SabkaCode/Papers/Emerging Technologies for Business

MBA · Semester III · Examination 2025–26

Emerging Technologies for Business — Complete Question Paper Solution

Section-wise theory answers with clear definitions, detailed explanations, business examples, comparison tables and process flowcharts.

Section A

7 short answers · 14 marks

Section B

5 alternatives · attempt any 3

Section C

Questions 3–7 · both options

Section A · 14 marks

Short-answer questions

Original question labels 1(a)–1(g) are retained.

Section A · 1(a)2 marks · Exam-oriented answer

What is the Fourth Industrial Revolution?

The Fourth Industrial Revolution (Industry 4.0) is the current phase of industrial development in which digital technologies connect physical machines, software, data and people. It combines technologies such as artificial intelligence (AI), the Internet of Things (IoT), robotics, cloud computing, big-data analytics and cyber-physical systems to make operations more connected, automated and data-driven.

Main features

Its major features include intelligent automation, real-time data collection, machine-to-machine communication, predictive maintenance and integration across the supply chain. Unlike simple mechanization, Industry 4.0 enables systems to sense conditions, analyse data and adjust operations with limited human intervention.

Suitable example

A smart factory may use sensors to monitor machine vibration, AI to predict a possible failure and automated systems to schedule maintenance before production stops.

Section A · 1(b)2 marks · Exam-oriented answer

What is digital transformation?

Digital transformation is the strategic use of digital technologies to redesign an organization's processes, products, services and customer experiences. It is not merely converting paper documents into digital files; it may change how the business operates, makes decisions, serves customers and creates value.

Key elements

Common elements include cloud platforms, data analytics, automation, digital customer channels, redesigned workflows, employee skills and organizational change. Successful transformation requires clear business objectives, leadership support, cybersecurity, data governance and measurement of results.

Suitable example

A bank that introduces mobile account opening, digital identity checks, online support and automated loan assessment is transforming its customer journey rather than simply digitizing forms.

Section A · 1(c)2 marks · Exam-oriented answer

Define cloud computing.

Cloud computing is the on-demand delivery of computing resources—such as servers, storage, databases, networking and software—over a network, usually the internet. Users can provision resources when required and pay according to a subscription or usage model, depending on the service.

Benefits and models

Cloud computing can provide scalability, faster deployment, remote access and reduced need to purchase and maintain physical infrastructure. Main service models are Infrastructure as a Service (IaaS), Platform as a Service (PaaS) and Software as a Service (SaaS). Organizations must still manage access, configuration, cost, data protection and compliance.

Suitable example

Using an online email service or renting virtual servers from a cloud provider are examples of cloud computing.

Section A · 1(d)2 marks · Exam-oriented answer

Give an example of Human–Machine Interaction.

Human–Machine Interaction (HMI) refers to the ways in which a person communicates with, controls or receives information from a machine or computer system. It includes physical controls, graphical user interfaces, touchscreens, voice assistants, gestures and other interaction methods.

Purpose

A good HMI makes a system understandable, efficient, accessible and safe. Clear feedback, consistent controls, error prevention and suitable alerts are especially important in industrial and medical systems.

Suitable example

A factory operator using a touchscreen control panel to monitor machine temperature and stop a production line is interacting with a machine through an HMI.

Section A · 1(e)2 marks · Exam-oriented answer

What are programmable networks?

Programmable networks are networks whose behaviour, configuration and services can be controlled or automated through software instead of requiring every change to be made manually on individual devices. They often use centralized controllers, APIs, automation scripts and policy-based configuration.

Software-Defined Networking

Software-Defined Networking (SDN) is an important approach in which the control logic is separated from the data-forwarding devices, allowing a controller to manage network behaviour. Programmability can make networks more flexible and easier to configure, but APIs and controllers must be secured and changes carefully validated.

Suitable example

A data centre may use software-defined networking to automatically allocate network paths to applications according to traffic and security policies.

Section A · 1(f)2 marks · Exam-oriented answer

What are smart cities?

A smart city uses digital technologies, connected sensors, data analytics and communication networks to improve urban services, sustainability and residents' quality of life. It may integrate transport, energy, water, waste management, public safety and citizen services.

Characteristics

Smart-city systems can monitor traffic, manage street lighting, detect water leaks, optimize waste collection and provide online public services. They should be designed with accessibility, privacy, cybersecurity, reliability and inclusion in mind so that technology benefits the whole community.

Suitable example

Adaptive traffic signals that use vehicle-flow sensors to adjust signal timing are one example of a smart-city application.

Section A · 1(g)2 marks · Exam-oriented answer

What is data privacy?

Data privacy concerns the proper collection, use, sharing, retention and deletion of personal information. It addresses what data is collected, why it is collected, who can access it and what rights individuals have over its use.

Core principles

Important principles include transparency, purpose limitation, data minimization, accuracy, security, limited retention and appropriate consent or other lawful grounds. Privacy is related to but different from security: security protects data from unauthorized access or alteration, while privacy governs appropriate handling of personal data.

Suitable example

A shopping app should explain why it needs a customer's location, request permission where required, protect the data and avoid using it for unrelated purposes without a valid basis.

