8. Smart Factories, Cyber-Physical Systems and Industrial IoT

Introduction and Context

Manufacturing has progressively moved from mechanization to automation, and from isolated automation to connected, intelligent production systems. The contemporary factory is no longer defined only by machines, operators, material-handling equipment and production schedules. It is increasingly characterized by continuous data exchange among physical assets, software platforms, people and supply-chain partners. This transformation is commonly associated with Industry 4.0.

Industry 4.0 does not simply mean installing sensors or replacing paper records with digital screens. It represents a change in the design and management of production systems. Machines acquire the ability to sense and communicate; software systems interpret operational data; production resources respond to changing conditions; and decisions can be made at the point where information is generated. The resulting production environment is often called a smart factory.

For production managers, the significance of smart manufacturing lies in improved visibility, flexibility, quality, asset utilization and responsiveness. However, digital connectivity also introduces new forms of complexity. A connected machine can become a cybersecurity vulnerability, a poorly designed dashboard can increase rather than reduce decision latency, and an isolated pilot project may fail to generate enterprise value. Smart manufacturing therefore requires both technological understanding and managerial judgement.

This chapter examines the architecture and application of smart factories, cyber-physical systems (CPS) and the Industrial Internet of Things (IIoT). It connects sensors, RFID, industrial networks, edge and cloud computing, analytics, interoperability and decentralized control to the practical concerns of production management.

Learning Objectives

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

1. Explain the meaning and evolution of Industry 4.0, smart factories, CPS and IIoT.

2. Describe the layers and information flows in a smart-factory architecture.

3. Distinguish sensors, actuators, RFID systems, PLCs, industrial networks, edge platforms and cloud platforms.

4. Evaluate the roles of interoperability, digital twins, analytics and decentralized control.

5. Select appropriate technologies for production visibility, quality, maintenance and material tracking.

6. Analyse managerial, technical, cybersecurity, ethical and organizational challenges.

7. Develop a structured approach for implementing connected manufacturing systems.

From Automated Production to Intelligent Production

The first industrial revolution used mechanical power, particularly steam, to transform production. The second introduced electrical power and large-scale production. The third incorporated electronics, programmable logic controllers, information technology and computer-based automation. The fourth extends automation through pervasive connectivity, cyber-physical integration, data analytics, artificial intelligence and distributed decision-making.

The distinction between the third and fourth industrial revolutions is not absolute. Many modern factories contain equipment from several generations. A new machine may communicate through an industrial Ethernet network, while an older press may provide only electrical signals or manually recorded readings. Industry 4.0 must therefore be understood as an integration challenge, not merely as a replacement programme.

A conventional automated system generally performs a programmed sequence. A smart production system additionally observes its condition, communicates with other systems, interprets context and adapts within defined limits. For example, an automated conveyor may move parts at a fixed speed. A connected conveyor can report motor temperature, detect abnormal vibration, identify the product being transported and adjust its operation in coordination with downstream equipment.

The value of intelligence arises from the relationship among data, models, decisions and action. Data without context may be unusable; analytics without operational authority may not produce improvement; and automation without governance may amplify errors. A smart factory is therefore a socio-technical system in which technology, processes, people and organizational rules operate together.

The Smart-Factory Architecture

A useful architecture begins with the physical layer. This includes machines, robots, tools, products, conveyors, energy systems and workers. Sensors measure variables such as temperature, pressure, position, speed, current, vibration, force, flow and image characteristics. Actuators convert decisions into physical action by controlling motors, valves, robots, heaters and positioning mechanisms.

The control layer contains devices such as programmable logic controllers (PLCs), distributed control systems, numerical controllers and robot controllers. These systems perform deterministic control tasks, often under strict timing requirements. A PLC may receive a proximity-sensor signal, execute a control programme and command a pneumatic cylinder within milliseconds. Control systems prioritize reliability and predictable response.

