BCSE313L– Fundamentals of Fog and Edge Computing
MODULE - 1
Internet of Things (IoT) and New
Computing Paradigms
Introduction - Relevant Technologies - Fog and Edge Computing
Completing the Cloud - Hierarchy of Fog and Edge Computing -
Business Models – Edge Computing Platforms - Opportunities and
Challenges
Internet of Things and New Computing Paradigms
2
IoT
• The Internet of Things (IoT) minimizes the need
for human interaction in machine operation.
• Applied IoT links devices, vehicles, and
appliances over the internet.
• Data is collected from linked devices and shared
without the need for human interaction.
• Analyzing the data collected, the operation or
functionality can be decided by the device itself.
3
Enabling Technologies in IoT
1. Connectivity - Internet connectivity is necessary and essential to establish
an IoT platform. Network protocols for the internet make it easy for
connected things to transfer data efficiently.
2. Low-cost, low-power sensor technology - Sensor technology collects
data. Advancements in sensor technology make it more reliable and
affordable.
3. Cloud computing - Cloud computing helps businesses scale without the
need to physically manage the infrastructure for expansion. Data collected
from sensors is stored in the cloud server for predictive analysis.
4. Machine learning and data analytics - New machine learning techniques
and data analytics provide insights from the data collected and stored in
the cloud.
5. Artificial intelligence - Natural language processing (NLP) used in IoT is
enabled by developments in neural networks. Artificial intelligence plays a
role in establishing viable IoT applications.
4
Cloud Centric IoT
• It enables offering sensing resources and instruments as
a service through cloud platforms based on the pay as
you go fashion.
• Common IoT system involves three major technologies:
– Embedded systems
– Middleware, and
– Cloud services
• Embedded systems provide intelligence to the front-end
devices,
• Middleware interconnects the heterogeneous embedded
systems of front-end devices to the cloud and finally,
• the Cloud provides comprehensive storage, processing,
and management mechanisms.
5
Cloud Centric IoT (CIOT) – Layered Architecture
6
Cloud Centric IoT Architecture
7
CIOT - Connectivity
• Tiers
• Connectivity
• Brokers
8
CIOT
• Definition
• Cloud-Centric IoT is an IoT architecture where
the cloud acts as the central processing,
storage, analytics, and coordination hub,
while devices focus only on sensing and
transmitting data.
• Why Cloud-Centric?
• Unlimited storage & compute
• Scalable data processing
• Easy integration with AI/ML models
• Centralized management & security
• Global access anytime, anywhere
9
CIOT
• Main Components
• IoT Devices / Sensors
• Gateway / Edge Forwarder
• Cloud Platform (Core Hub)
• Application Services
10
Layered Components
• Device Layer (Sensing Layer)
• Sensors, actuators, RFID, cameras
• Collect real-time environmental data
• Lightweight processing only
• 2. Network/Communication Layer
• WiFi, 4G/5G, LoRaWAN, ZigBee
• MQTT/HTTP/CoAP protocols
• Transmits data securely to cloud
11
Layered Components
• 3. Cloud Layer (Core of CCIoT)
• Big Data storage
• Data processing pipelines
• AI/ML analytics
• API gateway, device management
• Security & identity management
• 4. Application Layer
• User dashboards
• Smart city apps, healthcare apps, industrial apps
• Event notifications & automation rules
12
Scenarios
• Scenario 1: Smart Healthcare
• Wearables → Mobile → Cloud
• Cloud performs anomaly detection (ECG, SpO₂)
• Doctors monitor via web dashboard
• Scenario 2: Smart Home Automation
• Sensors send data to AWS/Azure IoT Core
• Cloud triggers alerts (fire, intrusion, energy
misuse)
• User controls devices remotely
13
Contd.
• Scenario 3: Industrial IoT (IIoT)
• Machines send vibration/temperature to cloud
• Cloud analytics predict machine failure
• Maintenance scheduled automatically
• Key Advantages
• High scalability
• Global availability
• Powerful analytics & ML models
• Centralized security and updates
14
CIoT Challenges
• CIoT faces challenges in BLURS (Bandwidth, Latency,
Uninterrupted, Resource-constraint, and Security)
• Bandwidth. The increasingly large and high-frequent
rate data produced by objects in IoT will exceed the
bandwidth availability.
• Latency. Cloud faces the challenges of achieving the
requirement of controlling the end-to-end latency within
tens of milliseconds.
• Uninterrupted. The long distance between cloud and
the front-end IoT devices can face issues derived from
the unstable and intermittent network connectivity.
