BETCK205H Introduction to IoT

MODULE- 5

IOT CASE STUDIES & FUTURE TRENDS

Chapter 1: Vehicular IoT

Introduction: Vehicular IoT also refers to as 'Automotive IoT' or ‘connected

vehicles system’. It is nothing but to embedding IoT technologies into automotive
systems to create new applications and solutions which enable vehicles to provide a
smarter, safer, more efficient, and more comfortable driving experience.

The use of connected vehicles is increasing rapidly across the globe.

Consequently, the number of on-road accidents and mismanagement of traffic is
also increasing. The increasing number of vehicles gives rise to the problem of
parking. However, the evolution of IoT helps to form a connected vehicular
environment to manage transportation systems efficiently. Vehicular IoT systems
addressed the different aspects of the transportation ecosystem, including

• On-road to off-road traffic management

• Driver safety for heavy to small vehicles
• Security in public transportation.
• Vehicles are capable of communicating and sharing their information.
• A vehicle owner easily tracks his vehicular asset remotely

The architecture of a Vehicular IoT system. The simple architecture of the vehicular IoT is
divided into THREE sublayers: (see fig 1.1)

1. Device layer

2. Fog layer

3. Cloud layer

Dept. of ECE, GSSSIETW, Mysuru Page 1

 BETCK205H Introduction to IoT

1. Device layer:

- This layer consists of the basic infrastructure of the connected vehicle.

- This layer includes the vehicles and road side units (RSU).

- These vehicles contain certain sensors which gather internal information about the vehicles.

- The RSU works as a local centralized unit that manages the data from the vehicles.

2. Fog layer:

- Fog computing provides fast decisions in real-time, much near to the devices.

- In vehicular IoT systems, fast decision-making is important to avoid accidents and traffic
mismanagement. In such situations, the fog layer helps to minimize data transmission time in a
vehicular IoT system.

3. Cloud layer:

Dept. of ECE, GSSSIETW, Mysuru Page 2

 BETCK205H Introduction to IoT

- Processing of huge data, fog computing is not enough, in such a situation, cloud computing is
used.

- In a vehicular IoT system, cloud computing helps to handle processes a huge amount of data,
and long-term storage

1.1 Components of vehicular IoT:

The processes involved in a vehicular IoT have several components, along with many practical
challenges. Figure 1.2 depicts the components required for vehicular IoT systems. The main
components are as follows

a. Sensors

b. Satellite

c. Wireless Connectivity.

d. Road side unit (RSU)

e. Cloud and for Computing

f. Analytics

Dept. of ECE, GSSSIETW, Mysuru Page 3

 BETCK205H Introduction to IoT

Figure 1.2 Components of vehicular IoT

a). Sensors: In vehicular IoT, sensors monitor different environmental conditions and make the
system more economical, efficient, and robust. Two types of sensors are used

(i) internal sensors:

- The sensors are used within the vehicle to sense parameters that are directly associated with
the vehicle.

- These internal sensors in a vehicle are connected to the processor board and actuator, to which
they transmit the sensed data.

- Further, the sensed data are processed to take certain predefined actions.

- Examples of internal sensors are GPS, fuel gauge, ultrasonic sensors, proximity sensors,
accelerometer, pressure sensors, and temperature sensors.

Dept. of ECE, GSSSIETW, Mysuru Page 4

 BETCK205H Introduction to IoT

(ii) external sensors:

- It quantifies information like smart traffic systems and vacant parking lots outside the vehicle.

- The still images and videos from cameras are important inputs to generate decisions in a
vehicular IoT system.

- Therefore, on-road cameras are widely used as external sensors to capture still images and
videos.

- The captured images and videos are processed, either in the fog or in the cloud layer, to take
certain pre-programmed actions.

- As an example, a camera sensor can capture the image of the license plate of an over speeding
vehicle at a traffic signal.

b). Satellites:

- In vehicular IoT systems, automatic vehicle tracking and crash detection are among the
important available features.