Section B · 21 marks

Detailed theory answers

All five options are included; the paper asks you to attempt any three.

Section B · 2(a)7 marks · Exam-oriented answer

What are the three stages in a Data Pipeline? Explain.

A data pipeline is a sequence of steps that moves data from one or more sources to a destination where it can be stored, analysed or used by applications. In a basic model, its three stages are data ingestion, data processing and data storage or delivery. Some architectures split transformation and serving into separate stages, but the core objective remains to move reliable data from source to useful destination.

1. Data ingestion

Data ingestion collects data from sources such as databases, business applications, IoT sensors, log files, websites and external APIs. It may run in batch mode at scheduled intervals or in streaming mode as events arrive. The ingestion stage should validate source connectivity, record timestamps, handle duplicate events and preserve the necessary metadata.

2. Data processing and transformation

The raw data is cleaned, validated, standardized, joined, filtered or aggregated. Processing can correct formats, remove duplicates, handle missing values, convert units and apply business rules. Data quality checks are important because incorrect or inconsistent data can produce misleading reports and models. Processing may occur before storage (ETL) or after loading into a destination (ELT).

3. Data storage and delivery

Processed data is stored or delivered to a destination such as a database, data warehouse, data lake, dashboard or machine-learning system. The destination should support the intended query patterns, access controls, retention rules and performance requirements. Monitoring and lineage help users understand where data came from and whether a pipeline completed successfully.

Operational considerations

A production pipeline also needs scheduling or orchestration, retries, error handling, logging, alerting, access control and recovery procedures. These capabilities are not always described as separate stages, but they are essential for dependable operation.

Flowchart / Process

Data sources

1. Ingestion: collect and validate

2. Processing: clean and transform

3. Storage / delivery: warehouse, lake or dashboard

Monitoring, quality checks and feedback

Suitable example

A retailer collects point-of-sale transactions every hour, cleans product codes and timestamps, aggregates sales by store, and loads the result into a warehouse for sales dashboards and demand forecasting.

Conclusion

Ingestion acquires data, processing turns it into a consistent form, and storage or delivery makes it available for use. Quality, monitoring and security should be applied throughout the pipeline.

Section B · 2(b)7 marks · Exam-oriented answer

What is Big Data? What is the difference between Big Data and Database?

Big Data refers to datasets whose size, speed, diversity or complexity makes them difficult to manage efficiently using traditional data-processing approaches alone. Big Data technologies distribute storage and computation across multiple machines and support analysis of large structured, semi-structured and unstructured datasets. A database, by contrast, is an organized data collection managed by a database management system; databases can be small or very large.

Characteristics of Big Data

Big Data is often explained through the 5 Vs. Volume refers to the amount of data; Velocity to the speed at which it is generated and processed; Variety to formats such as tables, text, images, video and sensor events; Veracity to quality and reliability; and Value to useful outcomes extracted from data. Not every Big Data project needs all five characteristics to be equally strong.

Traditional databases

Relational databases organize information into tables with schemas, keys and relationships. They are widely used for transactional applications that require consistent updates and queries. Modern databases can scale substantially, so it is inaccurate to say that databases are always small or that Big Data replaces them.

Big Data platforms

Big Data systems may use distributed file storage, parallel processing engines, streaming platforms, data lakes and cloud analytics. They are useful when data volume, velocity or variety exceeds the practical limits of a single-machine approach or when distributed analytics is required.

Basis	Database	Big Data
Meaning	Organized collection managed by a DBMS	Data and methods for handling challenging scale, speed or diversity
Typical data	Often structured; may include other formats	Structured, semi-structured and unstructured
Processing	Queries and transactions; may be distributed	Often distributed batch, streaming or parallel analytics
Examples	Banking transaction database	Clickstream, IoT sensor and social-media analytics

Suitable example

A bank may use a relational database to record account balances and transfers, while a Big Data platform analyses billions of click events and device signals to identify usage patterns or detect unusual behaviour.

Conclusion

A database is a way of organizing and managing data, whereas Big Data describes data challenges and the technologies and methods used to address them. They can work together in one information architecture.

Section B · 2(c)7 marks · Exam-oriented answer

Describe cloud computing service models. What is AWS?

Cloud computing delivers computing resources over a network on demand. Its three widely used service models—Infrastructure as a Service, Platform as a Service and Software as a Service—differ in how much of the technology stack is managed by the cloud provider and how much remains the customer's responsibility.

1. Infrastructure as a Service (IaaS)

IaaS provides fundamental computing resources such as virtual machines, storage and networking. The customer usually manages the operating system, middleware, applications, data and many security configurations, while the provider manages the physical infrastructure and virtualization layer. It offers flexibility but requires technical administration.

2. Platform as a Service (PaaS)

PaaS provides a managed environment for building, testing and deploying applications. The provider manages much of the underlying infrastructure and runtime platform, while the customer focuses mainly on application code and data. It can speed up development, although platform constraints and vendor dependence should be considered.

3. Software as a Service (SaaS)

SaaS delivers a complete application through a browser, app or API. The provider operates the application and underlying platform; customers configure settings, manage users and control their data and access. Examples include hosted email, online office applications and cloud-based customer relationship management systems.

What is AWS?