15
Contd…
• Resource-constrained. Commonly, many front-end devices are resource-
constrained in which they are unable to perform complex computational
tasks and hence, CIoT systems usually require front-end devices to
continuously stream their data to the cloud.
• Security. A large number of constraint front-end devices may not have
sufficient resources to protect themselves from the attacks.
Specifically, outdoor-based front-end devices, which rely on the distant
cloud to keep them updated with the security software, can be attackers’
targets, in which the attackers are capable of performing a malicious activity at
the edge network where the front-end devices are located and the cloud does
not have full control on it.
Furthermore, the attacker may also damage or control the front-end device and
send false data to the cloud.
16
Contd…
• Fog-enabled IoT system can distribute the simple data classification tasks
to the IoT devices and assign the more complicated context reasoning tasks
at the edge gateway devices.
• Further, for the analytics tasks that involve terabytes of data, which requires
higher processing power, the system can further move the processes to the
resources at the core network.
• The industry has seen fog as the main trend for the practical IoT systems,
and the leading OpenFog consortium has established collaboration with
major industrial standard parties such as European Telecommunications
Standards Institute (ETSI) multi-access edge computing (MEC) and IEEE
Standard for fog computing and networking [9] to hasten the fog.
• Furthermore, the fog market research report stated that the market value of
fog will grow from $3.7 billion by 2019 up to $18.2 billion by 2022 across
different fields, where the top five utilization domains of fog will be
energy/utilities, transportation, healthcare, industry, and agriculture.
17
IoT applications and environments with supporting
computing paradigms
18
Data Collection from sensors to Gateway-
Healthcare
• Physical Layer – Sensors Used
• ECG sensor (electrocardiogram)
• PPG sensor (for heart rate & SpO₂)
• Temperature sensor
• Blood pressure sensor
• Accelerometer (3-axis) – detects motion
• Gyroscope – posture and fall detection
• GSR sensor – stress monitoring
• Respiration sensor
19
Example: Sensors for Health Care
20
Sensors – predictive Maintenance
• Physical Layer – Sensors Used
• Vibration Sensor (IMU / Accelerometer)
• Tri-axial vibration sensor
• Temperature sensor (Thermocouple/RTD)
• Current sensor (ACS series)
• Ultrasonic sensor for misalignment
• Pressure sensor
• Acoustic sensor – detects bearing noise
• Proximity sensor – shaft rotation
• Oil quality sensor (contamination detection)
21
Predictive Maintenance - Sensors
22
Smart Cities : Sensors
• Physical Layer – Sensors Used
• Soil moisture sensor
• Soil temperature sensor
• pH sensor
• NPK sensor (nutrient measurement)
• Ambient temperature sensor
• Humidity sensor
• Rain sensor
• Weather station sensors:
– Wind speed sensor
– Wind direction sensor
– Rain gauge
– Barometric pressure sensor
• Cattle sensors:
– GPS tracker
– Accelerometer (movement)
– Heart-rate tag
23
Smart Agriculture Sensors: Examples
24
Soil Sensors
25
Iot Protocols
26
Case Study 1
• Smart Agriculture – Cattle Health & Feeding
Automation (IoT + Edge + Cloud)
27
Case Study 1
• Context
• Large dairy farms face challenges in real-time
health monitoring, feeding optimization, and
disease prediction.
• Problem
• Delayed detection of illness, manual feeding
variations, and no predictive analytics.
28
Case Study 1
• IoT Solution
• Sensors: Temperature, heart rate, rumination,
movement accelerometer.
• Edge Node: ESP32/ThingsBoard Gateway for
preprocessing anomalies.
• Connectivity: LoRaWAN → Gateway → MQTT
→ Cloud.
• Cloud: Azure IoT Hub / ThingsBoard CE.
• Dashboard: Real-time health & feeding
analytics.
29
Contd.
• Workflow
• Sensors send vitals to ESP32.
• Edge node filters noise + performs anomaly
scoring.
• Gateway pushes data to cloud via MQTT.
• Cloud ML model predicts mastitis / stress.
• Automated feeder dispenses optimal quantity.
• Outcome
• 30–40% reduction in illness detection time,
optimized feeding, improved milk yield.
30
Case Study 2
Smart City Traffic Flow Optimization (IoT + Edge AI)
31
Case Study 2
32
Case Study 2
• Context
• Growing metropolitan cities struggle with
congestion and unpredictable traffic flow.
