improvement rather than solely for employee surveillance. Performance measures should distinguish behaviour within the driver’s control from factors such as congestion, weather, loading delays or defective planning. Transparent policies, proportional monitoring and opportunities for explanation are essential for legitimacy.

Cloud platforms and data interoperability

Cloud platforms provide shared computing, storage and application services through network connections. They allow shippers, carriers, suppliers and logistics providers to access common information without each organisation maintaining identical infrastructure. Cloud systems support scalability, remote access, software updates and integration across transport, warehouse, procurement and sustainability applications.

Their sustainability contribution arises mainly through coordination. A shared platform can reduce duplicated data entry, improve vehicle and warehouse utilisation, and support collaborative transport planning. Nevertheless, interoperability remains difficult because organisations use different data formats, definitions and identification systems. “On-time delivery,” for example, may be measured against a promised date, a requested date or a delivery appointment. Without common definitions, apparently precise analytics may conceal inconsistent measurement.

Control towers

A control tower is an integrated visibility and decision-support capability that combines data from multiple systems and partners. It may display shipment status, inventory risk, exceptions, capacity, weather, congestion, energy use and emissions. Its purpose is not merely to provide a dashboard but to support coordinated intervention.

In sustainable logistics, a control tower can identify consolidation opportunities, prioritise low-emission transport, detect temperature excursions, compare carrier performance and manage disruptions. Effective control towers generally include exception management: the system identifies events requiring attention and directs them to an accountable person or team. A dashboard with hundreds of indicators but no decision rights is not a control tower in the managerial sense.

Digital twins

A digital twin is a dynamic digital representation of a physical asset, process or network, updated by operational data and used for analysis or simulation. A warehouse twin may represent storage locations, labour, equipment and flows. A fleet twin may represent vehicles, routes, charging infrastructure and utilisation. A supply network twin may model suppliers, facilities, transport modes, inventories and demand.

Digital twins support sustainability by allowing alternatives to be tested before implementation. Managers can compare a new distribution centre, different inventory policies, alternative vehicle mixes or revised delivery frequencies. A twin can estimate service, cost, energy and emissions under different demand and disruption scenarios. Its results are only as reliable as its assumptions, data and model boundaries. If upstream production, empty return journeys or customer travel are excluded, the twin may understate total environmental impact.

Blockchain and distributed traceability

Blockchain is a distributed ledger technology in which transactions are recorded in linked blocks and validated through an agreed consensus mechanism. In supply chains, it may support shared records of product origin, custody transfers, certifications or compliance events.

Its potential sustainability value lies in traceability and trust. A buyer may verify that a material passed through declared stages, while a regulator or auditor may examine the history of a transaction. Smart contracts can automate actions when predefined conditions are met. However, blockchain does not guarantee that the original data is truthful. If a fraudulent or inaccurate event is entered, the ledger may preserve the error. Governance over data entry, participant identity and access remains essential. The energy implications also depend on the consensus mechanism and system design; not every traceability problem requires a blockchain.

Artificial Intelligence in Sustainable Logistics

Artificial intelligence refers broadly to computational methods that perform tasks associated with human reasoning, learning, perception or decision-making. In logistics, machine learning is particularly important. It identifies patterns in historical and real-time data to produce predictions or recommendations.

Forecasting and inventory decisions

AI-supported forecasting can combine order history with promotions, weather, calendar effects, economic indicators, local events and online behaviour. More accurate forecasts can reduce emergency transport, excess inventory, obsolescence and stockouts. Forecasting models should be evaluated not only by statistical accuracy but also by operational and environmental consequences. A small forecast error for a high-volume product may matter more than a large error for a low-volume product. The cost of holding excess stock, the emissions of expedited delivery and the social cost of stockouts should be considered together.

Human judgement remains relevant because structural changes, unusual disruptions and ethical priorities may not be represented in historical data. A model trained on normal conditions may fail during a pandemic, port closure or severe weather event.

Routing and optimisation

AI and operations research can support vehicle routing, load consolidation, delivery sequencing and multimodal choices. A sustainable routing model may minimise a weighted objective involving distance, fuel, emissions, delivery time, congestion, driver hours and service commitments. The “shortest route” is not necessarily the lowest-emission route: road gradient, traffic, vehicle type, payload and idling can change the result.

