Internet of Things (IoT)
Unit I: Introduction to Internet of Things
1.1 Evolution of Internet of Things
The concept of IoT has evolved from early ideas of ubiquitous computing and machine-to-
machine (M2M) communication. The evolution can be broadly categorized into three phases:
1 Pre-IoT (1980s–1990s): Characterized by early M2M communication, primarily in
industrial settings (e.g., SCADA systems). Connectivity was proprietary, expensive,
and limited to closed networks.
2 Early IoT (2000s): The term "Internet of Things" was coined in 1999. This phase
saw the rise of RFID tags and early sensor networks. The focus was on connecting
physical objects to the internet, primarily for inventory and tracking.
3 Modern IoT (2010s–Present): Driven by the proliferation of affordable, powerful
microprocessors (like Raspberry Pi), ubiquitous wireless connectivity (4G/5G), and
the maturity of cloud computing. The focus shifted from simple tracking to data-
driven decision-making and creating new business models.
Business Implication: The evolution highlights a shift from technology-centric M2M to a
business-centric IoT, where the value lies not in the connection itself, but in the data and
insights generated.
1.2 Physical Design of IoT
The physical design refers to the "Things" themselves—the devices, sensors, and actuators
that interact with the physical world.
• Things: Devices with unique identities that collect data (sensors) or perform actions
(actuators).
• IoT Protocols: The communication standards that govern how these devices connect
and exchange data. These span multiple layers, from the physical layer (e.g., Wi-Fi,
Bluetooth) to the application layer (e.g., HTTP, MQTT, CoAP).
Example: In a Smart Warehouse scenario, the Physical Design includes:
• Sensors: Temperature and humidity sensors (collect data).
• Actuators: Automated ventilation systems (perform actions).
• Protocols: Wi-Fi or ZigBee for local communication, and MQTT for sending data to
the cloud.
1.3 Logical Design of IoT
The logical design defines the abstract components of an IoT system, including the software,
applications, and processes that handle the data flow. Key components include:
• IoT Functional Blocks:
◦ Device: Sensing, actuation, and local processing.
◦ Communication: Handling data transfer (protocols).
◦ Services: Device management, data management, and security.
◦ Application: The user-facing interface for monitoring and control.
• IoT Communication Models:
◦ Request-Response: Client requests data, server responds (e.g., a user
checking a sensor reading via a web app).
◦ Publish-Subscribe (Pub/Sub): Devices publish data to a topic, and interested
applications (subscribers) receive it (e.g., a temperature sensor publishing data
to a "Warehouse/Temp" topic). This is highly scalable.
◦ Push-Pull: Data is pushed to a queue by producers and pulled by consumers
(e.g., a data processing pipeline).
Business Implication: The logical design dictates the scalability, reliability, and security
of the IoT solution, which are critical managerial considerations.
1.4 IoT Enabling Technologies
IoT relies on a convergence of several technologies:
Enabling
Role in IoT Business Value
Technology
Wireless Sensor Provides real-time visibility into
Collects data from the environment.
Networks (WSN) operations.
Provides scalable storage and
Reduces capital expenditure and enables
Cloud Computing processing power for massive IoT
global deployment.
data.
Extracts meaningful patterns and Supports predictive maintenance,
Big Data Analytics
insights from the collected data. optimized operations, and new services.
Communication Ensures seamless and efficient data Guarantees reliable connectivity, even in
Protocols transfer (e.g., MQTT, CoAP). low-power or constrained environments.
1.5 IoT Levels and Deployment Templates
IoT systems are categorized into six levels based on their complexity, number of devices, and
data processing location. MBA students should view these levels as a roadmap for investment
and system complexity.
Level Description Example (Business Context)
Level Single device, local analysis, A smart thermostat in a single office that adjusts temperature
1 no cloud. based on local sensor data.
Level Single device, data stored in A smart home security camera that uploads video footage to a
2 the cloud. cloud server for remote viewing.
Level Single device, cloud analysis, A high-end industrial machine that streams performance data to
3 data is "Big Data." the cloud for complex predictive maintenance modeling.
Level Multiple devices, cloud A fleet of delivery trucks with GPS trackers sending location data
4 analysis, no controller. to a central logistics platform.