The connectivity layer transports data among equipment and information systems. It may include fieldbus systems, industrial Ethernet, wireless networks, gateways and application programming interfaces. Industrial communication differs from ordinary office networking because it must often satisfy requirements for low latency, synchronization, availability, electromagnetic robustness and safety.

The edge layer places computing capability near machines and production lines. Edge devices can filter data, perform local analytics, translate protocols, store information temporarily and trigger immediate responses. Edge processing is particularly useful when a system must operate despite unreliable connectivity to a remote cloud platform or when sending every raw signal to the cloud would create unnecessary cost and latency.

The platform layer stores, manages and contextualizes data. It may include manufacturing execution systems (MES), historians, asset-management systems, data lakes, industrial IoT platforms and digital-twin applications. Contextualization links a value such as “72” to its meaning: perhaps 72 degrees Celsius measured at a particular bearing, on a specified machine, during a particular production order.

The enterprise and application layer supports planning, coordination and decision-making. Enterprise resource planning (ERP), supply-chain management, quality management, maintenance management and business-intelligence systems operate at this level. The objective is not to move all control into enterprise software. Rather, the architecture should ensure that each decision is made at the appropriate level and that information can flow vertically and horizontally.

A practical architecture can therefore be represented as:

Physical assets and products → sensing and control → industrial connectivity → edge processing → platforms and data models → applications and management decisions.

Information also flows in the reverse direction. Production plans, quality limits and maintenance policies influence machine settings, work instructions and local actions. This bidirectional flow is what makes the system cyber-physical rather than merely data-collecting.

Core Concepts and Terminology

A cyber-physical system integrates computational elements with physical processes. Sensors observe the physical process, software models interpret observations, and actuators influence the process. Feedback is central: the system continuously compares actual conditions with desired conditions and adjusts its behaviour.

The Industrial Internet of Things refers to the connection of industrial assets, instruments, products and systems so that they can collect, exchange and use data. IIoT differs from consumer IoT in its emphasis on operational continuity, safety, long asset lifecycles, deterministic performance and integration with industrial control.

Interoperability is the ability of different systems to exchange information and use the exchanged information meaningfully. Syntactic interoperability concerns compatible data formats and communication protocols. Semantic interoperability concerns shared meaning. A machine may send a value successfully while the receiving system still misinterprets its unit, timestamp or equipment identity.

A digital twin is a digital representation of a physical asset, process or system that is updated with operational information. It may range from a simple condition model to a sophisticated simulation. A digital twin is not automatically a complete virtual replica; its usefulness depends on the quality of its data, assumptions and intended decision.

A digital thread is the connected flow of information across the lifecycle of a product or process. It can link design requirements, production parameters, inspection results, maintenance events and field performance. The digital thread supports traceability and learning across organizational boundaries.

An industrial gateway connects equipment using different protocols or data models. It may convert signals from legacy machines into a format usable by modern platforms. A historian is a specialized database that records time-series data from industrial processes. An edge device performs computing close to the data source, whereas cloud computing provides scalable remote storage and processing.

Manufacturing execution systems coordinate production activities between planning systems and shop-floor control. They may manage dispatching, work-in-process, genealogy, quality records, labour reporting and electronic instructions. RFID uses radio-frequency identification to identify and track tagged objects without requiring direct line-of-sight scanning. Radio-frequency identification is particularly valuable for reusable containers, work-in-process carriers, tools and high-value components.

Sensing, Identification and the Creation of Context

Sensors are the primary observation mechanism of a CPS. Their selection should begin with the decision to be supported, not with the availability of a fashionable device. A vibration sensor is justified when vibration information can improve failure detection, process diagnosis or product quality. Important selection criteria include measurement range, accuracy, sampling rate, environmental resistance, calibration requirements, installation method and lifecycle cost.

Sensor data becomes valuable when it is associated with context. A temperature reading needs an asset identity, location, time reference, unit of measure and operating state. Production managers should distinguish between raw data, information and actionable knowledge. Raw data may show motor current; information may show that current increased during a particular product family; knowledge may indicate that the increase predicts a tooling problem.

