The Wider Emerging Technology Landscape

Introduction

Generative artificial intelligence (AI) is not an isolated business technology. Its value increasingly depends on the wider technology environment in which data are collected, processed, verified, visualised and acted upon. Cloud computing supplies scalable infrastructure; edge computing reduces latency; the Internet of Things (IoT) connects physical objects; blockchain can support shared records and programmable transactions; digital twins model real assets; extended reality (XR) changes human interaction; robotics converts digital decisions into physical action; and quantum computing may eventually transform selected optimisation, simulation and cryptographic tasks.

For business analysts, the central question is not whether a technology is fashionable. It is whether a particular combination of technologies solves a measurable business problem better than feasible alternatives. This requires assessing value, data requirements, integration complexity, risk, skills, governance, security and organisational readiness.

Learning Objectives

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

1. Explain the principles and business roles of cloud and edge computing.
2. Analyse how IoT generates operational and customer data.
3. Evaluate blockchain, digital twins, XR, robotics and quantum computing for appropriate use cases.
4. Explain how these technologies converge with AI and generative AI.
5. Compare opportunities, limitations, risks and governance requirements.
6. Recommend an emerging-technology architecture for a business problem.

1. A Systems View of Emerging Technologies

A useful analytical model separates an emerging-technology solution into five layers:

The layers are interdependent. A predictive-maintenance model is valuable only if sensors collect reliable data, connectivity is sufficiently available, the model is deployed where a timely decision can be made, and maintenance staff can act on the recommendation. Similarly, a blockchain does not make inaccurate data true: it can preserve a record after entry, but it cannot automatically verify whether the original input was honest or correct.

A business analyst should begin with the decision or process, not the technology. Important diagnostic questions include:

• What outcome is to be improved: cost, quality, speed, safety, revenue or resilience?
• What data are required, and who owns them?
• Does the use case require real-time response?
• Are multiple parties that do not fully trust one another sharing records?
• Is the environment physical, digital or both?
• What level of automation is acceptable?

2. Cloud Computing and Edge Computing

Cloud computing

Cloud computing provides on-demand access to computing resources over a network. Its principal characteristics are resource pooling, elasticity, measured usage, broad network access and self-service provisioning. Instead of purchasing all servers in advance, an organisation can obtain infrastructure, platforms or software as a service.

The major service models are:

• Infrastructure as a Service (IaaS): virtual machines, storage and networks.
• Platform as a Service (PaaS): managed tools for developing and deploying applications.
• Software as a Service (SaaS): complete applications accessed by users.
• Serverless computing: the provider manages server capacity while customers execute functions in response to events.

Deployment may be public, private, hybrid or multi-cloud. Cloud economics can improve agility and convert capital expenditure into usage-related operating expenditure, but cloud is not automatically cheaper. Data transfer, storage growth, licensing, idle resources, security controls and migration costs must be included in total cost of ownership.

A simple annual cost model is:

Here, each term is an annual cost in the chosen currency. The model is a planning identity rather than a universal pricing formula. It assumes that relevant costs can be estimated and that one period, such as one year, is being compared consistently.

Cloud platforms are especially useful for enterprise data lakes, model training, business intelligence, collaboration, disaster recovery and applications with variable demand. Generative AI often uses cloud-based foundation models because training requires substantial computing capacity and diverse data. However, organisations must address data residency, confidential information, model-provider dependency and service availability.

Edge computing

Edge computing places storage or computation close to where data are produced or consumed. An edge device may be a sensor gateway, industrial computer, vehicle computer, retail appliance or local server. Edge computing does not eliminate the cloud; it creates a distributed architecture in which some work occurs locally and other work occurs centrally.

Edge is appropriate when latency, connectivity, privacy or bandwidth is important. Examples include robotic collision avoidance, factory quality inspection, vehicle control and health monitoring. A local model can identify an event immediately, while the cloud aggregates data for longer-term analysis and model improvement.

The trade-off is operational complexity. Edge devices may have limited processing power, difficult physical access, heterogeneous operating systems and weaker security controls. Model updates, device identity, monitoring and patch management become essential.

Cloud–edge convergence with AI

A typical architecture uses:

1. Sensors and operational systems to generate data.
2. Edge analytics to filter, detect or respond quickly.
3. Cloud platforms to store historical data and train models.
4. A deployment pipeline to send approved models back to edge devices.
5. Human oversight and monitoring to identify drift, errors and unsafe actions.

This arrangement is often called distributed or hybrid AI. It should be selected because it satisfies business constraints, not because it is technologically fashionable.