• Problem
• Static traffic signal timing fails during peak hours
→ high delays & pollution.
33
Case Study 2
• IoT Solution
• Sensors: Edge cameras, pollution sensors, loop
detectors.
• Edge AI: Jetson Nano classifies vehicle density.
• Fog Layer: Regional traffic controller
applications.
• Cloud: City-wide analytics; long-term planning.
34
Case Study 2
• Workflow
• Edge cameras detect vehicle count every
second.
• Fog controller aggregates density from multiple
junctions.
• AI adjusts green/red cycles dynamically.
• Cloud collects historical data for prediction.
• Outcome
• 20–45% reduction in waiting time, fuel savings,
and CO₂ emissions.
35
Case Study 3
Industrial IoT – Predictive Maintenance for
Factory Machines
36
Case Study 3
37
Case Study 3
• Context
• Motor-based systems in manufacturing fail
unexpectedly causing downtime.
• Problem
• No real-time monitoring → unplanned
shutdowns & repair costs.
38
Case Study 3
• IoT Solution
• Sensors: Vibration (IMU), current, temperature.
• Edge Analytics: Raspberry Pi performing FFT +
ML anomaly detection.
• Connectivity: Modbus TCP → OPC-UA →
Cloud.
• Dashboard: Failure prediction & maintenance
scheduling.
39
• Workflow
• Sensors capture vibration signatures.
• Edge device performs frequency analysis (FFT).
• ML model predicts bearing wear stages.
• Alerts sent to maintenance team with RUL
(Remaining Useful Life).
• Outcome
• 30–60% reduced downtime, improved machine
life, reduced maintenance cost.
40
Fog and Edge Computing Completing the Cloud
• FEC provides a complement to the cloud in IoT
by filling the gap between cloud and things
toward providing service continuum.
41
Fog Edge Completing Cloud
42
Fog Edge Completing Cloud
• Three logical tiers:
• Sensors & Controllers (Physical / Perception
layer) → Edge (Intra-net) → Cloud (Internet).
• The Edge (intra-net) hosts local compute,
storage and control; the Cloud (internet)
performs centralized processing, long-term
analytics and global control.
• Primary functions shown:
• Sensors/controllers collect raw signals.
43
Fog Edge Completing Cloud
• IoT Gateways aggregate sensor traffic and connect
device protocols to IP/MQTT/HTTP.
• Edge Servers / Network Switch / Local Storage perform
local filtering, quick analytics, short-term caching and
local visualization.
• Internet Gateway / Cloud Apps perform central
processing of summary data, big-data analytics, complex
ML training and overall orchestration.
44
Fog Edge Completing Cloud
• Fog = Mini-Cloud inside the intranet
• Fog nodes are more powerful than Edge devices but closer than Cloud.
They perform regional processing, coordination, aggregation, and medium-
latency analytics.
• Typical fog components you should add in diagram
• Under the “Intranet” section, above the Edge and below the Cloud, add:
• Fog Server Cluster
• Fog Analytics Engine
• Fog Storage / Data Lake (short-term)
• Fog Orchestrator
• Fog Security Manager
• Fog API Gateway
45
Fog and Edge Computing
46
Advantages
• Advantages of FEC
SCALE - Security, Cognition, Agility, Latency, and
Efficiency
47
Security
• FEC supports additional security to IoT devices to ensure safety and
trustworthiness in transactions.
• For example, today’s wireless sensors deployed in outdoor environments
often require a remote wireless source code update in order to resolve the
security-related issues.
• However, due to various dynamic environmental factors such as unstable
signal strength, interruptions, constraint bandwidth etc., the distant central
backend server may face challenges to perform the update swiftly and,
hence, increases the chance of cybersecurity attack.
• On the other hand, if the FEC infrastructure is available, the backend can
configure the best routing path among the entire network via various FEC
nodes in order to rapidly perform the software security update to the
wireless sensors.
48
Cognition
• FEC enables the awareness of the objectives of their
clients toward supporting autonomous decision-making
in terms of where and when to deploy computing,
storage, and control functions.
• self-adaptation, self-organization, self-healing, self-
expression, and so forth shifts the role of IoT devices
from passive to active smart devices that can
continuously operate and react to customer
requirements without relying on the decision from the
distant Cloud.
49
Agility
• FEC enhances the agility of the large scope IoT system deployment.
• In contrast to the existing utility Cloud service business model, which relies
on the large business holder to establish, deploy, and manage the
fundamental infrastructure.