RFID and related automatic identification technologies support material visibility. A tag may identify a component, container, batch or work order. Readers installed at process points can create an event history: receipt, issue, machining, inspection, storage and dispatch. RFID reduces manual recording and can improve genealogy, but it does not eliminate the need for correct process design. Read failures, metal interference, tag damage and duplicate events must be managed.

Machine vision adds another sensing capability. Cameras and image-processing systems can inspect dimensions, surface defects, orientation, presence and assembly conditions. Vision systems are often connected to PLCs or edge computers so that a defective product can be diverted immediately. Where artificial intelligence is used for image classification, training-data quality, model drift and human review remain important.

Industrial Networks and Interoperability

Industrial networks may be categorized according to their role and performance requirements. Field-level networks connect sensors and actuators. Control-level networks connect PLCs, drives, robots and controllers. Supervisory networks carry information to monitoring and manufacturing systems. Enterprise networks connect production data with planning, finance, procurement and customer systems.

Common industrial communication technologies include fieldbus protocols, industrial Ethernet variants, wireless local networks, cellular technologies and specialized machine-to-machine protocols. The choice depends on latency, determinism, distance, mobility, environmental conditions, cybersecurity, installed base and required data volume.

Open standards can reduce dependence on a single supplier and support integration, but “open” does not necessarily mean simple. Interoperability requires common information models, naming conventions, time synchronization, authentication and clear ownership of interfaces. Technologies such as OPC UA are often used to expose structured industrial data, while message-oriented protocols such as MQTT can support efficient event transmission. The specific protocol is less important than the architectural discipline with which it is implemented.

Legacy equipment creates a significant challenge. Retrofitting may involve reading existing controller tags, installing external sensors, using protocol converters or capturing electrical signatures. A phased approach is usually preferable to replacing all equipment. Critical assets should be prioritized according to business impact, data availability, integration effort and expected improvement.

Edge, Cloud and Hybrid Computing

The edge-cloud relationship should be designed around operational requirements. Edge computing is appropriate for high-frequency data, rapid response, local autonomy, privacy-sensitive information and continued operation during network interruption. Cloud computing is useful for enterprise-wide aggregation, long-term storage, cross-site benchmarking, large-scale analytics and collaboration.

A hybrid architecture combines both. For example, an edge device may calculate a vibration feature every second and issue a local warning if a threshold is exceeded. The summarized features and event records may then be transmitted to a cloud platform for fleet-level analysis. Sending only relevant information reduces bandwidth and supports faster response.

Centralized cloud systems can provide economies of scale, but excessive centralization may create dependence on connectivity and increase exposure to a common failure. Conversely, completely decentralized systems may duplicate data and make enterprise analysis difficult. The appropriate design separates time-critical control from higher-level analysis. Safety and basic process control should not depend on an uncertain external network.

Data, Analytics and Decentralized Control

Smart manufacturing generates several types of analytics. Descriptive analytics explains what has happened, such as downtime by reason code. Diagnostic analytics investigates why it happened, perhaps by relating downtime to temperature, operator, material lot and tooling condition. Predictive analytics estimates what may happen, such as the probability of bearing failure. Prescriptive analytics recommends an action, such as rescheduling a job or changing a maintenance interval.

Condition-based maintenance illustrates this progression. Sensors monitor pressure, vibration and temperature. A model detects deviation from a normal operating pattern. The system generates a maintenance recommendation, but a planner evaluates production impact, spare-parts availability and safety requirements before authorizing work. Automation should support responsible decisions rather than conceal uncertainty.

Decentralized control distributes decisions among machines, cells and local software agents. A production cell may negotiate sequencing with neighbouring cells, or a workpiece may carry an electronic identity that informs machines of its required processing route. Decentralization can improve responsiveness and resilience, but it requires clear authority, conflict-resolution rules, synchronized data and safeguards against unstable behaviour.