- Satellites help the system track vehicles and detect on-road crashes, detecting on-road
congestions, and roadblocks.

c). Wireless connectivity:

- As vehicular IoT deals with connected vehicles, communication is an important enabling

component.

- It carries the sensed data from multiple sensors to RS from RSUs to the cloud.

- In the vehicular IoT scenario, the high mobility of the vehicles necessitates the connectivity
type to be wireless for practical and real-time data transmission.

- Different communication technologies, such as Wi-Fi, Bluetooth, and GSM, are common in
vehicular IoT systems.

Dept. of ECE, GSSSIETW, Mysuru Page 5

 BETCK205H Introduction to IoT

d). Road Side Unit (RSU):

- The RSU is a static entity that works collaboratively with internal and external sensors.

- The RSUs are equipped with sensors, communication units, and fog devices.

- To take decisions in real-time, the fog devices attached to the RSUs process the sensed data and
take necessary action promptly.

- If a vehicular system involves heavy computation, the RSU transmits the sensed data to the
cloud end.

- RSUs also work as intermediate communication agents between two vehicles.

e). Fog computing:

- It handles the lightweight processes geographically closer to the vehicles than the cloud.

- It is used for faster decision-making in vehicular IoT systems. For example, find the traffic jam
at closer locations with the help of sensed data.

- Further, the congestion information can be shared by the RSU among another road vehicles,
thereby suggesting that they avoid the congested road.

f). Cloud computing:

- It is used for heavyweight processes & handles a huge amount of data in V_IoT systems.

- It provides more scalability of resources as compared to fog.

- The choice of the application of fog and cloud computing depends on the situation and
applications in vehicular IoT systems.

- Determining regular on-road congestion, predictions are typically handled with the help of
cloud computing.

- For the regular congestion prediction, the cloud end needs to process a huge amount of
instantaneous data, as well as, historical data for that stretch of road spanning back a few
months to years.

Dept. of ECE, GSSSIETW, Mysuru Page 6

 BETCK205H Introduction to IoT

g). Analytics:

- In vehicular IoT, analytics is a crucial component.

- It is used to predict different dynamic and static conditions in the traffics.

- For example, strong data analytics is required to predict on-road traffic conditions that may
occur at a location after an hour.

1.2 Advantages of vehicular IoT:

The advantages of vehicular IoT are depicted in the figure below.

(i) Easy tracking: In a vehicular IoT system, the tracking of vehicles is straightforward; the
system can collect information about the location and vehicle information.

(ii) Fast decision-making: Most of the decisions in the connected vehicle environment are time
critical. Fast and active decision-making is important for avoiding accidents. In the vehicular IoT
environment, cloud and fog computing help to make fast decisions with the data received from
the sensor-based devices.

Dept. of ECE, GSSSIETW, Mysuru Page 7

 BETCK205H Introduction to IoT

(iii) Connected vehicles: A vehicular IoT system provides an opportunity to remain connected
and share information among different vehicles.

(iv) Easy management: With the help of sensors, a communication unit, processing devices, and
GPS, the management of the vehicle becomes easy. The connectivity among different
components in a vehicular IoT enables systems to track every activity in and around the vehicle.

(v) Safety: Both the internal &external sensors placed at different locations provide safety to the
vehicle, its occupants, as well as the people around it.

(vi) Record: The record may be of any form, such as video footage, still images, and
documentation. By taking advantage of cloud and fog computing architecture, vehicular IoT
systems keep all the required records stored in their database.

1.3 Crime assistance in a smart IoT transportation system:

This section provides a case study on smart safety in vehicular IoT infrastructure. The system
highlights a fog framework for intelligent public safety in vehicular environments (fog-FISVER).
The primary aim of this system is to Ensure Smart Transportation Safety (STS) in public bus
services.

The system works through the following three steps:

a. The vehicle is equipped with a smart surveillance system, which is capable of executing video
processing and detecting criminal activity in real-time.

b. A fog computing architecture works as the mediator between a vehicle and a police vehicle.

c. A mobile application is used to report the crime to a nearby police agent.