Managers must define the objective carefully. If the model is rewarded only for cost reduction, it may select solutions that increase emissions or transfer burdens to subcontractors. Multi-objective optimisation makes trade-offs visible, while constraints can protect minimum service levels, legal driving hours, safety and labour conditions.

Predictive maintenance

Predictive maintenance uses sensor and maintenance data to estimate the likelihood or timing of equipment failure. It differs from corrective maintenance, which occurs after failure, and preventive maintenance, which occurs at scheduled intervals. By servicing assets according to condition, organisations may reduce breakdowns, spare-part waste, emergency transport and inefficient operation.

A predictive system should be assessed through maintenance outcomes rather than prediction accuracy alone. False alarms may generate unnecessary part replacement, while missed failures can create serious disruption. Maintenance staff should be able to challenge a prediction and record contextual information that the model cannot observe.

Emissions visibility and carbon intelligence

Digital systems can estimate emissions by combining activity data—such as kilometres, fuel, electricity, payload, mode and shipment weight—with appropriate emission factors. Higher-quality systems distinguish actual measured fuel or electricity from assumptions and show the confidence or quality of the estimate. They may allocate transport emissions to shipments using weight, volume, distance, time or another documented rule.

Emissions dashboards support hotspot analysis, supplier comparison, customer reporting and target monitoring. Yet digital precision should not be confused with environmental truth. Different boundaries, allocation rules and emission factors can produce different results. Organisations should document methodology, avoid double counting and disclose material assumptions.

A Decision Framework for Digital Sustainability

A practical evaluation can follow six questions.

First, what decision is being improved? Technology should be connected to a business and sustainability problem, such as empty running, spoilage, poor asset utilisation or unreliable emissions data.

Second, what data is required? Identify sources, frequency, ownership, granularity, retention and permitted uses.

Third, is the data fit for purpose? Examine accuracy, completeness, consistency, timeliness, validity, uniqueness and representativeness. A dataset can be technically complete but conceptually unsuitable.

Fourth, what model or workflow converts data into action? Specify the algorithm, assumptions, thresholds, human roles and escalation procedures.

Fifth, what are the total impacts? Assess cost, emissions, energy use, social consequences, resilience, privacy and cybersecurity, not merely software performance.

Sixth, how will performance be governed? Establish indicators, audit trails, review intervals, accountability and procedures for correcting errors or challenging automated decisions.

This framework prevents technology-first investment. It also encourages a life-cycle view of digital systems, including equipment manufacture, connectivity, data-centre energy, software maintenance, replacement and disposal.

Data Quality, Cybersecurity and Responsible AI

Data quality is a sustainability issue because poor data can lead directly to wasteful decisions. Duplicate shipment records may cause unnecessary dispatches; missing payload data may distort emissions allocation; delayed location data may cause avoidable expediting. Data quality management should include validation rules, master-data governance, reconciliation between systems, sensor calibration and clear definitions.

Cybersecurity is equally important. Connected vehicles, warehouses and control systems expand the attack surface. A cyberattack may interrupt deliveries, manipulate temperature records, disable charging infrastructure or expose commercially sensitive information. Basic controls include strong identity management, least-privilege access, encryption, network segmentation, patch management, supplier assurance, backups and tested incident-response plans. Resilience requires the ability to operate safely when connectivity or automated services are unavailable.

Responsible AI requires attention to explainability, fairness, accountability, privacy, safety and human oversight. Algorithmic bias can arise from unrepresentative training data, historical discrimination, proxy variables or unequal data quality among suppliers and regions. For example, a carrier-rating model may penalise firms operating in congested urban areas without distinguishing congestion from poor management. A workforce scheduling model may allocate undesirable shifts disproportionately to a particular group if historical patterns are treated as neutral.

Organisations should conduct impact assessments before deployment, test performance across relevant groups and operating conditions, document model versions, monitor drift and provide mechanisms for appeal. Human oversight must be meaningful rather than symbolic: a person should have the authority, competence and time to question an automated recommendation.

Illustrative Case: A Regional Retailer’s Low-Emission Distribution Programme

Background

The following is an illustrative case. Meridian Retail, a regional retailer with 80 stores and an online channel, operates two distribution centres and contracts a mixed fleet of company-owned and third-party vehicles. Management has committed to reducing logistics emissions while maintaining next-day service.