The degree of autonomy should match the consequence of error. A system may automatically adjust a non-critical parameter within validated limits, while a safety-related or quality-critical decision may require human approval. This principle is especially important when machine-learning models are used in changing production environments.

Applications in Production Management

Real-time production visibility is one of the most immediate applications. Connected machines can report operating state, cycle time, micro-stoppages and output. Dashboards can display actual production against plan, but meaningful implementation requires consistent definitions of availability, performance, quality and downtime. Otherwise, apparent precision merely digitizes inconsistent reporting.

In quality management, process data can be linked to product genealogy. If a defect is identified, the organization can trace the relevant material lot, machine settings, tool, operator, inspection result and environmental conditions. Statistical models can identify combinations of variables associated with defects. Closed-loop quality systems may automatically stop a process or adjust a parameter when validated limits are exceeded.

In maintenance, connected assets support remote diagnosis, condition monitoring, work-order prioritization and spare-parts planning. Predictive maintenance is not appropriate for every asset. A simple preventive replacement may be economically superior when failure consequences are low and sensor deployment is costly. The managerial question is whether additional information improves the maintenance decision enough to justify its total cost.

Material and inventory control can benefit from RFID, barcode integration, location sensors and automated storage systems. Real-time location information reduces searching, supports line-side replenishment and improves work-in-process accuracy. However, visibility does not itself solve excess inventory or poor scheduling; it enables better decisions when linked to operating policies.

Energy management is another important application. Smart meters and machine-level sensors can reveal energy consumption by product, process, shift or operating state. Production schedules may be adjusted to reduce peak demand, while abnormal consumption can indicate leaks, inefficient equipment or idle running.

Implementation as a Managerial Programme

A successful smart-factory programme begins with a business problem. Common starting points include excessive unplanned downtime, poor traceability, long changeovers, quality escapes, high energy use or unreliable production reporting. The organization should define a baseline, identify decision owners and specify measurable outcomes before selecting technology.

The next step is an assessment of the current state. This includes asset criticality, data availability, network capability, cybersecurity exposure, system interfaces, workforce skills and process maturity. A value-versus-complexity matrix can help prioritize use cases. High-value, manageable applications should normally be implemented before ambitious plant-wide platforms.

Pilot projects should be designed as operational experiments rather than technology demonstrations. The pilot must include process owners, maintenance personnel, operators, information-technology specialists and cybersecurity representatives. Measures may include downtime reduction, improvement in first-pass yield, shorter response time, reduced manual reporting or improved schedule adherence.

Scaling requires standardization. Data naming, asset models, access rights, network segmentation, device configuration, maintenance responsibilities and supplier interfaces should be documented. Without standards, each pilot becomes a separate technical island. Governance should also specify who owns data, who may modify machine logic, how software is updated and how incidents are reported.

Workforce development is essential. Operators need to understand new alerts and electronic work instructions; maintenance technicians may require skills in networks, instrumentation and data interpretation; managers must learn to question data quality and model assumptions. Technology should augment professional judgement, not treat employees as obstacles to automation.

Illustrative Case: Connected Visibility in a Components Plant

Background

The following is an illustrative case based on a typical medium-sized components manufacturer. The plant produces several families of machined parts using computer numerical control machines, washing equipment and inspection stations. It has a modern manufacturing execution system for some lines, but several older machines record production through end-of-shift entries.

Situation and Problem

Management reports that planned production is frequently missed, although reported machine utilization appears high. Quality investigations also take several days because material lots, machine settings and inspection records are not consistently linked. Maintenance personnel believe that certain machines exhibit warning signs before failure, but those signs are not systematically captured.

Analysis

A cross-functional team maps the information flow and discovers three problems. First, “running” is defined differently in the production and maintenance systems. Second, older machines lack reliable cycle-completion signals. Third, inspection data is stored by batch while production data is stored by machine and shift, making genealogy difficult. The team concludes that a large cloud platform alone will not solve the problem.