3. The Internet of Things

The IoT is a network of physical objects with sensors, software, processing capability and connectivity that enables them to collect and exchange data. Industrial IoT (IIoT) applies these ideas to factories, energy systems, logistics, mining and infrastructure.

An IoT system commonly contains:

• Sensors, which measure variables such as temperature, vibration, location or pressure.
• Actuators, which change physical conditions, such as opening a valve or controlling a motor.
• Connectivity, including wired networks, Wi-Fi, cellular, low-power wide-area networks and industrial protocols.
• Gateways, which translate protocols, aggregate data and perform local processing.
• Platforms, which manage devices, data and applications.
• Analytics and AI, which identify patterns, predict events and recommend actions.

IoT data are often high-volume, time-stamped and streaming. Data quality problems include sensor drift, missing observations, duplicated messages, inconsistent units and insecure devices. A data dictionary should define the measurement, unit, frequency, precision, source and acceptable range.

One common use case is predictive maintenance. Instead of servicing equipment only on a calendar schedule or waiting for failure, a model estimates the likelihood of failure within a specified horizon. The business value may arise from reduced downtime, lower spare-parts inventory and improved worker safety. The analysis must also consider false alarms, missed failures, maintenance capacity and the cost of sensor deployment.

IoT introduces an expanded attack surface. Weak passwords, unpatched firmware, insecure application programming interfaces and poorly controlled remote access can allow an attacker to move from a device into enterprise systems. Security should therefore include unique device identity, encryption, least-privilege access, secure boot, patching, network segmentation and retirement procedures.

4. Blockchain and Distributed Ledger Technology

A blockchain is a type of distributed ledger in which transactions are grouped into blocks and linked using cryptographic techniques. Copies of the ledger may be maintained by multiple participants. Consensus rules determine how the network accepts additions, while digital signatures help authenticate transaction origin.

Important distinctions include:

• Permissionless networks: participation and validation are broadly open.
• Permissioned networks: identified organisations control participation.
• Smart contracts: software that automatically executes agreed rules when conditions are met.
• Tokenisation: representing an asset, right or claim digitally, subject to legal and operational validity.

Blockchain is potentially useful when multiple parties need a shared, tamper-evident record but do not want one participant to control the database. Examples include provenance in food and pharmaceuticals, trade documentation, inter-organisational settlements and selected identity applications.

It is usually unnecessary when one trusted organisation already operates an adequate database. Blockchain may add latency, governance complexity, privacy challenges and integration costs. Immutability can also be a disadvantage when personal data must be corrected or deleted. A practical design often stores sensitive or large data off-chain and records only a hash or reference on-chain.

AI and blockchain can complement one another. Blockchain can record model versions, approvals, data-access events or transactions, while AI can detect fraud, forecast demand or interpret documents. Yet blockchain does not guarantee the quality of AI training data, and AI does not remove the need for governance of participants and smart-contract logic.

5. Digital Twins

A digital twin is a digital representation of a physical object, process, system or environment that is connected to its real-world counterpart through data. The connection may support monitoring, simulation, prediction or control. A static computer-aided design file is not necessarily a digital twin because it may lack live data and operational linkage.

Digital twins range from a component twin, such as a motor, to a system twin, such as a production line, and an ecosystem twin, such as a city or supply network. Typical components are:

• a physical entity;
• sensors and operational data;
• a digital model;
• a data and integration layer; and
• applications for visualisation, simulation and decision-making.

AI can estimate unobserved conditions, detect anomalies, forecast performance and optimise operating parameters. Generative AI can provide a natural-language interface to engineering records, explain simulated scenarios or draft maintenance procedures. Such outputs must remain grounded in authoritative data and reviewed where safety is involved.

A digital twin is valuable when decisions are repeated, consequences are material and the system can be observed sufficiently. It is less suitable when the asset is poorly instrumented, the model cannot be validated or the cost of maintaining the twin exceeds the benefit.

6. Extended Reality

Extended reality includes virtual reality (VR), augmented reality (AR) and mixed reality (MR).

• VR places the user in a largely simulated environment.
• AR overlays digital information on the physical environment.
• MR allows digital and physical objects to interact in a spatially integrated environment.

Business applications include employee training, remote assistance, design review, medical education, retail visualisation, tourism and collaborative product development. AR can display a technician’s repair instructions while leaving the physical machine visible. VR can simulate hazardous or expensive conditions without exposing trainees to actual danger.