Amazon Web Services (AWS) is a cloud-computing platform that offers services for compute, storage, databases, networking, analytics, artificial intelligence, security and application deployment. Examples include Amazon EC2 for virtual compute, Amazon S3 for object storage, Amazon RDS for managed relational databases and AWS Lambda for event-driven serverless functions. The exact service chosen depends on the workload and architecture.

Shared responsibility

Cloud providers secure the underlying cloud infrastructure, while customers remain responsible for aspects such as identity permissions, data classification, secure configuration and application security, depending on the service model. Cloud use does not automatically guarantee compliance, low cost or security.

Model	Provider typically manages	Customer typically manages
IaaS	Physical infrastructure and virtualization	OS, applications, data and configuration
PaaS	Infrastructure, OS/runtime and platform	Application code, data and access
SaaS	Complete application and platform	Users, settings, access and appropriate data handling

Flowchart / Process

Choose workload and requirements

Select IaaS, PaaS or SaaS

Configure identity, network and data security

Deploy and monitor service

Review cost, performance and compliance

Suitable example

A company may run a custom application on EC2 (IaaS), deploy code on a managed application platform (PaaS-style service), and use a hosted email product (SaaS).

Conclusion

IaaS offers the most infrastructure control, PaaS reduces platform administration, and SaaS provides a ready-to-use application. AWS is a major provider offering services across these and other cloud categories.

Section B · 2(d)7 marks · Exam-oriented answer

How does Data Science help in Decision Making?

Data science combines statistics, programming, domain knowledge and analytical methods to extract insights from data. It supports decision-making by describing what has happened, diagnosing why it happened, predicting what may happen and recommending actions. It improves evidence-based decisions but does not remove uncertainty or the need for human judgment.

1. Data collection and preparation

Data scientists combine relevant information from transactions, sensors, customer interactions, surveys and external sources. They clean errors, handle missing values, standardize formats and assess bias. Reliable preparation is essential because poor-quality input can undermine even a sophisticated model.

2. Descriptive and diagnostic analysis

Descriptive analysis summarizes trends through totals, averages, distributions and dashboards. Diagnostic analysis investigates possible causes by comparing groups, examining correlations and testing hypotheses. Correlation alone does not establish causation, so conclusions must be interpreted carefully.

3. Predictive analysis

Statistical and machine-learning models can forecast sales, estimate customer churn, predict equipment failure or identify fraud risk. Predictions should be evaluated on suitable data and monitored for changes in real-world performance. A prediction is a probability or estimate, not a guarantee.

4. Prescriptive analysis

Optimization, simulation and scenario analysis can compare possible actions under constraints such as cost, capacity, time or risk. For example, a logistics business may compare delivery routes based on distance, traffic and delivery deadlines. Decision-makers should understand the assumptions and trade-offs used by the model.

5. Measurement and continuous improvement

After implementing a decision, teams compare actual results with objectives and use the findings to improve future decisions. Dashboards and controlled experiments can help assess whether an intervention produced the intended outcome. Privacy, fairness, explainability and accountability should be considered when decisions affect people.

Flowchart / Process

Define decision question

Collect and prepare data

Analyse patterns and causes

Build forecasts or compare scenarios

Make informed decision

Measure outcomes and improve

Suitable example

A retailer analyses past sales, weather and holiday periods to forecast demand. The purchasing manager uses the forecast to set stock levels, then compares predicted and actual sales to improve the next forecast.

Conclusion

Data science helps organizations make more timely, measurable and evidence-based decisions through descriptive, diagnostic, predictive and prescriptive analysis. Its value depends on reliable data, appropriate methods and responsible interpretation.

Section B · 2(e)7 marks · Exam-oriented answer

What is an IoT device and how does it work?

An Internet of Things (IoT) device is a physical object containing sensors or actuators, processing capability and network connectivity that allows it to collect data, exchange information or perform actions. IoT devices connect the physical world with digital applications and can support monitoring, automation and remote control.

Main components

A typical IoT device may include sensors to measure temperature, movement, pressure or location; a microcontroller or processor to interpret readings; connectivity such as Wi-Fi, Bluetooth, cellular or a low-power network; firmware; and a power source. Some devices also contain actuators that perform actions such as opening a valve or switching a motor.

Working process

The sensor collects a measurement, and the device processes or packages it. Data is sent directly to a gateway or cloud platform over a network. Software stores and analyses the data, displays it to a user or applies a rule. If the system supports control, a command can be sent back to an actuator. The device may operate locally when connectivity is unavailable, depending on its design.

Benefits and risks

IoT enables remote monitoring, automation, predictive maintenance and better resource use. Risks include insecure default passwords, weak updates, exposed data, unreliable connectivity and physical tampering. Secure provisioning, encryption, authentication, updates and limited data collection are important.

Flowchart / Process

Physical sensor measures condition

Device processes reading

Network or gateway transmits data

Platform stores and analyses data

Dashboard or rule generates insight

Optional command activates actuator

Suitable example

A smart thermostat measures room temperature, sends readings to a home controller, compares them with the desired temperature and switches heating or cooling equipment on or off.

Conclusion

An IoT device senses or affects the physical environment and exchanges data with connected systems. Its usefulness depends on reliable connectivity, suitable analytics, safe control and strong security.