A smart agriculture system where a gateway collects data from
Level Multiple devices, coordinator
soil sensors and weather stations before sending aggregated data to
5 node, cloud analysis.
the cloud.
A large-scale Smart City traffic management system where a
Level Multiple devices, centralized
central control center manages traffic lights based on data from
6 controller, cloud analysis.
thousands of street sensors.
1.6 Domain Specific to IoTs
IoT applications are transforming numerous industries, creating new market opportunities:
• Smart Cities: Traffic management, smart parking, waste collection.
• Smart Agriculture (AgriTech): Precision farming, soil monitoring, automated
irrigation.
• Industrial IoT (IIoT): Predictive maintenance, asset tracking, quality control.
• Healthcare (IoMT): Remote patient monitoring, smart wearables, asset tracking in
hospitals.
• Retail: Inventory management, personalized in-store experiences, supply chain
visibility.
Unit II: IoT Architecture
IoT architecture provides a structured framework for understanding the components and their
interactions. Several standards bodies have proposed models, including ETSI, IETF, and
OGC.
2.1 ETSI, IETF, OGC Architectures
Architectur
Focus Key Components (Managerial View)
e
Machine-to-Machine Service Capability Layer (SCL): The core layer
Communication. Focuses on service that provides reusable functions (e.g., security,
ETSI M2M
layer capabilities for managing data management) to applications, reducing
devices and data transfer. development time.
6LoWPAN, CoAP, ROLL: Lightweight
Internet Protocol (IP) Stack. protocols that enable low-power devices to
IETF Focuses on adapting standard internet connect directly to the internet using IPv6,
protocols for constrained IoT devices. ensuring global addressability and
interoperability.
Sensor Web Enablement (SWE): Standards for
Geospatial and Sensor Web.
describing, accessing, and retrieving sensor data,
OGC Focuses on standards for location-
critical for applications like environmental
based data and sensor observation.
monitoring and smart cities.
2.2 IoT Reference Model
The IoT Reference Model is a conceptual framework that defines the key aspects of an IoT
system. Understanding these models is crucial for designing and evaluating IoT solutions.
Domain Model
Defines the physical entities in the IoT system and their relationships.
• Entities: Devices, Resources, Services, and Applications.
• Business Relevance: Helps in defining the scope of the IoT solution and the assets
that need to be digitized.
Information Model
Defines the data structures and semantics of the information exchanged.
• Key: Standardizing data formats (e.g., temperature, location) to ensure
interoperability between different devices and applications.
• Business Relevance: Ensures data quality and consistency, which is vital for accurate
analytics and decision-making.
Functional Model
Defines the functional groups and capabilities required to manage and operate the IoT
system.
• Groups: Device Management, Communication Management, Data Management, and
Security.
• Business Relevance: Provides a blueprint for the required software and operational
capabilities, guiding technology procurement and organizational structure.
Communication Model
Defines the protocols and mechanisms for data exchange (e.g., Pub/Sub, Request-
Response).
• Business Relevance: Impacts system latency, bandwidth requirements, and overall
network cost.
2.3 IoT Reference Architecture
A comprehensive, layered view of the entire IoT system, often depicted in a 7-layer model
(e.g., Cisco's model) or a 5-layer model (e.g., simplified academic model).
Layer Function Business Focus
Physical devices (sensors, Asset Digitization: Cost and deployment of
1. Sensing/Perception
actuators) collecting data. physical assets.
Connectivity Strategy: Choosing the right network
Secure and reliable data
2. Network/Transport technology (e.g., 5G, LoRaWAN, Wi-Fi) based on
transmission.
range, power, and cost.
Latency Reduction: Enabling immediate actions
3. Edge/Fog Local data processing and
(e.g., industrial control) and reducing cloud
Computing real-time decision-making.
bandwidth costs.
Aggregation, cleaning,
4. Data Data Governance: Ensuring data security,
and storage of massive
Processing/Storage compliance, and readiness for analytics.
data.
5. User-facing applications Value Creation: Developing new services,
Application/Business and business logic. optimizing operations, and generating revenue.