Situation and problem

The company’s transport data was fragmented. Fuel records were available monthly, carrier invoices used different distance measures, and delivery exceptions were recorded inconsistently. Vehicles often returned partially empty, while urgent replenishment orders generated additional trips. Managers could not identify whether emissions were driven primarily by routing, vehicle choice, low utilisation or demand volatility.

Analysis

Meridian first created a common data model for orders, vehicles, routes, payloads, energy use and delivery events. Telematics was installed on the company fleet, while contracted carriers supplied standardised activity data through a cloud platform. A control tower combined transport management, warehouse and order information. A pilot digital twin modelled delivery frequency, consolidation rules and the gradual introduction of electric vehicles.

An AI forecasting model was then tested for selected product categories. It improved predictions in stable categories but performed poorly during promotional events, so planners retained an override process. A routing model considered distance, vehicle capacity, delivery windows, driver-hour constraints and estimated emissions rather than distance alone.

Approach and solution

The programme was implemented in stages. Data definitions and governance were established before automation. A baseline was calculated using documented boundaries and emission factors. Managers introduced weekly reviews of empty kilometres, energy per delivered case, urgent shipments, vehicle utilisation and delivery service. The control tower generated exceptions, but dispatchers retained responsibility for final decisions.

Meridian also created a cybersecurity plan, restricted access to driver data and established a retention period. Drivers were consulted about telematics and were given explanations of how information would be used. The company rejected blockchain for the pilot because its immediate traceability problem could be addressed more simply through signed digital records and stronger supplier controls.

Outcome

The pilot identified avoidable empty running, improved the timing of replenishment and allowed electric vehicles to be assigned to routes compatible with their range and charging availability. The organisation did not treat the early model outputs as proof of permanent savings. Instead, it continued measurement across seasons and monitored whether consolidation affected product availability or driver workload.

Learning points

The case illustrates five principles. First, digital transformation begins with a decision problem, not with a fashionable technology. Second, standardised data is a prerequisite for credible comparison. Third, AI should supplement operational expertise, particularly when conditions change. Fourth, a digital twin is useful only when its boundaries and assumptions are explicit. Fifth, sustainability performance requires simultaneous attention to emissions, service, labour, privacy, security and resilience.

Managerial and Professional Implications

Managers must develop digital literacy without assuming that they need to become software engineers. They should be able to interrogate data definitions, understand model limitations, compare total cost of ownership, evaluate vendors and identify governance gaps. Procurement teams should require interoperability, data portability, cybersecurity evidence, service-level commitments and clear ownership of generated data.

Sustainability professionals increasingly need competence in data architecture and measurement, while information technology specialists need to understand logistics processes and environmental accounting. Cross-functional governance is therefore essential. A digital steering group may include operations, IT, procurement, sustainability, legal, security, finance and workforce representatives.

Performance indicators should combine leading and lagging measures. Leading indicators include data completeness, sensor uptime, model review frequency, cybersecurity training and percentage of decisions subject to human review. Lagging indicators include fuel use, emissions, spoilage, empty kilometres, equipment downtime, safety events and service failures. Indicators should be interpreted together: reducing vehicle kilometres by delaying deliveries may lower one metric while worsening customer service or inventory waste.

Challenges and Limitations

Digital transformation can create vendor dependence, high implementation costs and integration difficulties. Small suppliers may lack the financial or technical capacity to participate in sophisticated platforms. Unequal digital capability can therefore shift power toward large firms and exclude smaller partners from valuable networks.

Automation may also deskill work or increase work intensity if targets are imposed without regard to operational reality. Continuous monitoring can undermine trust. Data sharing may expose commercial secrets or personal information. Digital systems can fail through connectivity loss, model drift, poor configuration or deliberate attack.

There is also a rebound effect. More efficient delivery may reduce the cost of rapid fulfilment and stimulate additional orders, vehicle movements and packaging. Data centres, sensors, batteries, network equipment and computing hardware have material and energy footprints. Sustainable digitalisation must therefore measure the net effect of technology, not only the efficiency achieved in the immediate process.

Common mistakes include purchasing dashboards without decision rights, treating estimated emissions as exact measurements, assuming immutable records are truthful, using historical data without testing for bias, and automating a process that has not been standardised. Another mistake is to define success solely as return on investment while ignoring resilience, worker welfare and long-term environmental consequences.