Architecture of Fog-Intelligent Safety in Vehicular Environments (fog-FISVER).

- Fog-FISVER framework is designed for low-latency fog computing architecture needed for the
connected vehicle.

Dept. of ECE, GSSSIETW, Mysuru Page 8

 BETCK205H Introduction to IoT

- Fog-FISVER consists of different IoT components and it developed based on a three-tiered

architecture

- Figure 1.3 describes each of the tiers as follows:

Tier1: In-vehicle FISVER smart transportation safety(STS) Fog:

o This tier accumulates the real sensed data from within the vehicle and processes it to detect
possible criminal activities inside the vehicle.

o This tier is responsible for creating crime-level metadata and transferring the required
information to the next tier.

o For performing all the activities, Tier 1 consists of two subsystems:

(1) Image processor (2) Event dispatcher

Dept. of ECE, GSSSIETW, Mysuru Page 9

 BETCK205H Introduction to IoT

• Image Processor: It is a potent component, which has the capability of detecting criminal
activities. The processor system used a deep-learning-based approach for enabling image
processing techniques. To implement the fog computing architecture in the vehicle, a Raspberry-
Pi-3 processor board is used, which is equipped with a high-quality camera. The image
processor stores a set of crime object templates in Tier 2 of the system. The image processor is
divided into the following three parts:

(a) Crime definition downloader: It periodically checks the new crime object template
definitions in fog, If available, it is stored locally.

(b) Crime definition storage: It is used to store all the possible crime object template definitions.

(c) Algorithm launcher: It matches the template with the video captured by the camera attached
to the vehicles. If a crime object is matched with the video, criminal activity is confirmed.

• Event dispatcher: It is responsible for accumulating the data sensed from vehicles and the
image processor. After the successful detection of criminal activity, the information is sent to the
fog-FISVER STS fog infrastructure. The components of the event dispatcher are as follows:

(a) Event notifier: It transfers the data to the fog-FISVER STS fog infrastructure, after receiving it
from the attached sensor nodes in the vehicle.

(b) Data gatherer: This is an intermediate component between the event notifier and the
physical sensor; it helps to gather sensed data.

(c) Virtual sensor interface: It helps to maintain a particular procedure to gather data. This
component also cooperates to register the sensors in the system.

Tier 2: FISVER STS Fog Infrastructure: Primarily, this tier has three responsibilities—keep
updating the new object template definitions, classifying events, and finding the most suitable
police vehicle to notify the event. It is divided into TWO sub-components:

(1) Target Object Training:

- It is responsible for creating, updating, and storing all possible crime object definitions.

- The algorithm launcher in tier 1 uses these definitions for the template-matching process.

Dept. of ECE, GSSSIETW, Mysuru Page 10

 BETCK205H Introduction to IoT

- The template definition includes different features of the crime object such as color gradient
and shape format

(2). Notification Factory:

- This sub-component receives a notification about the events in a different vehicle with the
installed system. Further, this component receives and validates the events. To handle multiple
events, it maintains a queue.

Tier 3: It consists of mobile applications that are executed on the users’ devices. The application
helps a user, who witnesses a crime, to notify the police.

Dept. of ECE, GSSSIETW, Mysuru Page 11

 BETCK205H Introduction to IoT

Chapter 2: Healthcare IoT

1. Introduction

o IoT made a huge impact on the medical field, especially wearable healthcare. These
technologies have given rise to small, power-efficient, health monitoring and diagnostic systems.

o Currently, various IoT-enabled healthcare devices are in wide use around the globe for
diagnosing human diseases, monitoring human health conditions, and caring /monitoring for
elders, children, and even infants.

o Moreover, IoT-based healthcare systems and services help to increase the quality of life for
common human beings; in fact, it has a promising scope of revolutionizing healthcare in
developing nations.

o IoT-based healthcare devices provide access and knowledge about human physiological
conditions through handheld devices.

o In IoT-based healthcare services, the sensors are specifically designed to measure and quantify
different physiological conditions of their users/patients.