Section C · Questions 3–7

Long-answer questions

Original question numbering and both alternatives are preserved for Questions 3 through 7.

Section C · 3(a)7 marks · Exam-oriented answer

What is Artificial Intelligence? What are the competitive advantages of using AI in business?

Artificial Intelligence (AI) is a field of computing concerned with building systems that perform tasks commonly associated with human intelligence, such as understanding language, recognizing patterns, making predictions, planning and supporting decisions. Modern business AI includes machine learning, natural-language processing, computer vision, recommendation systems and generative AI. AI capabilities vary by system; they should not be assumed to possess human understanding or to be correct in every situation.

1. Automation and productivity

AI can automate repetitive information-processing tasks such as document classification, invoice extraction, support-ticket routing and report preparation. This can reduce processing time and allow employees to focus on complex work. Processes still need oversight, exception handling and quality checks.

2. Better forecasting and decisions

Predictive models can analyse historical and current data to estimate demand, identify churn risk, detect unusual transactions or forecast equipment failures. Businesses can use these estimates to plan inventory, staffing, maintenance and marketing. Model performance should be tested and monitored as conditions change.

3. Customer experience and personalization

Recommendation engines can suggest relevant products, while chatbots can answer routine questions at any time. Language tools can help classify feedback and summarize customer interactions. Poorly designed automation may frustrate customers, so escalation to a human should be available for complex or sensitive cases.

4. Cost control and operational efficiency

AI can optimize routes, improve energy use, detect manufacturing defects and support predictive maintenance. Reduced waste and downtime may lower operating costs. However, total cost includes data preparation, infrastructure, licensing, integration, monitoring and specialist skills.

5. Innovation and competitive differentiation

Businesses can use AI to develop new digital services, improve product features and discover patterns that are difficult to detect manually. Early, responsible adoption may create differentiation, but competitors can adopt similar tools; sustainable advantage often depends on proprietary data, process integration, expertise and customer trust rather than the model alone.

6. Risk management and fraud detection

AI systems can flag unusual transaction patterns, suspicious logins, product defects or compliance anomalies for review. These systems should be evaluated for false positives, false negatives and unequal impact. High-impact decisions should have appropriate human review and an appeal or correction process.

7. Requirements and limitations

Successful AI adoption needs a clear business objective, representative data, reliable infrastructure, employee training, privacy and security controls, governance and measurable performance indicators. Risks include bias, hallucinations, data leakage, explainability gaps, model drift and over-reliance on automated outputs. AI should be selected where it creates demonstrable value and managed throughout its life cycle.

Flowchart / Process

Identify a business problem

Prepare and govern relevant data

Select and validate an AI method

Integrate with the business process

Monitor accuracy, fairness, cost and security

Improve the system with human oversight

Suitable example

An online retailer can use AI to forecast product demand and recommend items. Better forecasting may reduce stockouts, while recommendations can improve product discovery. The retailer must still protect customer data, evaluate recommendation quality and avoid misleading or discriminatory outcomes.

Conclusion

AI can provide competitive advantages through productivity, forecasting, personalization, efficiency, innovation and risk detection. These advantages are sustainable only when the system is reliable, responsibly governed and connected to real business needs.

Section C · 3(b)7 marks · Exam-oriented answer

What do you understand by Machine Learning?

Machine Learning (ML) is a branch of AI in which computer systems learn patterns from data to make predictions, classifications, recommendations or decisions without requiring every rule to be explicitly programmed. A learning algorithm uses examples to build a model, and the model is evaluated on data to determine whether it generalizes to new cases.

1. Supervised learning

Supervised learning uses labelled examples, where the correct output is known during training. Classification predicts categories, such as whether a transaction is potentially fraudulent; regression predicts numerical values, such as expected sales. The model is evaluated on data that was not used for training to estimate performance on new cases.

2. Unsupervised learning

Unsupervised learning works with data that does not have target labels. Clustering can group customers by similar behaviour, while dimensionality reduction can summarize complex data. The patterns found must be interpreted and validated because an algorithm may identify groups that are not useful for the business question.

3. Reinforcement learning

Reinforcement learning trains an agent to choose actions in an environment using rewards or penalties. It can be applied to certain control, scheduling and optimization problems. Training needs careful design because a poorly specified reward can encourage undesirable behaviour.

4. ML life cycle

The process starts by defining the problem and collecting relevant data. The data is cleaned and divided into training, validation and test sets. A model is selected and trained, its performance is measured using suitable metrics, and it is deployed only when it meets requirements. Monitoring is needed to detect data drift, performance decline and unexpected outcomes.

5. Advantages and limitations

ML can discover complex patterns, automate predictions and improve as relevant data and methods improve. Limitations include data requirements, bias, overfitting, interpretability challenges, computation cost and changing real-world conditions. A model can perform well on training data but poorly on unseen data if it memorizes examples rather than learning general patterns.

Flowchart / Process

Define problem and success metric

Collect, clean and label data where needed

Split data and choose algorithm

Train model and tune parameters

Evaluate on unseen data

Deploy, monitor and retrain when justified

Suitable example

A bank may train a supervised classification model on historical transactions labelled as legitimate or fraudulent. The model scores new transactions for review. It must be evaluated for false alarms and missed fraud, and its decisions should be monitored as fraud patterns change.

