Webinar on AI in IoT applications KCG Connect Alumni Digital Series by Rajkumar

Time : 6.30 pm – 8.30 pm
AI in IOT Applications
KCG Connect Alumni Digital Series
Date : 24.04.2020
R. Rajkumar
AP | CSE | SRM IST
KCG Alumni

R. Rajkumar
AP | CSE | SRM
KCG Connect Alumni Digital Series
AI in IOT Applications

Purpose / Objectives of the Webinar
Contents
 To get knowledge in AI
 To get knowledge in WEKA tool
 To get knowledge in IOT
 To know about Arduino & Raspberry Pi
 To know the possibilities of combining AI with IOT
Outcomes
 To plan for an IOT project
 To plan for an AI project
 To plan for an AI + IOT project

Imagine a world without Internet.!
Without Internet it is really hard to imagine life.
We are so much dependent on Internet not only in
terms of social media but professionally as well
If Internet is shut down for a day then imagine
about websites like Google, Zoom, Webex, Amazon,
Flipkart and many more.

TURING TEST
• Can Machine think ?
• A way to test machine intelligence.
• Todetermine whether responses come from
computer or human

Learning
What is Human Intelligence?
Understanding of Language FeelingPerceiving
Reasoning

For your thinking
Touch
Sight
Hearing
Smell
Taste

For your thinking
Touch
Sight
Hearing
Smell
Taste
Pressure, temperature, light touch, vibration, pain
and other sensations are all part of the touch sense

For your thinking
Touch
Sight
Hearing
Smell
Taste
A 576-megapixel resolution
People without sight may compensate
with enhanced hearing, taste, touch
and smell

Webinar on AI in IoT applications KCG Connect Alumni Digital Series by Rajkumar

For your thinking
Touch
Sight
Hearing
Smell
Taste
The inner ear is connected to the vestibulocochlear
nerve, which carries sound and equilibrium
information to the brain.

For your thinking
Touch
Sight
Hearing
Smell
Taste
Humans may be able to smell over 1 trillion scents

For your thinking
Touch
Sight
Hearing
Smell
Taste
Salty, sweet, sour, spicy and bitter.
Taste is sensed in the taste buds.
Adults have 2,000 to 4,000 taste buds

What is Artificial Intelligence?
AI = Machine Intelligence
• Artificial Intelligence is the science and engineering of making
intelligence machine
-John McCarthy
• A computer performing tasks that are normally though to require
human intelligence
• Getting a computer to do in real life what computers do in the movies

KNOWLEDGE IS
POWER
SIR FRANCIS BACON (1561 - 1626)

IS KNOWLEDGE
STILL POWER?
SIR FRANCIS BACON (1561 - 1626)

“Artificial Intelligence would be the ultimate version of Google.
We’re nowhere near doing that now.
However, we can get incrementally closer to that, and that is
basically what we work on”
- Larry Page
CEO, Google, October 2000
“We’ve been building the best AI team and tools for
years, and recent breakthroughs will allow us to do
even more.
We will move from mobile first to an AI first world.”
- Larry Page
CEO, Alphabet, April 28, 2016

Artificial Intelligence Evolution

“Success in creating AI would be the biggest event in human history.
Unfortunately, It might also be the last, unless we learn how to avoid
the risks”
- Steven Hawking

Thinking
Goals ofAI
System that think like humans System that think rationally
System that act like humans System that act rationally
Acting
Human
This model from Russell and Norving.
Rational

Thinking
Goals ofAI
System that think like humans
“Cognitive Science”
• Neuron Level
• Neuroanatomical Level
• Mind Level
Acting
Human Rational

Thinking
Goals ofAI
• Neuron Level
• Mind Level
System that act like humans
“The Turing Test”
• Understand language
• Game AI, Control NPCs
• Control the body
Acting
Human Rational

Thinking
Goals ofAI
• Neuron Level
• Mind Level
System that think rationally
“Laws of thought”
• Logic
• A is X and X are Y then A is Y
Acting
Human Rational

Thinking
Goals ofAI
• Neuron Level
• Mind Level
System that think rationally
“Laws of thought”
• Logic
• A is X and X are Y then A is Y
System that act rationally
“Doing the right thing”
• Maximize the goal achievement, given
information
• Doesn’t necessary involve thinking
• It involve solving
Acting
Human Rational

Mathematical Model / Conditions

What’s required for a machine to be Intelligent?