Architecture for healthcare IoT: The architecture into FOUR layers. It is shown in Figure 1.4.
The detailed description of these layers is as follows:

(i) Layer 1: It contains different physiological sensors that are placed on the human body. These
sensors collect the values of various physiological parameters. The physiological data are
analyzed to extract meaningful information.

(ii) Layer 2: It collects short-term storage and low-level processing data from layer 1. Layer 2 is
equipped with local processing units (LPU). These units collect the sensed data from the
physiological sensors attached to the body and process it based on the architecture’s
requirements. Further, LPUs forward the processed data to Layer 3.

Dept. of ECE, GSSSIETW, Mysuru Page 12

 BETCK205H Introduction to IoT

(iii) Layer 3: This layer consists of cloud architecture or high-end servers that perform
application-specific high-level analytics. The data from multiple patients, which may be from the
same or different locations, are accumulated in this layer. Post-analysis of data, some inferences
or results are provided to the application in Layer 4.

(iv) Layer 4: The end-users directly interact with Layer 4 through receiver-side applications.
The modes of accessibility of these services by an end user are typically through cell phones,
computers, and tablets.

Components of healthcare IoT: A typical IoT healthcare architecture is composed of several

components. Figure 1.5 depicts different components and their usage in an IoT healthcare
system

Dept. of ECE, GSSSIETW, Mysuru Page 13

 BETCK205H Introduction to IoT

(i) Sensors: Uses physiological sensors that collect the physiological parameters of the patient.
Commonly used physiological sensors and their uses are depicted in Table 1.1.

Dept. of ECE, GSSSIETW, Mysuru Page 14

 BETCK205H Introduction to IoT

(ii) Wireless Connectivity: Wireless connectivity is preferably used to communicate between

the wearable sensors and the Local Processing Unit (LPU) with the help of Bluetooth and ZigBee.
On the other hand, the communication between the LPU and the cloud or server takes place with
Internet connectivity such as Wi-Fi and WLAN. The healthcare data are received by the end
users with different devices such as laptops, desktops, and cellphones by use of 3G/4G/5G or
WiFi

(iii) Privacy and Security: It is a major concern in healthcare IoT services. In a healthcare IoT
architecture, several devices connect with the external world. If any of the devices are
compromised, it may result in the theft of the health data of a patient, leading to serious security
breaches and ensuing lawsuits. To increase the security of healthcare data, different healthcare

Dept. of ECE, GSSSIETW, Mysuru Page 15

 BETCK205H Introduction to IoT

service providers and organizations are implementing healthcare data encryption and
protection schemes.

(iv) Analytics: It converts raw data into meaningful information. The doctors, nurses, and
patients access healthcare information in a different customized format. Analytics is also used
for diagnosing a disease from the raw physiological data available.

(v) Cloud and Fog Computing: It plays a pivotal role in the storage of massive volumes of
heterogeneous health data generated from the patient body. These data are used for checking
the patient’s history, current health status, and future for diagnosing different diseases and the
symptoms of the patient. To store health data in a healthcare IoT system, cloud storage space is
used. Analytics of the stored data in cloud storage space is used for drawing various inferences.
The major challenges in storage are security and delay in accessing the data.

(vi) Interface: The interface is the most important component for users in a healthcare

IoT system. Among IoT applications, healthcare IoT is a very crucial and sensitive application.
Thus, the user interface must be designed in such a way that it can depict all the required
information clearly and, if necessary, reformat or represent it such that it is easy to understand.
Moreover, an interface must also contain all the useful information related to the services

Advantages and risks of healthcare IoT: Different advantages and risks of healthcare IoT as
depicted in Figure 1.5

Dept. of ECE, GSSSIETW, Mysuru Page 16

 BETCK205H Introduction to IoT

Advantages of healthcare IoT:

a. Real-time:

- A healthcare IoT system enables users, such as doctors, end users at the patient side, and staff
in a healthcare unit, to receive real-time updates about the healthcare IoT components.