Approach and Solution

The plant selects one production cell for a staged pilot. External sensors are installed on two critical machines to capture vibration, current and temperature. A gateway collects controller data and sensor readings at the edge. RFID tags identify reusable work-in-process carriers, and readers are installed at the cell entrance and inspection point. The edge application translates machine signals into common states such as production, setup, blocked, starved and fault.

An MES interface links carrier identity, work order, machine state and inspection result. Time-critical alarms remain local, while summarized events are transmitted to a central platform. A dashboard displays schedule attainment, actual cycle time, downtime categories and material location. Maintenance receives alerts only after a defined pattern is confirmed, and the planner decides whether the alert warrants immediate intervention.

Outcome

Within the pilot, the organization gains a consistent definition of machine state and can identify the location and processing history of each tagged carrier. Investigation of quality deviations becomes faster because machine and inspection records are connected. Maintenance personnel report that alerts are more useful when accompanied by operating context rather than by a raw sensor threshold.

The pilot also reveals limitations. RFID readers require repositioning near a metal enclosure, and some operators initially bypass the intended carrier route. The company therefore improves reader placement, revises standard work and introduces training rather than assuming that technical installation guarantees compliance.

Learning Points

The case demonstrates that value comes from combining sensing, identification, data modelling, process discipline and human response. It also shows why a local edge layer is useful for immediate operational information, while a central platform supports analysis across time. Most importantly, the project begins with production problems and definitions, not with a decision to purchase a particular technology.

Risks, Limitations and Governance

Connectivity increases the attack surface of a factory. Threats include unauthorized access, ransomware, manipulation of control logic, theft of production data, insecure remote maintenance and disruption of availability. Industrial cybersecurity should include asset inventories, network segmentation, least-privilege access, strong authentication, secure backups, patch management, monitoring and tested recovery procedures. Safety systems require special treatment because cybersecurity incidents can create physical harm.

Data quality is another limitation. Sensors may drift, timestamps may be inconsistent, machines may report false states and operators may enter incomplete reason codes. Analytics built on unreliable data can generate false alarms and erode trust. Data governance should define ownership, quality standards, retention, lineage and permitted uses.

Vendor lock-in may occur when proprietary platforms, interfaces or data formats make it difficult to change suppliers. Contracts should address data portability, interface documentation, cybersecurity responsibilities, software support, licensing and system exit. Long asset lifecycles also mean that today’s architecture should not unnecessarily restrict future integration.

Privacy and employment concerns must be addressed when systems monitor operators, location or performance. Data collected for safety or process improvement should not automatically be repurposed for punitive individual surveillance. Organizations should communicate purpose, limit access and comply with applicable labour, privacy and data-protection requirements.

Automation can redistribute rather than eliminate work. Some routine tasks may decline, while interpretation, exception management, system maintenance and cybersecurity responsibilities increase. Ethical implementation requires consultation, retraining and fair treatment of affected employees.

Emerging Directions

Industrial artificial intelligence is increasingly being applied to visual inspection, anomaly detection, process optimization and production scheduling. Its practical value depends on explainability, representative data and integration with work processes. Models must be monitored for drift because materials, tools, products and operating conditions change.

Industrial 5G and advanced wireless systems may support mobile robots, connected tools and flexible production layouts. Their adoption should be justified by mobility, capacity or performance requirements rather than by the label alone. Digital twins are also developing from asset-level monitoring toward process and factory-level simulation, enabling what-if analysis for capacity, layout, energy and resilience decisions.

Additive manufacturing, autonomous mobile robots, collaborative robots and machine vision are increasingly connected to broader production architectures. Edge AI may enable local inference without transmitting sensitive raw data. Sustainable manufacturing applications combine operational data with energy, emissions and resource information to support more informed trade-offs.

The emerging factory is therefore unlikely to be completely autonomous. More plausibly, it will combine automated local actions, human supervision, enterprise coordination and continuous learning. The central managerial challenge is to design an appropriate distribution of authority between people and machines.