• Artificial Intelligence (AI) :
A field of computer science dedicated to the study of computer software making
intelligent decisions, reasoning, and problem solving.
• Machine Learning (ML) :
A field of AI focused on getting machines to act without being programmed to do
so. Machines "learn" from patterns they recognize and adjust their behavior
accordingly.
• Natural Language Processing (NLP) :
The ability of computers to understand, or process natural human languages and
derive meaning from them. NLP typically involves machine interpretation of text
or speech recognition.
AI Terms

• Data Mining :
The process by which patterns are discovered within large sets of data
with the goal of extracting useful information from it.
• Deep Learning (DL):
A subset of machine learning that uses specialized algorithms to
model and understand complex structures and relationships among
data and datasets.
AI Terms

• Algorithm:
Formula that represents a relationship between things. It’s self-contained, step-
by-step set of operations that automates a function, like a process,
recommendation or analysis.
• Neural Network:
Computational approach that loosely models how the brain solves problems with
layers of inputs and outputs. Rather than being programmed, the networks are
trained with several thousand cycles of interaction.
AI Terms

AI Terms
• Intelligent Agents - Example : Taxi Driver
• Performance Measure:
Safe, Fast, Legal, Comfortable Trip, Maximize profits
• Environment:
Roads, other traffic, pedestrians, customers
• Actuators:
Steering wheel, accelerator, brake, signal, horn
• Sensors:
Cameras, Sonar, Speedometer, GPS, engine sensor

Uses of AI spreading Rapidly andWidely

Facebook CEO Mark Zuckerberg talks about the company's 10-year road map during
@ Facebook’s F8 developers conference in April,2016

• A Sub-Field of AI
• Construction and study of systems that can learn from data
What is MachineLearning?

• Types of Learning
• Supervised Learning:
Reliance on algorithm trained by human input, reduce expenditure on manual
review for relevance and coding
• Unsupervised Learning:
High reliance on algorithm for raw data, large expenditure on manual review
for review for relevance and coding
• Semi-Supervised Learning:
Reliance on analytics trained by human input, automated analysis using
resulting model
• Reinforcement Learning:
Algorithm is continually trained by human input, can be automated once
maximally accurate

• Evolved from pattern recognition and computational learning theory
• Subfield of Artificial Intelligence
• Study of Algorithms that iteratively learn form data
• Make predictions
• Find hidden insights without explicit programming

In other words…
“Learning is any process by which a system improve
performance from experience”
“Machine Learning is concerned with computer
programs that automatically improve their
performance through experience”
- Herbert Simon

• Google
• TensorFlow
• Keras
• Facebook
• Caffe2
• Amazon
• DSSTNE
• Microsoft
• CNTK
Deep Learning Framework by Big TechCompanies
Credit - https://twitter.com/fchollet/status/776455778274250752/photo/1

Natural Language Processing(NLP)
• What is NLP?
• Study of interaction between computer and human languages
• A Sub-Field of AI
• Aim : Tobuild intelligent computer that can interact with human being like a human being !!
NLP = Computer Science + AI + Computational Linguistics
Language Language
Generation
(NLG)
Understanding
(NLU)

Top10 Hot AI Technologies
The 10 Hottest AI Technologies:
1. Natural Language Generation
2. Speech Recognition
3. Virtual Agents
4. Machine Learning Platforms
5. AI-optimized Hardware
6. Decision Management
7. Deep Learning Platforms
8. Biometrics
9. Robotic Process Automation
10. Text Analytics and NLP
38% of enterprises are already using AI,
growing to 62% by 2020