- It enables a doctor to observe a patient’s health condition in real-time even from a remote
location and can suggest the type of care to be provided to the patient. On the other hand, users
at the patient end can easily take different decisions, such as where to take a patient during
critical situations.

- Moreover, the staff in a healthcare unit are better aware of the current situation of their unit,
which includes the number of patients admitted, availability of the doctors and beds, total
revenue of the unit, and other such information.

b. Low cost:

- Healthcare IoT systems facilitate users with different services at low cost.

- For example, an authorized user can easily find the availability of beds in a hospital with simple
Internet connectivity and a web-browser-based portal.

- Moreover, multiple registered users can retrieve the same information simultaneously.

c. Easy management:

- Healthcare IoT is an infrastructure that brings all its end users under the same umbrella to
provide healthcare services.

- However, healthcare IoT facilitates easy and robust management of all entities.

d. Automatic processing:

- Healthcare IoT enables end-to-end automatic processing in different units and also
consolidates the information across the whole chain: from a patient’s registration to discharge.

e. Easy record-keeping:

Dept. of ECE, GSSSIETW, Mysuru Page 17

 BETCK205H Introduction to IoT

- A healthcare IoT enables the user to keep these records in a safe environment and deliver them
to the authorized user as per requirement. Moreover, these recorded data are accessible from
any part of the globe.

f. Easy diagnosis:

- In a healthcare IoT system, the diagnosis of the disease becomes easier with the help of certain
learning mechanisms along with the availability of prior datasets.

Risk in healthcare IoT: The various risks associated with a healthcare IoT system.

a. Loss of connectivity:

- Intermittent connectivity may result in data loss, which may result in a life-threatening
situation for the patient. Proper and continuous connectivity is essential in a healthcare IoT
system.

b. Security:

- The healthcare system must keep the data confidential. This data should not be accessible to
any unauthorized person. On the other hand, different persons and devices are associated with a
healthcare IoT system. In such a system, the risk of data tampering and unauthorized access is
quite high.

c. Error:

- In the healthcare system, errors in data may lead to misinterpretation of symptoms and lead to
the wrong diagnosis of the patient. It is a challenging task to construct an error-free healthcare
IoT architecture.

2. Case Studies: AmbuSens system – This case study deals with

Overview of AmbuSens system

The architecture of the AmbuSens system

The hardware of the AmbuSens system

Dept. of ECE, GSSSIETW, Mysuru Page 18

 BETCK205H Introduction to IoT

Front End of AmbuSens system

Overview of AmbuSens system

o It is an IoT-based healthcare system developed by Smart Wireless Applications and

Networking (SWAN) laboratory at IIT Kharagpur.

o The system was primarily funded by the Ministry of Human Resource and Development
(MHRD) of the Government of India.

o AmbuSens collects the physiological parameters of the remote patients and performs real-time
analysis of the collected data.

o The primary objectives of the AmbuSens system are summarized as follows:

- Digitization and standardization of the healthcare data, which can be easily accessed by the
registered hospital authorities.

- Real-time monitoring of the patients who are in transit from one hospital to another. At both
hospitals, doctors can access the patient’s health conditions.

- Accessibility by which multiple doctors can access the patient’s health data at the same time.

- Provision of confidentiality to the health data of the patients in the cloud.

- In the AmbuSens system, wireless physiological sensor nodes are used. These sensor nodes
make the system flexible and easy to use.

Architecture AmbuSens system

- The AmbuSens system is equipped with different physiological sensors along with a local hub.

- These sensors sense the physiological parameters from the patient’s body and transmit those
to a local data processing unit (LDPU).

- The physiological sensors and LDPU form a wireless body area network (WBAN).

- Further, this local hub forwards the physiological data to the cloud for storing and analyzing
the health parameters.