AI strengthens XR through speech interfaces, object recognition, gesture interpretation, adaptive tutoring and generative creation of environments. XR also generates sensitive data, including body movement, gaze, voice and spatial maps. Organisations should address consent, biometric privacy, accessibility, motion sickness, workplace monitoring and intellectual-property protection.

A sound business case should compare XR with ordinary video, manuals, simulation software or instructor-led training. XR is justified when spatial understanding, embodied practice or realistic interaction creates measurable additional value.

7. Robotics and Autonomous Systems

Robotics combines sensing, computation, control and mechanical action. Industrial robots perform structured, repeatable tasks; collaborative robots are designed to work near people under controlled safety conditions; mobile robots navigate changing environments; and service robots interact with customers or workers.

Robotic systems typically include perception, planning, control, actuation and safety monitoring. AI supports visual inspection, object recognition, navigation, language interaction and adaptation to variation. Generative AI may translate natural-language instructions into candidate workflows, but execution should be constrained by permissions, testing and deterministic safety rules.

Automation does not always mean replacing workers. It may remove dangerous, repetitive or ergonomically harmful tasks while shifting employees toward supervision, exception handling, maintenance and customer-facing work. Organisations should analyse job redesign, training, accountability and worker participation.

Key performance measures include throughput, cycle time, availability, defect rate, safety incidents, human intervention frequency and total cost. A robot that operates quickly but requires frequent manual recovery may not improve the process.

8. Quantum Computing

Quantum computing uses quantum-mechanical properties such as superposition and entanglement to process information with quantum bits, or qubits. Quantum algorithms do not simply make every computation faster. Advantage is expected only for particular problem structures and sufficiently capable hardware.

Potential business areas include molecular simulation, portfolio and logistics optimisation, machine learning subroutines and cryptographic analysis. Current systems face noise, limited scale, error correction challenges, specialised programming requirements and uncertain performance advantages. Many proposed applications can currently be solved more reliably using classical high-performance computing or heuristics.

Businesses should distinguish three stages:

1. Awareness: identify problems with potential quantum relevance and monitor developments.
2. Experimentation: use cloud-accessible quantum services, proofs of concept and hybrid quantum–classical methods.
3. Deployment: adopt a solution only when validated performance, cost, security and reliability exceed classical alternatives.

Quantum developments also create a cryptographic issue. A sufficiently capable future quantum computer could threaten some widely used public-key cryptographic schemes. Organisations should inventory cryptographic dependencies and plan migration toward post-quantum cryptography according to recognised standards and risk assessments. This is a long-term resilience activity, not evidence that current commercial quantum computers can break all existing encryption.

9. Convergence Patterns

The most important innovation frequently occurs at the intersection of technologies.

Convergence increases both capability and risk. A connected autonomous system may combine physical safety risk, cybersecurity risk, privacy risk and algorithmic bias. Governance must therefore cover the whole system rather than each component separately.

10. Evaluation Framework for Business Analysts

A structured evaluation can use six dimensions:

1. Strategic fit: Does the technology support a clear organisational priority?
2. Economic value: Are benefits measurable and greater than total lifecycle costs?
3. Technical feasibility: Are data, connectivity, skills and integration available?
4. Operational fit: Can employees use, supervise and maintain the system?
5. Risk and compliance: Are privacy, safety, security, resilience and legal duties addressed?
6. Scalability: Can a successful pilot operate reliably across locations and processes?

A simple net-benefit expression is:

When comparing projects over several years, analysts should discount future cash flows. For a cash flow CF_t received at time t, discount rate r, and planning horizon T:

The formula assumes a constant discount rate and cash flows measured consistently. Non-financial outcomes—safety, resilience, service quality and environmental impact—should be assessed alongside financial value rather than forced into unreliable monetary estimates.

A responsible pilot should define a baseline, success metrics, a limited scope, user involvement, security controls, fallback procedures and an exit criterion. Pilots should not be designed merely to demonstrate that a technology can function; they should test whether it improves a business outcome.

Illustrative Case: Smart Cold-Chain Distribution

Background

A fictional pharmaceutical distributor operates refrigerated warehouses and vehicles. Temperature excursions can damage products, while manual inspection provides incomplete visibility.

Problem

The organisation needs earlier warning of equipment failure, continuous shipment visibility and auditable evidence that products remained within permitted conditions.

Analysis

Sensors measure temperature, humidity, location and door openings. Vehicle connectivity is intermittent, so immediate detection cannot depend entirely on the cloud. Multiple parties—manufacturer, carrier, distributor and pharmacy—need trustworthy event records, but personal patient data should not be placed on a shared ledger.