• FEC brings the opportunity to individual and small businesses to participate
in providing FEC services using the common open software interfaces or
open Software Development Kits (SDKs).
50
Latency
• FEC systems are capable of supporting time-sensitive functions.
• Moreover, the softwarization feature of FEC, in which the behavior
of physical devices can be fully configured by the distant central
server using software abstraction, provides a highly flexible platform
for rapid re-configuration of the IoT devices.
51
Efficiency
• FEC enhances the efficiency of CIoT in terms of
improving performance and reducing the
unnecessary costs.
• By applying FEC, the ubiquitous healthcare or
eldercare system can distribute a number of
tasks to the Internet gateway devices of the
healthcare sensors and utilize the gateway
devices to perform the sensory data analytics
tasks.
52
How FEC Achieves These Advantages: SCANC
• SCANC - Storage, Compute, Acceleration,
Networking, and Control
53
Storage
• The mechanism of storage in FEC corresponds to the
temporary data storing and caching at the FEC nodes in
order to improve the performance of information or
content delivery.
• For example, content service providers can perform
multimedia content caching at the FEC nodes that are
most close to their customers in order to improve the
quality of experience.
54
Storage
• Storage (S): FEC brings data storage closer to
devices, reducing dependence on cloud.
• What it enables:
• Low-latency access to local data
• Local backups for intermittent networks
• Pre-processing before upload
• Data privacy & compliance at the edge
• Example Scenario
• Smart Healthcare Wearables
Patient vitals (ECG, SpO₂) stored temporarily on a home
gateway → processed instantly → only critical alerts sent
to cloud.
55
Compute
• FEC nodes provide the computing mechanisms mainly in two models—
infrastructure or platform as a service (I/PaaS) and software as a service
(SaaS).
• In general, FEC providers offer I/PaaS based on two approaches—
hypervisor virtual machines (VMs) or containers engines (CEs), which
enable flexible platforms for FEC clients to deploy the customized software
they need in a sandbox environment hosted in FEC nodes.
• Besides the I/PaaS, the SaaS is also promising in FEC service provision.
• To enumerate, SaaS providers can offer two types of services—on-demand
data processing (ODP) and context as a service (CaaS)
56
Compute
• Compute (C)
• Edge devices run analytics and processing without needing the cloud.
• What it enables:
• Real-time decision making
• Reduced cloud cost
• Running ML inference at device level
• Handling high-frequency sensor streams
• Example Scenario
• Smart Traffic Systems
Jetson Nano computes vehicle density at the signal → adjusts lights →
avoids delays.
57
Acceleration
• FEC nodes support acceleration in two aspects -
networking acceleration and computing acceleration.
• Networking acceleration. Initially, most network
operators have their own configuration for message
routing paths and their clients are unable to request their
own customized routing tables.
• Computing acceleration. Researchers in fog computing
have envisioned that the FEC nodes will provide
computing acceleration by utilizing advanced embedded
processing units such as graphics processing units
(GPUs) or field programmable gate arrays (FPGA)
58
Acceleration
• Acceleration (A)
• Hardware acceleration (GPU/TPU/FPGA) boosts real-
time performance.
• What it enables:
• Fast AI inference
• Enhanced video analytics (edge cameras)
• Real-time anomaly detection
• Reduced compute load on cloud
• Example Scenario
• Industrial Predictive Maintenance
FPGA-based vibration analysis detects bearing wear
within milliseconds.
59
Networking
• Networking of FEC involves vertical and
horizontal connectivities.
• Vertical networking interconnects things and
cloud with the IP networks;
• whereas, horizontal networking can be
heterogeneous in network signals and protocols,
depending on the supported hardware
specification of the FEC nodes.
60
Networking
• Fog/Edge nodes serve as local communication
hubs with optimized routing.
• What it enables:Lower network congestion
• Adaptive bandwidth allocation
• Multi-protocol interoperability (WiFi, LoRaWAN,
ZigBee, 5G)
• Secure local messaging between IoT devices
• Example Scenario
• Smart Campus IoT
All sensors connect to local fog gateways →
only compressed summaries sent to cloud → 61
seamless campus-wide networking.
Control
• Basic types of control
– deployment, actuation, mediation, and security
– Deployment control allows clients to perform
customizable software program deployment
dynamically.
– Further, clients can configure FEC
nodes to control which program the FEC node should
execute and when it should execute it.
–
62
Contd…
• Actuation control represents the mechanism
supported by the hardware specification and the
connectivities between the FEC nodes and the
connected devices.