Dept. of ECE, GSSSIETW, Mysuru Page 19

 BETCK205H Introduction to IoT

- Finally, the data are accessed by different users. The detailed layered architecture of the
AmbuSens system is depicted in Figure 1.7

Layer 1:

- This layer consists of multiple WBANs attached to a patient’s body.

- These WBANs acquire the physiological data from the patient and transmit them to the upper
layer.

- The physiological sensors are heterogeneous, that is, each of these sensors senses different
parameters of the body.

- The formation of WBAN takes place by connecting multiple physiological sensor nodes to the
LDPU so that the sensors can transmit the data to the LDPU, simultaneously.

Layer 2:

- It is responsible for handling cloud-related functions.

- WBANs attached to the different patients deliver data to the cloud end.

- The cloud is used for the long-term analysis and storage of data in the AmbuSens system.

Dept. of ECE, GSSSIETW, Mysuru Page 20

 BETCK205H Introduction to IoT

- A huge volume of health data is produced by the WBANs, which are handled by the cloud with
the help of big data analytics for providing real-time analysis.

Layer 3:

- In this layer, an algorithm is designed to generate a dynamic hash value for each patient to keep
the patient’s identity anonymous.

- In the AmbuSens system, at different time instants, a new hash value is generated for the
patients. T

- The entire hashing mechanism of the AmbuSens is performed in this layer.

Layer 4: The users simply register into the system and use it as per requirement.

The hardware of AmbuSens system: In the AmbuSens system, a variety of hardware components
are used such as

- Sensors

- Communication units

- Other computing devices.

o Sensors: The sensors used in the AmbuSens system are non-invasive. The description of the
sensors used for forming the WBAN in the AmbuSens system are as follows:

(i) Optical Pulse Sensing Probe: It senses the photoplethysmograph (PPG) signal. Typically, PPG
signals are sensed from the ear lobe, fingers, or other locations of the human body.

(ii) Electrocardiogram (ECG) unit and sensor: It is a kit that contains ECG electrodes, alcohol
swabs, and a wrist strap. Typically, the ECG sensor measures the electrical impulses of the heart
to sense the heart’s responses and other factors affecting cardiac health.

(iii) Electromyogram (EMG) sensor: The EMG sensor is used to measure different electrical
activities related to muscle contractions; it also assesses nerve conduction, and muscle response
in injured tissue.

Dept. of ECE, GSSSIETW, Mysuru Page 21

 BETCK205H Introduction to IoT

(iv) Temperature sensor: The body temperature of patients changes with the condition of the
body. Therefore, a temperature sensor is included in the AmbuSens system, which can easily be
placed on the body of the patient.

(v) Galvanic Skin Response (GSR) sensor: The GSR sensor is used for measuring the change in
electrical characteristics of the skin.

o Communication Module: Each sensor node consists of a Bluetooth module. The

communication between the sensor nodes and the LDPU takes place with the help of Bluetooth,
which supports a maximum communication range of 10 meters in line of sight. The LDPU
delivers the data to the cloud with 3G/4G communication.

o Local Data Processing Unit (LDPU): An LDPU is a small processing board with limited
computation capabilities. The connectivity between the sensors and the LDPU follows a single-
hop star topology. The LDPU is programmed in such a way that it can receive physiological data
from multiple sensor nodes, simultaneously

Front End of AmbuSens system

- In the AmbuSens system, the doctor, paramedic/nurse, and patient can participate and use the
services.

- The web interface is designed as per the requirements of the users of the system.

- Each user has the option to log in and access the system.

- The confidentiality of a patient and their physiological data is important in a healthcare system.
Therefore, the system provides different scopes for data accessibility based on the category of an
actor.

- In AmbuSens, the database is designed in an efficient way such that it can deliver customized
data to the respective user.

- Each of the users has to register with the system to avail of the service of AmbuSens. Therefore,
in this system, the registration process is also designed in a customized fashion, For example, a
doctor must enter his/her registration number in the registration form.