Approach and solution

Edge gateways analyse temperature locally and issue alerts when readings exceed validated thresholds. The cloud stores time-series data, trains an anomaly-detection model and provides dashboards. A digital twin represents each shipment and refrigeration unit. Generative AI summarises incidents from approved records and drafts an investigation report, subject to quality review. A permissioned ledger records selected custody events and hashes of compliance documents, while detailed sensor data remain in controlled storage.

Outcome

In the pilot, the distributor measures excursion frequency, warning lead time, false-alert rate, response time, product loss and operating cost against a pre-pilot baseline. The organisation expands only if the system improves response without creating unmanageable alert volume or compromising data protection.

Learning points

The case illustrates that no single technology solves the problem. Edge computing provides timely response, IoT supplies observations, cloud computing supports enterprise analytics, the digital twin organises the operational context, blockchain addresses selective inter-organisational evidence, and AI supports detection and communication. Governance determines whether the combined system is trustworthy.

Challenges, Ethics and Governance

Emerging technologies raise several cross-cutting concerns:

• Privacy: connected devices and XR may collect location, voice, biometric or behavioural data.
• Security: more devices and interfaces create more attack paths.
• Safety: autonomous decisions can cause physical harm if models fail.
• Bias and exclusion: AI systems may perform unequally across groups or environments.
• Transparency: users need understandable explanations, audit trails and appeal routes.
• Work and skills: automation can improve jobs but may also displace or intensify work.
• Environmental impact: data centres, sensors, batteries and hardware supply chains consume resources.
• Interoperability: proprietary formats can create vendor lock-in.
• Accountability: contracts must specify responsibility for errors, downtime, updates and data misuse.

Best practice includes privacy and security by design, data minimisation, strong identity management, human oversight for high-impact decisions, model and device monitoring, documented change control, independent testing, accessibility review and incident-response planning. Organisations should maintain an inventory of deployed systems and periodically reassess whether the original purpose, risks and controls remain valid.

Contemporary Direction

Several developments are shaping the landscape. AI is moving toward multimodal systems that combine text, images, audio, sensor streams and spatial information. Smaller models are increasingly deployed at the edge to reduce latency and data movement. Digital twins are becoming more operational as organisations connect engineering, maintenance and supply-chain data. Robotics is becoming more adaptable through improved perception and language interfaces. Standards for responsible AI, cybersecurity, digital identity and post-quantum cryptography are becoming strategically important.

The likely competitive advantage will not come from adopting every technology. It will come from integrating selected capabilities into reliable processes, building proprietary data assets, developing workforce skills and governing technology responsibly.

Chapter Summary

Cloud computing provides elastic, shared digital infrastructure, while edge computing places processing near data sources when latency, connectivity or privacy requires it. IoT connects physical objects and creates operational data, but it also increases security and data-quality responsibilities. Blockchain offers a shared, tamper-evident record for selected multi-party situations; it is not a replacement for every database and does not validate inaccurate inputs. Digital twins connect physical assets with digital models for monitoring, simulation and optimisation. XR supports spatial training, visualisation and collaboration. Robotics converts sensing and intelligence into physical action, while quantum computing remains an emerging capability with potential in selected problem classes and important cryptographic implications.

These technologies create the greatest value when deliberately combined with AI. Business analysts should evaluate strategic fit, measurable value, feasibility, operational adoption, risk and scalability. Responsible implementation requires privacy, cybersecurity, safety, interoperability, human oversight and clear accountability. The appropriate question is therefore not “Which emerging technology should we buy?” but “Which technology-enabled system best improves this decision or process under real organisational constraints?”

Key Terms

Cloud computing: On-demand network access to shared computing resources.

Edge computing: Processing and storage located near the source or user of data.

Internet of Things: Connected physical objects that sense, communicate and sometimes act.

Industrial IoT: IoT applied to industrial operations, equipment and infrastructure.

Gateway: A device that connects networks, translates protocols or aggregates data.

Digital twin: A connected digital representation of a physical entity or process.

Blockchain: A cryptographically linked distributed ledger of recorded transactions.

Distributed ledger: A shared record maintained across multiple network participants.

Smart contract: Software that executes programmed rules under specified conditions.

Extended reality: The combined field of VR, AR and MR technologies.

Virtual reality: Immersive computer-generated environments.

Augmented reality: Digital information overlaid on the physical environment.

Robotics: The design and use of machines capable of sensing, computing and acting.

Autonomous system: A system that performs tasks with limited direct human control.