Emerging Directions and Best Practice

Several developments are shaping the next phase of sustainable logistics. Edge computing allows some data processing close to the vehicle or sensor, reducing latency and dependence on continuous cloud communication. Generative AI may assist planners by summarising disruptions, explaining alternatives or interacting with logistics data, but its outputs require verification. Computer vision is being used for warehouse safety, inventory recognition and damage inspection, raising important questions about privacy and accuracy.

Digital product information and product passports may improve material traceability, repairability and end-of-life management. Greater use of electric vehicles will increase the importance of charging orchestration, battery health analytics and renewable electricity information. Digital twins are likely to become more integrated across facilities and networks, although model governance will remain a major challenge.

Best practice is characterised by the following principles:

• Begin with a defined sustainability and operational decision.
• Establish common data definitions and quality controls before scaling.
• Integrate sustainability indicators into ordinary planning and performance systems.
• Use the least complex technology that reliably solves the problem.
• Design cybersecurity, privacy and resilience into the system from the outset.
• Keep accountable human decision-makers for material consequences.
• Test algorithms for accuracy, bias, drift and unintended effects.
• Assess the full life-cycle footprint of digital equipment and services.
• Involve employees, suppliers and affected communities in system design.
• Review outcomes continuously rather than treating implementation as completion.

Chapter Summary

Digital transformation can make sustainable logistics more visible, predictive and coordinated. IoT and telematics collect operational information; cloud platforms integrate participants; control towers coordinate exceptions; digital twins test alternatives; blockchain can support shared traceability; and AI can improve forecasting, routing, emissions analysis and predictive maintenance. Their value depends on the quality of data and on the connection between information and accountable decisions.

Technology is not inherently sustainable. Its benefits must be weighed against cybersecurity threats, privacy concerns, algorithmic bias, vendor dependence, exclusion of smaller partners and the energy and material footprint of digital infrastructure. A responsible approach combines technical architecture with governance, human judgement, transparent measurement and life-cycle evaluation. The strategic objective is not maximum automation but better logistics decisions that simultaneously support service, resilience, economic performance, environmental protection and social legitimacy.

Key Terms

Artificial intelligence: Computational capability that performs tasks involving learning, reasoning, perception or decision support.

Algorithmic bias: Systematic and unfair distortion in an algorithm’s outputs arising from data, design or implementation.

Blockchain: A distributed ledger in which validated transactions are recorded in linked and tamper-evident blocks.

Cloud platform: Network-accessible computing and data services that support shared applications and integration.

Control tower: An integrated visibility and decision-support capability that coordinates supply-chain events and exceptions.

Data quality: The degree to which data is accurate, complete, timely, consistent, valid and fit for purpose.

Digital maturity: The organisational capability to use digital technologies, data, processes and people effectively.

Digital twin: A dynamic digital representation of a physical asset, process or network used for monitoring and simulation.

Edge computing: Processing data near the location where it is generated rather than sending all data to a central system.

Emissions visibility: The ability to measure, estimate, allocate and analyse greenhouse-gas emissions across logistics activities.

Internet of Things: A network of connected physical objects that collect and exchange data.

Machine learning: Methods through which systems identify patterns in data and improve predictions or classifications.

Predictive maintenance: Maintenance based on predicted equipment condition or failure risk.

Smart contract: Software that automatically executes agreed actions when specified conditions are satisfied.

Telematics: The collection and communication of vehicle or asset data through telecommunications and information systems.

Traceability: The ability to follow a product, material, transaction or custody event through supply-chain stages.

Technology footprint: The environmental and social impacts associated with producing, operating, maintaining and disposing of digital technologies.

Total cost of ownership: The complete cost of acquiring, operating, maintaining, upgrading and retiring a system or asset.

Review Questions

Short Answer Questions

1. Distinguish between logistics visibility and logistics intelligence.

2. How can telematics contribute to fleet sustainability?

3. What is the purpose of a digital twin in supply-chain planning?

4. Why does blockchain not guarantee the truthfulness of supply-chain data?

5. Identify four dimensions of data quality relevant to logistics analytics.

Long Answer Questions

1. Explain how IoT, cloud platforms and control towers can work together to improve sustainable logistics.

2. Discuss the uses and limitations of AI in forecasting, routing and predictive maintenance.

3. Analyse the cybersecurity and privacy risks created by connected logistics operations.

4. Explain how algorithmic bias may arise in logistics systems and propose governance measures.

5. Evaluate the argument that digital transformation is not automatically sustainable.

Analytical / Application Questions

1. A carrier’s routing algorithm reduces kilometres but increases delivery failures and driver overtime. How should the objective function, constraints and performance measures be redesigned?