Dept. of ECE, GSSSIETW, Mysuru Page 22

 BETCK205H Introduction to IoT

CHAPTER 3: IoT Analytics

1. Introduction

- IoT sensors generated a huge volume of data in practice. These sensors collect data from the
environment and serve different IoT-based applications.

- IoT analytics is a data analysis tool that assesses the wide range of data collected from IoT
devices. IoT analytics assesses vast quantities of data and produces useful information from it.

- Data analytics is the process of exploring and analyzing large datasets to find hidden patterns,
and unseen trends, discover correlations, and derive valuable insights to make business
predictions.

- Some of the traditional data analytics tools are popular in the context of IoT applications. These
tools include k-means, decision tree (DT), random forest (RF), k-nearest neighbor (KNN), etc.

1.1 Machine learning :

- Machine learning is a “field of study that gives computers the ability to learn without being
explicitly programmed”.

- ML is a powerful tool that allows a computer to learn from past experiences and its mistakes
and improve itself without user intervention

- The term “machine learning” was coined by Arthur Lee Samuel, in 1959.

- Different ML models play a crucial role in designing intelligent systems in IoT by leveraging the
massive amount of generated data and increasing the accuracy of their operations.

- The main components of ML are statistics, mathematics, and computer science for drawing
inferences, constructing ML models, and implementation, respectively.

1.2 Advantages of ML: The different advantages of ML are as follows

Self-learner

Time-efficient

Dept. of ECE, GSSSIETW, Mysuru Page 23

 BETCK205H Introduction to IoT

Self-guided

Minimum Human Interaction Required

Diverse Data Handling

Diverse Applications

(i) Self-learner:

- An ML-empowered system is capable of learning from its prior and run-time experiences,
which helps in improving its performance continuously.

(ii) Time-efficient:

- ML tools are capable of producing faster results as compared to human interpretation.ML is

beneficial in predicting the weather with less delay and accuracy as compared to humans.

(iii) Self-guided:

- An ML tool uses a huge amount of data for producing its results. These tools have the capability
of analyzing a huge amount of data for identifying trends autonomously.

(iv) Minimum Human Interaction Required:

- The ML algorithm trains itself automatically, based on available data inputs. The ML algorithm
determines the disease with minimum human intervention, based on the health data available in
the system and matching the same with the symptoms of the patient.

(v) Diverse Data Handling:

- Typically, IoT systems consist of different sensors and produce diverse and multi-dimensional
data, which are easily analyzed by ML algorithms.

(vi) Diverse Applications:

- ML is flexible and can be applied to different application domains such as healthcare, industry,
smart traffic, smart home, and many others.

Dept. of ECE, GSSSIETW, Mysuru Page 24

 BETCK205H Introduction to IoT

1.3 Challenges in ML: A few major challenges in ML are listed as follows:

(i) Data Description:

- The data acquired from different sensors are required to be informative and meaningful. The
description of data is a challenging part of ML.

(ii) Amount of Data:

- To provide an accurate output, a model must have a sufficient amount of data. The availability
of a huge amount of data is a challenge in ML.

(iii) Erroneous Data:

- A dataset may contain noisy or erroneous data. The learning of a model is heavily dependent on
the quality of data. Since erroneous data misleads the ML model, its identification is crucial.

(iv) Selection of Model:

- Multiple models may be suitable for serving a particular purpose. However, one model may
perform better than others. In such cases, the proper selection of the model is pertinent for ML.

(v). Quality of Model:

- After the selection of a model, it is difficult to determine the quality of the selected model.
However, the quality of the model is essential in an ML-based system.

1.4 Types of ML:

Typically, ML algorithms consist of four categories:

(i) Supervised Learning: Classification and Regression

(ii) Unsupervised Learning: Clustering and association

(iii) Semi-supervised Learning

(iv) Reinforcement Learning

Dept. of ECE, GSSSIETW, Mysuru Page 25

 BETCK205H Introduction to IoT

(i) Supervised Learning:

- Supervised learning is the type of machine learning in which machines are trained using well
"labeled" training data and on basis of that data, machines predict the output.