Quantum computing: Computing based on quantum information principles and qubits.

Latency: The time delay between an event, its processing and a response.

Interoperability: The ability of systems to exchange and use information effectively.

Predictive maintenance: Maintenance informed by predicted equipment condition or failure.

Post-quantum cryptography: Cryptographic methods designed to resist future quantum attacks.

Technology convergence: The integration of multiple technologies into one capability.

Review Questions

Short Answer Questions

1. Distinguish cloud computing from edge computing.
2. What are the principal components of an IoT system?
3. Why is blockchain not required for every shared database?
4. Define a digital twin and explain how it differs from a static design model.
5. Identify two governance risks associated with robotics or autonomous systems.

Long Answer Questions

1. Explain how cloud, edge, IoT and AI can be combined in predictive maintenance.
2. Compare blockchain and conventional databases for inter-organisational record sharing.
3. Evaluate the business benefits and limitations of extended reality.
4. Discuss the current opportunities and constraints of quantum computing.
5. Develop a framework for evaluating an emerging-technology investment.

Analytical / Application Questions

1. A retailer wants real-time shelf monitoring across stores. Recommend an architecture using IoT, edge and cloud computing, and justify where each function should occur.
2. A manufacturer proposes a digital twin, robotics and generative AI system. Identify the data, safety, workforce and governance controls required before deployment.
3. An international supply chain has competing firms that need shared provenance records. Determine whether a permissioned blockchain is justified, and compare it with a conventional shared database.

Multiple Choice Questions with Answers

Question 1

Which situation most strongly justifies edge computing?

A. A quarterly financial report with no latency requirement

B. Real-time collision avoidance for an autonomous vehicle

C. Long-term archival storage of documents

D. Annual employee survey analysis

Correct Answer: B. Real-time collision avoidance for an autonomous vehicle

Question 2

Which statement about blockchain is most accurate?

A. It guarantees that all input data are truthful

B. It is always cheaper than a conventional database

C. It can provide a tamper-evident shared record among participants

D. It eliminates the need for governance

Correct Answer: C. It can provide a tamper-evident shared record among participants

Question 3

What distinguishes a digital twin from a static CAD model?

A. A digital twin must be printed in three dimensions

B. A digital twin is connected to its physical counterpart through operational data

C. A digital twin cannot use simulation

D. A digital twin is used only in construction

Correct Answer: B. A digital twin is connected to its physical counterpart through operational data

Question 4

Which is an example of augmented reality?

A. A fully simulated training environment viewed through a headset

B. A digital repair overlay displayed while a technician sees the real machine

C. A paper maintenance manual

D. A blockchain transaction record

Correct Answer: B. A digital repair overlay displayed while a technician sees the real machine

Question 5

What is a major security concern created by large-scale IoT deployment?

A. The complete elimination of network connectivity

B. A larger attack surface caused by many connected devices

C. The impossibility of collecting sensor data

D. The removal of all software dependencies

Correct Answer: B. A larger attack surface caused by many connected devices

Question 6

Which statement best describes current quantum computing?

A. It makes every business calculation faster today

B. It has no relationship to cryptography

C. It may offer advantages for selected problem structures but faces major hardware limitations

D. It replaces cloud and edge computing

Correct Answer: C. It may offer advantages for selected problem structures but faces major hardware limitations

Question 7

In the smart cold-chain case, why was detailed sensor data kept off-chain?

A. Because sensor data cannot be digitally stored

B. To reduce privacy, scalability and storage concerns on the shared ledger

C. Because blockchain cannot record hashes

D. To prevent all parties from seeing any event

Correct Answer: B. To reduce privacy, scalability and storage concerns on the shared ledger

Question 8

Which measure is most directly relevant to evaluating a predictive-maintenance pilot?

A. Number of technology press releases

B. Warning lead time and false-alert rate

C. Office colour scheme

D. Number of unused software licences

Correct Answer: B. Warning lead time and false-alert rate

Question 9

What is the principal role of a gateway in many IoT architectures?

A. To replace every sensor

B. To translate protocols, aggregate data and support local processing

C. To guarantee perfect predictions

D. To remove the need for cybersecurity

Correct Answer: B. To translate protocols, aggregate data and support local processing

Question 10

Which practice best supports responsible deployment of AI-enabled robotics?

A. Allowing unrestricted model-generated commands

B. Removing human oversight from safety-critical actions

C. Defining task boundaries, testing outputs and maintaining safety controls

D. Ignoring worker participation

Correct Answer: C. Defining task boundaries, testing outputs and maintaining safety controls