Conclusion

Machine learning enables systems to learn predictive patterns from data. Appropriate data, sound evaluation, monitoring and responsible use are essential for dependable business applications.

Section C · 4(a)7 marks · Exam-oriented answer

What is Blockchain? How many types of Blockchain are there?

Blockchain is a type of distributed ledger in which records are grouped into blocks linked cryptographically and maintained by a network according to a consensus mechanism. Participants can verify the ledger according to the network's rules. Depending on the design, blockchain can make unauthorized alteration more difficult, but it does not automatically guarantee that input data is true, that applications are secure or that every transaction is private.

How it works

Transactions are submitted to the network, checked against its rules and grouped into a proposed block. A consensus mechanism determines whether the block is accepted. The block references the previous block using a cryptographic hash, creating a linked history. Different blockchains use different consensus methods and finality rules.

1. Public blockchain

A public blockchain is generally open for anyone to read and, subject to protocol rules, participate in validation or submit transactions. It supports broad transparency and open participation, but may face scalability, transaction-cost, privacy and governance challenges. Public access does not mean every piece of data is confidential.

2. Private blockchain

A private blockchain restricts participation and often validation to an organization or authorized administrators. It may offer greater control, performance and confidentiality, but trust depends more heavily on the organization managing access and governance. A conventional database may be simpler when there is no genuine need for a shared ledger.

3. Consortium or federated blockchain

A consortium blockchain is governed by a group of selected organizations rather than one organization alone or the entire public. It may be useful when several known participants need a shared record and agreed rules. Governance must define membership, validation, dispute resolution and data access.

4. Hybrid blockchain

A hybrid blockchain combines restricted access to some information with selected public verification or transparency. An organization may keep sensitive business data private while publishing a proof or selected record to a public network. The exact balance depends on architecture and privacy requirements.

Applications and limitations

Potential uses include supply-chain traceability, shared asset records, cross-organization reconciliation and tokenized assets. Limitations include performance, integration, governance, energy use for some consensus designs, legal uncertainty and the challenge of correcting inaccurate off-chain data.

Type	Access	Typical setting
Public	Open participation under protocol rules	Open networks and public verification
Private	Controlled by one organization	Internal or restricted enterprise ledger
Consortium	Authorized group of organizations	Industry collaboration
Hybrid	Mix of restricted and public elements	Sensitive records with selective verification

Flowchart / Process

Transaction proposed

Network participants validate rules

Consensus accepts block

Block links to prior block cryptographically

Ledger replicas update

Suitable example

Several food distributors may use a permissioned consortium ledger to share shipment events. Each participant can verify the recorded history, while access rules restrict sensitive commercial information.

Conclusion

The four commonly taught blockchain types are public, private, consortium and hybrid. The correct choice depends on trust, governance, privacy, performance and whether a distributed ledger provides a genuine benefit.

Section C · 4(b)7 marks · Exam-oriented answer

What are smart contracts on blockchain? Elaborate.

A smart contract is program code deployed on a blockchain that executes according to predefined rules when it receives a valid transaction or trigger. It can record state changes, transfer digital assets or enforce conditions within the blockchain's capabilities. A smart contract is not automatically a legally enforceable contract; legal status depends on jurisdiction, wording and circumstances.

Working mechanism

Developers write contract logic, test it and deploy it to a blockchain address. Users submit transactions that call its functions. Network nodes execute the code according to protocol rules, and valid state changes are recorded on the ledger. Because deployed code can be difficult to change, upgrade and emergency procedures must be designed carefully.

Main characteristics

Smart contracts can provide deterministic execution, transparent rules and an auditable history. They may reduce some manual reconciliation and allow multiple participants to rely on the same programmed conditions. However, transparency can expose sensitive data if the design is careless, and execution depends on the blockchain's security and availability.

Oracles and external information

A blockchain contract usually cannot independently verify facts outside its network, such as a real-world delivery or weather event. An oracle supplies external data. If an oracle is inaccurate, compromised or unavailable, the contract may act on incorrect information. Oracle design and fallback procedures are therefore important.

Applications

Possible applications include automated escrow, supply-chain milestones, token issuance, digital identity permissions, insurance processes and decentralized finance. The business process must be suitable for automation and all participants must understand the rules and consequences.

Risks and controls

Programming defects, access-control mistakes, compromised keys, oracle failures and unexpected interactions can cause losses. Controls include code review, independent security audits, test networks, formal verification where appropriate, limited privileges, monitoring and a carefully designed pause or upgrade mechanism. Legal, privacy and consumer-protection requirements still apply.

Flowchart / Process

Define rules and participants

Develop and test contract code

Deploy contract to blockchain

Authorized transaction triggers function

Network validates and executes rules

Result is recorded and monitored

Suitable example

In a simplified escrow process, a buyer deposits funds into a smart contract. After a trusted delivery confirmation is supplied, the contract releases funds to the seller. The reliability of the delivery confirmation and the contract's code are critical to the outcome.