- The labeled data means some input data is already tagged with the correct output.

- In supervised learning, the training data provided to the machines work as the supervisor that
teaches the machines to predict the output correctly.

- Supervised learning can be used for two types of problems: Classification and Regression.

- Example: Suppose we have an image of different types of fruits. The task of the supervised
learning model is to identify the fruits and classify them accordingly. So to identify the image in
supervised learning, we will train the model by the shape, size, color, and taste of each fruit.
Once the training is completed, we will test the model by giving the new set of fruit. The model
will identify the fruit and predict the output using a suitable algorithm.

(ii) Unsupervised Learning:

- Unsupervised learning does not use any labels in its operations.

Dept. of ECE, GSSSIETW, Mysuru Page 26

 BETCK205H Introduction to IoT

- Unsupervised learning is another machine learning method in which patterns are inferred from
the unlabelled input data.

- The goal of unsupervised learning is to find the structure and patterns from the input data.
Unsupervised learning does not need any supervision. Instead, it finds patterns from the data on
its own.

- Unsupervised learning can be used for two types of problems: Clustering and Association.

(iii) Semi-Supervised Learning:

- Semi-supervised learning belongs to a category between supervised and unsupervised

learning.

- Algorithms under this category use a combination of both labelled and unlabelled datasets for
training.

- Labelled data are typically expensive and are relatively difficult to label correctly.

- Unlabelled data is less expensive than labelled data. Therefore, semi-supervised learning
includes both labelled and unlabelled datasets to design the learning model.

(iv) Reinforcement Learning:

- "Reinforcement learning is a type of machine learning method where an intelligent agent

(computer program) interacts with the environment and learns to act within that."

- Reinforcement learning establishes a pattern with the help of its experiences by interacting
with the environment

- Consequently, the agent performs a crucial role in reinforcement learning models.

- It aims to achieve a particular goal in an uncertain environment.

- Reinforcement learning models continue learning iteratively from their experiences while
inducing correctness to the output.

Dept. of ECE, GSSSIETW, Mysuru Page 27

 BETCK205H Introduction to IoT

Supervised Learning Unsupervised Learning

Supervised learning algorithms are trained using Unsupervised learning algorithms are trained
labeled data. using unlabeled data.

A supervised learning model takes direct The unsupervised learning model does not
feedback to check if it is predicting the correct take any feedback.
output or not.

A supervised learning model predicts the output. An unsupervised learning model finds the
hidden patterns in data.

In supervised learning, input data is provided to In unsupervised learning, only input data is
the model along with the output. provided to the model.

The goal of supervised learning is to train the The goal of unsupervised learning is to find
model so that it can predict the output when it is hidden patterns and useful insights from the
given new data. unknown dataset.

Supervised learning needs supervision to train Unsupervised learning does not need any
the model. supervision to train the model.

Supervised learning can be categorized into Unsupervised Learning can be classified as

Classification and Regression problems. Clustering and association problems.

Supervised learning can be used for those cases Unsupervised learning can be used for those
where we know the input as well as cases where we have only input data and no
corresponding outputs. corresponding output data.

A supervised learning model produces an An unsupervised learning model may give a

accurate result. less accurate result as compared to supervised
learning.

Supervised learning is not close to true Artificial Unsupervised learning is closer to true
intelligence as in this, we first train the model for Artificial Intelligence as it learns similarly to a

Dept. of ECE, GSSSIETW, Mysuru Page 28

 BETCK205H Introduction to IoT

each data, and then only it can predict the correct child who learns daily routine things by his
output. experiences.

It includes various algorithms such as Linear It includes various algorithms such as

Regression, Logistic Regression, Support Vector Clustering, KNN, and Apriori algorithm
Machine, Multi-class Classification, Decision tree,
Bayesian Logic, etc.

Dept. of ECE, GSSSIETW, Mysuru Page 29