Unit III: Building IoT
3.1 IoT Systems and Logical Design using Python
Building an IoT system involves integrating hardware, software, and cloud services. Python
is a popular choice for the logical design and application layer due to its simplicity, extensive
libraries, and compatibility with single-board computers like Raspberry Pi.
Example: Smart Retail Shelf System
• System Goal: Monitor inventory levels and alert staff when a product is low.
• Logical Design (Python):
3.1 A Python script on a local gateway reads data from weight sensors under the
shelf.
3.2 The script processes the raw weight data to determine the number of items.
3.3 If the count falls below a threshold, the script uses an MQTT library to publish
an alert message to a cloud broker.
3.4 A cloud application (also potentially Python-based) subscribes to the alert
topic and sends a notification to the store manager's mobile app.
3.2 IoT Physical Devices and Endpoints
What is an IoT Device?
An IoT device is a piece of hardware with sensing, processing, and communication
capabilities that connects to a network to collect and exchange data.
Basic Building Blocks of an IoT Device
4 Sensor/Actuator: The interface to the physical world (input/output).
5 Microcontroller/Microprocessor: The "brain" that processes data and runs the
operating system/application logic.
6 Communication Module: Enables network connectivity (e.g., Wi-Fi, Bluetooth,
Cellular).
7 Power Supply: Battery or wired power source.
Exemplary Device: Raspberry Pi
The Raspberry Pi is a low-cost, credit-card-sized computer often used as a gateway or a
powerful IoT endpoint.
• Programming Raspberry Pi with Python: Python is the preferred language for the
Pi because it allows for rapid prototyping and easy interaction with the Pi's General
Purpose Input/Output (GPIO) pins, which connect to sensors and actuators.
• Other IoT Devices: Arduino (simpler, lower-power microcontrollers),
ESP32/ESP8266 (Wi-Fi enabled microcontrollers), and industrial gateways (rugged,
high-performance devices for IIoT).
Unit IV: IoT Data Platform
The sheer volume, velocity, and variety of data generated by IoT devices necessitate
specialized data platforms and analytics tools.
4.1 Data Analytics for IoT
IoT data is often unstructured, time-series based, and massive, requiring Big Data solutions.
• Batch Data Analysis: Processing large volumes of historical data over a period of
time to uncover long-term trends and patterns.
◦ Apache Hadoop: A framework for distributed storage (HDFS) and processing
(MapReduce) of large datasets across clusters of commodity hardware.
◦ Using Hadoop MapReduce for Batch Data Analysis: A programming
model for processing large data sets with a parallel, distributed algorithm on a
cluster. Business Use: Analyzing a year's worth of sensor data from a fleet of
trucks to identify common failure points and optimize maintenance schedules.
◦ Apache Oozie: A workflow scheduler system to manage and coordinate
Hadoop jobs (e.g., scheduling a daily MapReduce job to generate an end-of-
day report).
• Real-Time/Stream Data Analysis: Processing data as it arrives to enable immediate
action.
◦ Apache Spark: An open-source, distributed processing system used for large-
scale data processing. It is significantly faster than Hadoop MapReduce for
many workloads because it performs in-memory processing. Business Use:
Monitoring a factory floor for anomalies in real-time to shut down a machine
before a catastrophic failure occurs.
4.2 Tools for IoT: Chef
Chef is an automation platform that turns infrastructure into code. In an IoT context, it is a
crucial tool for managing large-scale deployments.
• Setting up Chef: Chef uses "recipes" and "cookbooks" to define the desired state of a
device (e.g., install Python, configure network settings, deploy the sensor reading
script).
• Business Value:
7.1 Scalability: Allows a business to deploy and manage thousands of identical
IoT devices quickly and consistently.
7.2 Consistency: Ensures every device is configured correctly, reducing errors
and security vulnerabilities.
7.3 Operational Efficiency: Automates updates and maintenance, significantly
lowering the operational cost of a large IoT fleet.
Unit V: Case Studies and Real-World Applications
5.1 Real-World Design Constraints
When designing an IoT solution, MBA students must consider several non-technical, real-
world constraints:
8 Power/Energy: Many devices are battery-powered and must operate for years
without maintenance (e.g., remote environmental sensors). This dictates the choice of
communication protocol (e.g., LoRaWAN over Wi-Fi).