Conclusion

Smart contracts automate agreed digital rules on a blockchain and can improve traceability and reduce manual processing. Their use requires secure code, reliable external data, sound governance and legal review.

Section C · 5(a)7 marks · Exam-oriented answer

What are the principles of 3D printing? What are the uses of 3D printing?

Three-dimensional (3D) printing, also called additive manufacturing, creates a physical object by adding material in successive layers based on a digital 3D model. Unlike subtractive manufacturing, which removes material from a larger piece, additive manufacturing builds the part layer by layer. The exact process depends on the material and printing technology.

Principles and process

First, a digital model is created or scanned using computer-aided design software. The model is checked and converted into a printable file, then slicing software divides it into layers and generates machine instructions. The printer deposits, cures, fuses or binds material according to the selected process. After printing, the object may need support removal, cleaning, curing, machining or inspection.

Major technologies

Fused Deposition Modelling (FDM) extrudes thermoplastic filament through a heated nozzle. Stereolithography (SLA) cures liquid resin with light. Selective Laser Sintering (SLS) uses a laser to fuse powder particles, while metal powder-bed fusion processes can build metal components. These methods differ in materials, accuracy, strength, cost and finishing requirements.

Uses in healthcare

3D printing can produce anatomical models for surgical planning, customized prosthetics, dental aligners, hearing-aid shells and selected medical-device components. Medical use requires appropriate validation, material safety and regulatory compliance; a printed model is not automatically suitable for implantation or clinical use.

Uses in manufacturing and business

Manufacturers use 3D printing for rapid prototypes, jigs, fixtures, spare parts, low-volume customized products and complex geometries that may be difficult to produce conventionally. Aerospace and automotive firms may use it for lightweight components where the material and certification requirements permit.

Education, architecture and consumer uses

Schools and universities use printed models to teach engineering and design. Architects can create scale models, while designers can test shape, fit and ergonomics. Consumers and small businesses can make customized cases, ornaments, models and replacement parts where permitted.

Advantages and limitations

Advantages include rapid iteration, customization, reduced tooling needs and the ability to create complex shapes. Limitations include material and machine costs, slower production for large volumes, surface finishing, build-size limits, anisotropic strength, quality control and intellectual-property concerns. Conventional manufacturing may remain more economical for mass production.

Flowchart / Process

Create or scan 3D digital model

Check model and slice into layers

Set material and printer parameters

Build object layer by layer

Remove supports and finish part

Inspect dimensions and quality

Suitable example

A product-design team can print a prototype enclosure, check whether the circuit board fits, modify the digital design and print a revised version before paying for expensive production tooling.

Conclusion

3D printing converts a digital model into a physical object through additive layer construction. It is valuable for prototyping, customization and complex or low-volume parts, but process selection and quality validation are essential.

Section C · 5(b)7 marks · Exam-oriented answer

What is Neuromorphic Computing? What are its advantages?

Neuromorphic computing is a computing approach inspired by the structure and operating principles of biological nervous systems. It uses hardware and algorithms designed around neuron-like processing elements and synapse-like connections, often with event-driven communication. It is different from conventional computing architectures that typically separate processing and memory more distinctly.

How it works

Many neuromorphic systems use spiking neural networks in which information is represented by timed events or spikes. Processing can occur when events arrive rather than continuously performing every calculation. Specialized chips may place memory and computation close together to reduce data movement and energy use for suitable workloads.

1. Energy efficiency

Event-driven processing can reduce unnecessary computation when inputs are sparse. This can be valuable for battery-powered sensors, robotics and edge devices that must operate within strict power budgets. Efficiency depends on the workload and hardware; it is not guaranteed to outperform conventional processors for every task.

2. Low-latency processing

Neuromorphic systems may respond rapidly to changing sensor events because processing can be distributed and event-driven. This can help applications such as motion detection, adaptive robotics and real-time perception where timely responses matter.

3. Parallel and adaptive computation

Neuron-like units can operate in parallel, and some neuromorphic learning approaches adapt connection strengths based on experience or activity. This can support pattern recognition and learning in dynamic environments, although training methods and development tools are still less mature than those for mainstream AI in many settings.

4. Edge intelligence

Neuromorphic hardware may perform selected inference tasks locally, reducing the need to send every sensor reading to a cloud server. Local processing can improve privacy and reduce communication requirements, provided the system is designed and secured appropriately.

Limitations and applications

Potential applications include robotics, event-based vision, always-on sensing and low-power pattern recognition. Limitations include specialized hardware, limited software ecosystems, difficult model development, benchmarking challenges and compatibility with conventional AI workflows. It is an emerging field rather than a universal replacement for CPUs or GPUs.

Flowchart / Process

Sensors produce events

Neuromorphic hardware receives spikes

Neuron-like units process events in parallel

Network produces recognition or control output

System responds locally or shares selected results

Suitable example

An event-based vision sensor on a small robot could feed motion events to a neuromorphic processor, allowing the robot to react to movement with low latency and potentially low energy consumption.

Conclusion

Neuromorphic computing draws inspiration from neural systems to support event-driven, parallel and potentially energy-efficient processing. Its benefits are most promising in suitable real-time and edge workloads, while tools and applications continue to evolve.