Best-Practice Principles

A sound smart-factory programme follows several principles:

• Begin with measurable operational and business problems.
• Separate safety-critical and time-critical control from higher-level analytics.
• Establish common asset identities, data definitions and time standards.
• Use edge computing where latency, resilience or privacy requires local processing.
• Apply open interfaces and documented information models to improve interoperability.
• Treat cybersecurity as a lifecycle responsibility, not a final installation activity.
• Pilot with operators, maintenance staff, engineers and managers together.
• Validate analytics against operational outcomes and monitor model performance.
• Design for legacy integration and gradual scaling.
• Define data ownership, access rights, retention and ethical use.
• Measure value through improved decisions and outcomes, not merely through the number of connected devices.

Chapter Summary

Smart factories integrate physical production resources with sensing, communication, computation and decision-making. Industry 4.0 builds on earlier automation but extends it through cyber-physical systems, IIoT, industrial networks, edge and cloud computing, analytics, digital twins and decentralized control. A smart-factory architecture connects physical assets, control systems, networks, edge devices, platforms and enterprise applications through bidirectional information flows.

Sensors and RFID create operational visibility, but data becomes useful only when it is contextualized through asset identity, time, location and process meaning. Edge computing supports rapid local response and resilience, while cloud platforms support aggregation, long-term analysis and cross-site coordination. Interoperability requires compatible protocols as well as shared semantics and governance.

Applications include real-time production monitoring, quality traceability, condition-based maintenance, material tracking, energy management and adaptive scheduling. Effective implementation begins with a business problem, proceeds through a focused pilot and scales through standards, cybersecurity, workforce development and data governance. The principal risks include cyberattack, poor data quality, vendor dependence, privacy concerns and inappropriate automation. Smart manufacturing is ultimately a socio-technical transformation in which technology must be aligned with operational purpose, human capability and responsible management.

Key Terms

Cyber-physical system: A system that integrates computational processes with physical processes through sensing, communication, feedback and actuation.

Industrial Internet of Things: The connection of industrial assets, products, instruments and systems for collecting, exchanging and using operational data.

Smart factory: A production environment that uses connected technologies, analytics and intelligent control to improve visibility, responsiveness and performance.

Industry 4.0: The contemporary phase of industrial development characterized by cyber-physical integration, connectivity, automation and data-driven decision-making.

Sensor: A device that detects a physical condition and converts it into a measurable signal.

Actuator: A device that converts a control signal into physical action.

PLC: A programmable logic controller used for reliable, deterministic industrial control.

Industrial gateway: A device or software service that connects equipment using different protocols or data models.

Edge computing: Processing data close to the source or machine where it is generated.

Cloud computing: Remote, scalable computing and storage delivered through networked infrastructure.

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

RFID: Radio-frequency identification technology used to identify and track tagged objects wirelessly.

Digital twin: A digital representation of a physical asset, process or system that uses operational data for analysis or control.

Digital thread: A connected flow of product and process information across lifecycle stages.

Manufacturing execution system: A system that coordinates and records production activities between planning and shop-floor control.

Industrial historian: A database designed to store and retrieve time-series industrial data.

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

Decentralized control: Distribution of decision-making among local machines, cells or software agents.

Data governance: Policies and responsibilities governing data quality, ownership, access, security and use.

Machine vision: The use of cameras and image-processing techniques to inspect or guide industrial operations.

Review Questions

Short Answer Questions

1. What distinguishes a cyber-physical system from a conventional automated machine?

2. Explain the difference between edge computing and cloud computing in manufacturing.

3. Why is contextualization necessary for industrial sensor data?

4. What role does RFID play in production visibility?

5. Define interoperability and distinguish syntactic from semantic interoperability.

Long Answer Questions

1. Explain the layers of a smart-factory architecture and describe the flow of information between them.

2. Discuss how IIoT can support production, quality, maintenance and material-management decisions.

3. Examine the managerial process for implementing a smart-factory pilot and scaling it across a plant.

4. Analyse the advantages and limitations of decentralized control in manufacturing.

5. Discuss the cybersecurity, privacy, workforce and governance issues associated with connected factories.

Analytical / Application Questions

1. A plant has frequent machine stoppages but no reliable downtime records. Propose a phased IIoT solution, including sensors, edge processing, data definitions, users and performance measures.