2. Design a data-governance and emissions-visibility plan for a logistics network using several third-party carriers with incompatible reporting systems.

3. A company is considering blockchain for product traceability. Apply the chapter’s decision framework to determine whether blockchain is justified or whether a simpler digital-record system would be preferable.

22. Multiple Choice Questions with Answers

Question 1

Which condition most directly converts IoT data into sustainability value?

A. Installing the largest possible number of sensors

B. Linking sensor information to defined decisions and response procedures

C. Storing all sensor data indefinitely

D. Replacing operational managers with automated dashboards

Correct Answer: B. Linking sensor information to defined decisions and response procedures

Question 2

What is the principal purpose of a logistics control tower?

A. To eliminate all human decisions from transport planning

B. To provide integrated visibility and coordinate responses to exceptions

C. To replace every supplier’s information system

D. To record financial transactions only

Correct Answer: B. To provide integrated visibility and coordinate responses to exceptions

Question 3

Why may the shortest vehicle route not be the lowest-emission route?

A. Distance has no relationship to fuel use

B. Emission outcomes also depend on factors such as congestion, gradient, payload and vehicle type

C. Longer routes always use less fuel

D. Routing algorithms cannot use vehicle information

Correct Answer: B. Emission outcomes also depend on factors such as congestion, gradient, payload and vehicle type

Question 4

Which statement best describes a limitation of blockchain-based traceability?

A. It cannot record transactions in sequence

B. It guarantees that every physical observation entered is accurate

C. It preserves recorded data but does not by itself guarantee that the original data is truthful

D. It can only be used for internal warehouse records

Correct Answer: C. It preserves recorded data but does not by itself guarantee that the original data is truthful

Question 5

A forecasting model performs well during normal periods but poorly during major disruptions. What is the most appropriate response?

A. Allow the model to operate without review

B. Discard all historical data permanently

C. Retain human override and test the model under unusual scenarios

D. Measure success only by the number of predictions produced

Correct Answer: C. Retain human override and test the model under unusual scenarios

Question 6

Which practice best supports responsible use of AI in logistics?

A. Preventing users from questioning model outputs

B. Using historical data without testing for representativeness

C. Monitoring accuracy, bias and model drift while maintaining accountable human oversight

D. Optimising only for transport cost

Correct Answer: C. Monitoring accuracy, bias and model drift while maintaining accountable human oversight

Question 7

What is the main distinction between predictive and preventive maintenance?

A. Predictive maintenance occurs only after failure

B. Preventive maintenance uses no schedule

C. Predictive maintenance uses condition or failure-risk information, whereas preventive maintenance commonly follows planned intervals

D. There is no operational distinction

Correct Answer: C. Predictive maintenance uses condition or failure-risk information, whereas preventive maintenance commonly follows planned intervals

Question 8

Which example illustrates poor data quality creating a sustainability problem?

A. A calibrated sensor reporting temperature at defined intervals

B. A duplicated shipment record causing an unnecessary dispatch

C. A documented emissions factor being reviewed periodically

D. A standardised carrier report using agreed definitions

Correct Answer: B. A duplicated shipment record causing an unnecessary dispatch

Question 9

Why should the technology footprint be included in a digital sustainability assessment?

A. Digital systems have no environmental impacts during operation

B. Hardware, connectivity, computing, replacement and disposal create material and energy impacts

C. Technology footprints are always larger than transport footprints

D. It makes cybersecurity unnecessary

Correct Answer: B. Hardware, connectivity, computing, replacement and disposal create material and energy impacts

Question 10

A company has a simple traceability problem involving authenticated custody transfers. What principle should guide its technology choice?

A. Use the most complex technology available

B. Select the least complex technology that reliably solves the defined problem

C. Adopt blockchain regardless of governance arrangements

D. Avoid documenting data ownership

Correct Answer: B. Select the least complex technology that reliably solves the defined problem