Section C · 6(a)7 marks · Exam-oriented answer

Differentiate among Virtual Reality, Augmented Reality and Mixed Reality.

Virtual Reality (VR), Augmented Reality (AR) and Mixed Reality (MR) are immersive technologies that change how users experience digital content. They differ mainly in how much of the physical world the user sees and how digital objects interact with that world. Terminology can vary across products, but the following distinctions are commonly used.

Virtual Reality (VR)

VR places the user inside a computer-generated environment, usually through a head-mounted display that blocks most of the physical surroundings. Users can look around and sometimes interact with virtual objects through controllers, hand tracking or other inputs. VR is used for training simulations, virtual tours, gaming and design reviews.

Augmented Reality (AR)

AR overlays digital information—such as labels, arrows, images or 3D objects—onto a view of the real world. It may use a phone camera or smart glasses. AR commonly adds information to the physical environment without necessarily allowing digital objects to interact deeply with real surfaces.

Mixed Reality (MR)

MR combines physical and digital environments so that digital objects can be spatially anchored and interact with the user's understanding of the real space. For example, a virtual object may appear attached to a real table and remain in place as the user moves. MR generally requires more detailed spatial mapping and environmental awareness than simple overlays.

Business applications

VR can train employees in dangerous or expensive scenarios without exposing them to the real hazard. AR can guide a technician through repair steps by displaying instructions over equipment. MR can allow engineers to inspect a spatially anchored 3D model in a real workshop and collaborate around it.

Selection considerations

Selection depends on the task, required immersion, device cost, field of view, user comfort, tracking accuracy, physical safety, privacy and available content. Immersive systems should be tested with actual users and environments rather than chosen only for novelty.

Feature	VR	AR	MR
Physical world	Mostly replaced by virtual scene	Visible with digital overlay	Visible and spatially integrated with digital objects
Typical device	VR headset	Phone, tablet or smart glasses	Spatial headset or MR-capable glasses
Example	Flight simulator	Phone showing furniture in a room	Virtual machine model anchored beside real equipment

Suitable example

A trainee using a fully simulated factory is experiencing VR; a phone showing labels over a real machine is AR; and a digital 3D machine model anchored to the real floor and responding to spatial surfaces is an MR-style experience.

Conclusion

VR immerses users in a virtual world, AR adds digital content to the real world, and MR more deeply integrates digital objects with physical space. Each serves different training, design and operational needs.

Section C · 6(b)7 marks · Exam-oriented answer

What are sensors in IoT? Give examples.

An IoT sensor is a component that detects a physical condition or event and converts it into a signal that an electronic system can measure, process and communicate. Sensors allow connected devices to observe the physical world and provide data for monitoring, analysis and automated control.

Types of sensors

Temperature sensors measure heat; humidity sensors measure moisture in air; pressure sensors measure gas or liquid pressure; proximity sensors detect nearby objects; motion and accelerometer sensors detect movement or acceleration; light sensors measure illumination; gas sensors detect selected gases; and GPS receivers estimate location. Cameras and microphones can also act as sensing components, but they raise additional privacy considerations.

How sensors work in an IoT system

A sensor measures a condition and produces an electrical or digital signal. A microcontroller may convert, filter or calibrate the reading and attach a timestamp. The device sends the data through Wi-Fi, Bluetooth, cellular or a gateway to an application or cloud platform. Software then displays trends, triggers alerts or activates an actuator based on rules.

Uses in business and daily life

Temperature and humidity sensors monitor cold-storage facilities; vibration sensors support predictive maintenance; soil-moisture sensors help automate irrigation; occupancy sensors help manage lighting; air-quality sensors monitor pollution; and wearable sensors can track selected body measurements. Sensor accuracy, calibration, placement and operating conditions affect data quality.

Security and limitations

Sensor systems may produce noisy or incomplete readings, and devices can be physically damaged or tampered with. Secure device identity, encrypted communication, signed updates, access controls and data validation help protect the system. Data collection should be limited to what is necessary, especially where sensors could reveal personal behaviour or location.

Flowchart / Process

Physical condition changes

Sensor detects and measures it

Controller validates and timestamps reading

Network sends data to application

System displays insight or triggers alert

Optional actuator performs response

Suitable example

A cold-chain warehouse uses temperature sensors in refrigerators. If a reading rises above a permitted threshold, the IoT platform sends an alert to staff so that they can inspect the equipment and protect stored goods.

Conclusion

IoT sensors provide the measurements that make connected monitoring and automation possible. Correct selection, calibration, secure connectivity and sensible alert rules are necessary for reliable results.

Section C · 7(a)7 marks · Exam-oriented answer

What are cybersecurity threats in cloud computing?

Cloud-computing cybersecurity threats are risks that can compromise the confidentiality, integrity or availability of cloud-hosted applications, infrastructure or data. Cloud environments can improve resilience and security capabilities, but misconfiguration, weak identity controls, vulnerable software and misunderstandings about shared responsibility can expose resources.

1. Misconfiguration and exposed storage

Incorrect permissions, public storage, open management ports and overly broad network rules can expose sensitive resources. Organizations should use secure defaults, configuration scanning, least-privilege access and regular reviews. Infrastructure-as-code and policy checks can help maintain consistent settings.