• Mediation control corresponds to the capability
of FEC in terms of interacting with external
entities owned by different parties.
• In particular, the connected vehicles supported
by different service providers can communicate
with one another, though they may not have a
common protocol initially.
63
Contd…
• Security control is the basic requirement of
FEC nodes that allows clients to control the
authentication, authorization, identity, and
protection of the virtualized runtime environment
operated on the FEC nodes.
64
Contd.
• Control (C)
• FEC provides local autonomy for IoT systems.
• What it enables:
• Immediate control actions without cloud
• Autonomous system behavior
• Fault tolerance
• Closed-loop operations at the edge
• Automated Agriculture
Soil sensor → edge node → irrigation pump activates
instantly without waiting for cloud approval.
65
Hierarchy of Fog and Edge Computing
• 3 Layers
– Inner-edge,
– middle-edge,
– outer-edge
66
Hierarchy of fog and edge computing
67
Inner-Edge
• Inner-edge (also known as near-the-edge)
corresponds to countrywide, statewide, and
regional WAN of enterprises, ISPs, the data
center of Evolved Packet Core (EPC) and
metropolitan area network (MAN).
• Initially, service providers at inner-edge only
offer the fundamental infrastructures for
connecting local networks to the global Internet.
• Evolved Packet Core (EPC) is a framework for
providing converged voice and data on a 4G
Long-Term Evolution (LTE) network.
68
Contd…
• However, the recent needs in improving the
Quality of Experience (QoE) of Web services
have motivated the geo-distributed caching and
processing mechanism at the network data
centers of WAN.
• For example, in the commercial service aspect,
Google Edge Network ([Link])
collaborates with ISPs to distribute data servers
at the ISPs’ data centers in order to improve the
response speed of Google’s cloud services.
69
Contd…
• Further, many ISPs (e.g. AT&T, Telstra,
Vodafone, Deutsche Telekom etc.) are aware
that many local businesses require low latency
cloud and hence, they have offered local cloud
within the country.
• Based on the reference architecture of fog
computing the WAN-based cloud data centers
can be considered as the fog of inner-edge
70
Middle-Edge
• Middle-edge corresponds to the environment of
the most common understanding of FEC, which
consists of two types of networks—local area
network (LANs) and cellular network.
• To summarize, LANs include ethernet, wireless
LANs (WLANs) and Campus Area Network
(CANs).
• The cellular network consists of the macrocell,
microcell, picocell, and femtocell.
• Explicitly, middle-edge covers a broad range of
equipment to host FEC servers.
71
Cells in the cellular network
72
Local Area Network
• virtualization technology-enabled server
computers located within the same subnet of
LAN or CAN (i.e. within the
one-hop range between the IoT device and the
computer) with the FEC nodes.
• Local cloud, local datacenter, cloudlet
73
Cellular Network
• Developed cities have wide coverage of cellular
networks provided by
– numerous types of BTSs, which are the ideal facilities
to serve as roadside FEC hosts for various mobile IoT
use cases such as
• connected vehicles,
• mobile healthcare, and
• virtual or augmented reality,
74
Outer-Edge
• Outer-edge, which is also known as extreme-
edge, far-edge, or mist, represents the front-end
of the IoT network.
• 3 types of devices
– constraint devices,
• Constraint devices such as sensors or actuators
are usually operated by microcontrollers that have
the very limited processing power and memory.
– integrated devices,
– IP gateway devices
75
Integrated Devices
• These devices are operated by processors that have
decent processing power.
• Further, integrated devices have many embedded
capabilities in networking (e.g. Wi-Fi and Bluetooth
connectivities), embedded sensors (e.g. gyroscope,
accelerator), and decent storage memory.
• Typically, Acorn RISC Machine (ARM), CPU-based
smartphones, and tablets (e.g. Android OS, iOS devices)
are the most cost-efficient commercial products of
integrated devices.
• [Link] - termux – no installation is
required.
76
IP Gateway Devices
• Hubs, or IP gateway devices, act as the mediator between the
constrained devices and the middle-edge devices
• Linux OS-based IP gateway devices such as Prota’s ([Link]),
Raspberry Pi, or ASUS Tinker Board can easily host virtualization
environment such as Docker Containers Engine.
77
Business Models
• X as a Service
– X of the X as a service (XaaS) corresponds to
infrastructure, platform, software, networking, cache
or storage, and many other types of resources
mentioned in general cloud services
– XaaS providers of FEC allow their clients to pay to
use the hardware equipment that supports SCANC
mechanisms
• Support Service
• Application Service
78
Application Service
• Application service providers provide application
solutions to help their clients in processing the
data within or outside of the client’s operation
environments.