2. A manufacturer is considering predictive maintenance for every machine. Develop a decision framework for determining which assets should be connected and which should retain conventional maintenance practices.

3. An RFID-based traceability project is producing duplicate and missing events. Diagnose possible technical and process causes and recommend corrective actions.

22. Multiple Choice Questions with Answers

Question 1

Which feature most clearly distinguishes a cyber-physical system from a conventional automated system?

A. It uses electrical power to operate machinery.

B. It combines sensing, computation, communication and physical feedback.

C. It replaces all human decision-making with robots.

D. It stores production records only after a shift ends.

Correct Answer: B. It combines sensing, computation, communication and physical feedback.

Question 2

A machine must issue a response within milliseconds even when the external network is unavailable. Which architecture is most appropriate?

A. Cloud-only processing

B. Manual data entry

C. Local edge processing with autonomous control capability

D. Periodic spreadsheet analysis

Correct Answer: C. Local edge processing with autonomous control capability

Question 3

Why is contextualization important when collecting machine temperature data?

A. It increases the physical temperature of the machine.

B. It identifies the asset, time, location, unit and operating condition associated with the value.

C. It removes the need for sensor calibration.

D. It guarantees that the data is accurate.

Correct Answer: B. It identifies the asset, time, location, unit and operating condition associated with the value.

Question 4

Which application best illustrates the use of RFID in a smart factory?

A. Calculating a motor’s electrical resistance without a sensor

B. Tracking reusable work-in-process carriers through production stages

C. Replacing all machine-control programmes

D. Increasing cloud-storage capacity

Correct Answer: B. Tracking reusable work-in-process carriers through production stages

Question 5

A system can exchange a value between two machines, but one interprets the value as Fahrenheit and the other as Celsius. What type of problem is primarily involved?

A. Semantic interoperability

B. Mechanical failure

C. Network bandwidth excess

D. Actuator calibration only

Correct Answer: A. Semantic interoperability

Question 6

Which analytics activity is predictive rather than descriptive?

A. Reporting yesterday’s downtime by machine

B. Displaying current production against schedule

C. Estimating the probability that a bearing will fail next week

D. Listing completed work orders

Correct Answer: C. Estimating the probability that a bearing will fail next week

Question 7

Why should time-critical machine control generally remain independent of a remote cloud platform?

A. Cloud platforms cannot store historical data.

B. Local control reduces dependence on network latency and connectivity.

C. Cloud systems cannot support any analytics.

D. Local control prevents all cybersecurity threats.

Correct Answer: B. Local control reduces dependence on network latency and connectivity.

Question 8

In the illustrative case, what enabled faster quality investigation?

A. Removing inspection records from the production system

B. Connecting carrier identity, machine history and inspection results

C. Increasing the number of end-of-shift reports

D. Sending all raw sensor data directly to operators

Correct Answer: B. Connecting carrier identity, machine history and inspection results

Question 9

Which practice most directly reduces the risk of a connected factory becoming a collection of isolated technology pilots?

A. Allowing every department to use different asset names

B. Avoiding documentation of supplier interfaces

C. Establishing common data models, standards and governance for scaling

D. Installing sensors without defining business outcomes

Correct Answer: C. Establishing common data models, standards and governance for scaling

Question 10

A company wants to use operator-location data collected for safety monitoring to rank individual employees for punishment. Which issue does this raise most directly?

A. Excessive actuator speed

B. Ethical and privacy concerns concerning secondary data use

C. Insufficient RFID reading range

D. Failure of deterministic control

Correct Answer: B. Ethical and privacy concerns concerning secondary data use