2. Account hijacking and credential theft

Phishing, leaked API keys, reused passwords and stolen session tokens can allow attackers to access cloud resources. Multi-factor authentication, short-lived credentials, secret management, restricted administrator access and suspicious-login monitoring reduce this risk.

3. Data breaches and data loss

Unauthorized access, accidental deletion, ransomware, weak encryption or improper sharing can expose or destroy information. Data classification, encryption, access logging, versioning and protected backups are important. Recovery procedures should be tested regularly.

4. Insecure APIs and application vulnerabilities

Cloud services are frequently managed through APIs. Weak authentication, excessive permissions, missing input validation and unpatched software can allow attackers to misuse them. Secure development, API gateways where appropriate, rate limiting, dependency updates and testing help reduce exposure.

5. Denial of service and service disruption

Distributed denial-of-service attacks, resource exhaustion, provider outages or configuration errors can make services unavailable. Rate limits, traffic filtering, autoscaling where suitable, multi-zone architecture and incident plans can improve resilience. Autoscaling must be configured carefully to avoid unexpected cost.

6. Insider, supply-chain and shared-responsibility risks

Employees or contractors may misuse legitimate access, and compromised software dependencies or vendors may introduce malicious changes. Organizations should review third-party access, audit privileged activity, verify software sources and understand which security tasks belong to the provider and which remain their own.

Prevention and response

A cloud security programme should inventory assets, assess risks, enforce least privilege, encrypt sensitive data, patch systems, monitor logs, protect backups, test recovery and rehearse incident response. Security responsibilities must be assigned clearly and reviewed as services or architectures change.

Flowchart / Process

Identify cloud assets and threat exposure

Apply least privilege and secure configuration

Protect data, APIs and workloads

Monitor logs and detect suspicious activity

Contain incident and recover safely

Review root cause and strengthen controls

Suitable example

If a storage bucket is accidentally made public, confidential files may be exposed. Preventive controls include private-by-default settings, automated policy checks, access reviews, encryption and alerts for unexpected public access.

Conclusion

Cloud threats include misconfiguration, credential theft, data exposure, insecure APIs, denial of service, insider misuse and supply-chain compromise. Effective defence combines provider capabilities with the customer's own configuration, identity, data and incident-management controls.

Section C · 7(b)7 marks · Exam-oriented answer

What is AI bias and ethics? Describe in brief.

AI bias occurs when an AI system produces systematically unfair or skewed outcomes for particular individuals or groups. Bias may arise from unrepresentative training data, historical discrimination, flawed labels, unsuitable objectives, measurement errors or the way a system is deployed. AI ethics concerns the principles used to develop and use AI responsibly, including fairness, transparency, accountability, privacy, safety and respect for human rights.

Sources of AI bias

Historical data may reflect past inequality; some groups may be underrepresented; labels may encode subjective judgments; and a model may use proxy variables that correlate with protected or sensitive characteristics. Bias can also emerge when a model is applied to a population or environment different from the one in which it was evaluated.

Fairness and non-discrimination

Organizations should examine performance across relevant groups, assess the consequences of false positives and false negatives, and choose fairness criteria suitable to the context. There is no single metric that resolves every fairness trade-off. High-impact uses such as hiring, lending or healthcare require especially careful validation and oversight.

Transparency and explainability

Users and affected people should receive appropriate information about when AI is used and how its outputs influence decisions. Explanations should be meaningful for the audience and should not claim certainty that the model cannot provide. Documentation of data, model limitations and evaluation results supports accountability.

Privacy and data governance

AI systems should use data lawfully and proportionately, protect sensitive information and avoid unnecessary collection. Data access, retention, sharing and model-training practices should be governed. Personal information should not be exposed through prompts, outputs or model logs.

Accountability, safety and human oversight

An organization should assign responsibility for the system, monitor performance after deployment, provide routes for correction or appeal and intervene when harm is detected. Human review is particularly important for high-impact decisions, although reviewers need sufficient information, authority and time to challenge automated outputs.

Practical ethical AI life cycle

Ethical AI begins by assessing whether AI is appropriate for the problem. Teams then evaluate data quality and representativeness, test model performance and fairness, document limitations, protect the system, monitor outcomes and respond to complaints or incidents. Independent review may be appropriate where the risks are significant.

Flowchart / Process

Define purpose and assess potential harm

Review data quality and representation

Test accuracy, fairness, privacy and safety

Document limits and assign accountability

Deploy with appropriate human oversight

Monitor outcomes and correct problems

Suitable example

A hiring model trained mainly on historical applications from one demographic group may unfairly rank other qualified applicants lower. The employer should investigate the training data, test group-level outcomes, review job-related criteria and retain meaningful human oversight rather than relying blindly on model scores.

Conclusion

Responsible AI requires active attention to bias, fairness, transparency, privacy, safety and accountability. Ethical review is an ongoing process throughout design, deployment and monitoring, not a one-time approval.

Exam-writing checklist

Begin with a definition, then explain the main points under clear headings.
Use a suitable business or technology example in long answers.
Draw labelled flowcharts where they explain a process, architecture or lifecycle.
Conclude with the business value, limitations or implementation considerations.