• Digital Twin for Aerospace applications
79
Opportunities and Challenges
• Out-of-Box Experience
– OOBE-Based Equipment
– OOBE-Based Software
• Open Platforms
– OpenStack++
– WSO2–IoT Server
– Apache Edgent
• System Management
– Design
– Implementation
– Adjustment
80
OOBE-based equipment
– product vendors have integrated the FEC runtime
platform with their products such as routers, switches,
or other gateway devices in which the consumers
who purposed the equipment can easily configure
and deploy FEC applications
– user interfaces, which is similar to the commercial
router products that have graphical user interfaces for
users to configure customized settings.
81
OOBE-Based Software
• Experience of Microsoft Windows in which the
users who own FEC-compatible devices can
purchase and install OOBE-based FEC software
on their devices toward enabling FEC runtime
environment and the SCANC mechanisms
without any extra low-level configuration.
82
Open Platforms
• Commercial platform - Cisco IOX for fog computing
• OpenStack++
– OpenStack++ is a framework developed by Carnegie Mellon
University, Pittsburgh for providing VM-based cloudlet platform
on regular x86 computers for mobile application offloading.
– Explicitly, since the recent trend intends to apply lightweight
virtualization technology-based FEC, OpenStack++ is less
applicable to most use cases such as hosting FEC servers on
routers or hubs.
– Further, it indicates that the virtualization technology used in
FEC is focused more on containerization such as the Docker
Containers Engine.
83
WSO2–IoT Server
• Available at ([Link]/iot, the WSO2–IoT
server is an extension of the popular open-
source enterprise service-oriented integration
platform WSO2 server that consists of certain
IoT-related mechanisms, such as connecting a
broad range of common IoT devices.
• Examples include Arduino Uno, Raspberry Pi,
Android OS devices, iOS devices, Windows 10
IoT Core devices
84
Apache Edgent
• Formerly known as Quarks, Apache Edgent is
an open-source runtime platform contributed by
IBM.
• By utilizing Edgent, a system can dynamically
migrate the stream data processing between the
cloud and edge, which ideally fulfils the need in
most use cases that involve edge analytics.
85
Scenarios in Fog and Edge
• Smart Traffic Management (Edge AI → Fog
Coordination → Cloud Analytics)
86
Smart Traffic Management
• Hierarchy Role
• Device Layer: Cameras, loop detectors, speed sensors.
• Edge Layer: Jetson Nano performs vehicle detection + congestion score.
• Fog Layer: Regional fog node coordinates multiple intersections.
• Cloud Layer: Long-term analytics, pattern prediction, city-wide dashboard.
• Flow
• Traffic cameras capture real-time video.
• Edge device detects vehicle density (AI inference).
• Fog node compares multiple junctions → optimizes signal timing.
• Cloud predicts peak hours & generates weekly traffic reports.
• Outcome
• Reduced congestion
Faster emergency routing
Data-driven city planning
87
Smart Agriculture – Automated Irrigation & Cattle
Monitoring
88
Contd..
• Hierarchy Role
• Device Layer: Soil moisture sensors, cattle trackers, temp sensors.
• Edge Layer: ESP32/Raspberry Pi decides irrigation on/off immediately.
• Fog Layer: Farm-level fog node aggregates data from all fields.
• Cloud Layer: Crop analytics, weather integration, prediction models.
• Flow
• Soil sensor detects low moisture.
• Edge node triggers irrigation instantly (no cloud delay).
• Fog layer manages water usage for entire farm cluster.
• Cloud predicts crop yield + water requirement for next week.
• Outcome
• Watersavings
• Automatedirrigation
Early disease/stress detection in cattle
89
Industrial IoT – Predictive Maintenance in Smart
Factory
90
Hierarchy Role
• Device Layer: Vibration sensors, temperature probes, current
sensors.
• Edge Layer: Local controller performs FFT + anomaly detection.
• Fog Layer: Factory fog server correlates multiple machines’ data.
• Cloud Layer: Predictive ML model gives Remaining Useful Life
(RUL).
• Flow
• Machines send continuous vibration patterns.
• Edge gateway detects abnormal vibration signatures.
• Fog node aggregates plant-wide data & prioritizes alerts.
• Cloud predicts component failure & schedules maintenance.
• Outcome
• ✔ Reduced machine downtime
✔ Early fault detection
✔ Accurate maintenance scheduling
91