9 Security and Privacy: IoT devices are often deployed in public or sensitive areas,
making them targets for cyberattacks. Data privacy (especially in healthcare/smart
home) is a major legal and ethical constraint.
10 Interoperability: Solutions must often integrate with legacy systems and devices
from different vendors. Lack of open standards can lead to vendor lock-in.
11 Cost and ROI: The total cost of ownership (TCO)—including hardware,
connectivity, cloud, and maintenance—must justify the return on investment (ROI).
5.2 IoT Physical Servers & Cloud Offerings
IoT solutions require robust backend infrastructure, typically provided by major cloud
vendors.
Cloud Offering Description Business Advantage
Tools for provisioning, updating, and
Device Reduces operational overhead and
monitoring the health of devices (e.g., AWS
Management ensures fleet reliability.
IoT Device Management).
Services that securely receive and route data
Data Provides massive scalability and
from millions of devices (e.g., Azure IoT Hub,
Ingestion/Broker security for data entry points.
Google Cloud IoT Core).
Cost-effective processing of data
Serverless Services that run code in response to events
streams, paying only for the
Computing (e.g., AWS Lambda, Azure Functions).
compute time used.
5.3 Case Studies Illustrating IoT Design
Case Study 1: Asset Management - Predictive Maintenance in Manufacturing
Challenge: A large manufacturing company was experiencing unplanned downtime due to
equipment failure, leading to high repair costs and production losses. Their existing
maintenance was time-based, not condition-based.
IoT Solution:
12 Deployment: Installed vibration and temperature sensors on critical machinery (e.g.,
motors, pumps).
13 Data Flow: Sensors streamed data via an industrial gateway (Level 5 IoT) to a cloud-
based IoT platform.
14 Analytics: Apache Spark was used to process the real-time data stream, and machine
learning models were applied to detect subtle anomalies that precede failure.
Business Outcome (Solution):
• Shift to Predictive Maintenance: The company moved from reactive or time-based
maintenance to predictive maintenance.
• ROI: Unplanned downtime was reduced by 25%, and maintenance costs were
lowered by 15% by replacing parts only when necessary, maximizing asset lifespan.
This demonstrates the strategic value of IoT in optimizing operational expenditure
(OpEx).
Case Study 2: Smart Cities - Intelligent Parking in Barcelona
Challenge: Barcelona faced severe traffic congestion, with up to 40% of city-center traffic
caused by drivers searching for parking, leading to increased pollution and wasted time.
IoT Solution:
15 Deployment: Embedded magnetic sensors in parking spaces across key areas.
16 Data Flow: Sensors reported real-time occupancy status to a central city data
platform.
17 Application: The data was fed to public display panels and a mobile application,
guiding drivers directly to available spots.
Business Outcome (Solution):
• Traffic Flow Optimization: Reduced the time spent searching for parking, easing
congestion.
• Environmental Impact: Lowered fuel consumption and CO2 emissions in the city
center.
• Revenue Optimization: Enabled dynamic pricing for parking spots based on
demand, increasing city revenue and encouraging turnover. This is a clear example of
IoT creating a smart service that improves both citizen experience and municipal
finance.
Case Study 3: Environment - Smart Waste Management
Challenge: A city was using a fixed schedule for waste collection, resulting in inefficient
routes (collecting half-empty bins) and overflowing bins in high-traffic areas.
IoT Solution:
18 Deployment: Installed ultrasonic fill-level sensors in public waste bins.
19 Data Flow: Sensors transmitted fill-level and temperature data (to detect fire risk) via
a low-power network (e.g., LoRaWAN) to a central platform.
20 Application: The platform used an optimization algorithm to generate the most
efficient collection routes daily, only servicing bins that were near full.
Business Outcome (Solution):
• Operational Efficiency: Reduced the number of collection trips by 30%, saving
significant costs on fuel and labor.
• Improved Service: Eliminated overflowing bins, improving public hygiene and
citizen satisfaction.
• Sustainability: Lowered the city's carbon footprint from waste collection vehicles.
This case highlights how IoT can drive sustainability and cost reduction
simultaneously.