Google Voice Search
Google’s Self Driving Car Google Translate
Gmail Spam Detection

• Machine Learning has a wide spectrum of applications including:
• Spam Email Detection
• Machine Translation (Language Translation)
• Image Search (Similarity)
• Clustering (KMeans) : Amazon Recommendations
• Classification : Google News
• Text Summarization : Google News
• Fraud detection : Credit card Providers
• Speech Understanding : iPhone with Siri
• Face Dectection : Facebook’s Photo tagging
ML Applications

Amazon Automated Warehouse
WAREHOUSE LOGISTICS
Original from TEDx Manchester: AI & The Future of Work (2017), Volker Hirsch

Amazon Prime Air
POSTMAN

KALMAR AUTOMATED VEHICLES, PORT OF LOS ANGELES, CALIFORNIA
SHIPPING LOGISTICS

BOSTON DYNAMICS WILDCAT
POLICEMAN

KNIGHTSCOPE K5
SECURITY GUARD

GENERAL ATOMICS MQ-9 REAPER
SNIPER

HADRIAN X
BUILDER
TEDx Manchester: AI & The Future of Work (2017), Volker Hirsch

FLIRTEY DRONE
PIZZA DELIVERY GUY

PIZZA DELIVERY GUY

moley robotic
CHIEF

HADRIAN X
DRIVER

LONG-HAUL TRUCKING

MUSICIAN

Magnum Autonomous Tractor
FARMER

ROBEAR
Care Taker

Purpose / Objectives of the Webinar
Contents
 To get knowledge about AI
 To get knowledge in WEKA tool
 To get knowledge in IOT
 To know about Arduino & Raspberry Pi
 To know the possibilities of combining AI with IOT
Outcomes
 To plan for an IOT project
 To plan for an AI project
 To plan for an AI + IOT project

Weka Installation
Kindly download the WEKA tool
and Install in your laptop/ desktop
website : www.cs.waikato.ac.nz › weka

SELFIE TIME..!!
GET REFRESH
BREAK FOR 2 MINS

Shall we continue our learning?

Content
W h a t is WEKA?
D a t a set in WEKA
T h e Explorer:
Preprocess data
Classification
Clustering
Association Rules
Attribute Selection
D a t a Visualization
8
2

What is WEKA?
Waikato Environment for Knowledge Analysis
 I t ’ s a data mining/machine learning tool developed by
Department of Computer Science, University of Waikato, New
Zealand.
 W e k a is a collection of machine learning algorithms for
data mining tasks.
 W e k a is open source software issued under the GNU
8
3

Main Features
 4 9 data preprocessing tools
 7 6 classification/regression algorithms
 8 clustering algorithms
 3 algorithms for finding association rules
 1 5 attribute/subset evaluators + 10 search
algorithms for feature selection
8
5

Main GUI
• Three graphical user interfaces
• “The Explorer” (exploratory data analysis)
• “The Experimenter” (experimental environment)
• “The KnowledgeFlow” (new process model inspired
interface)
• Simple CLI (Command prompt)
• Offers some functionality not available via the GUI
8
6
01/07/13

Datasets in Weka
Each entry in a dataset is an instance of the java class:
weka.core.Instance
Each instance consists of a number of attributes
Nominal: one of a predefined list of values
e.g. red, green, blue
Numeric: A real or integer number
String: Enclosed in “double quotes”
Date
Relational

ARFF File
(Attribute – Relation File Format)
@relation weather
@attribute outlook { sunny, overcast, rainy }
@attribute temperature numeric
@attribute humidity numeric
@attribute windy { TRUE, FALSE }
@attribute play { yes, no }
@data
sunny, 85, 85, FALSE, no
sunny, 80, 90, TRUE, no
overcast, 83, 86, FALSE, yes
rainy, 70, 96, FALSE, yes
rainy, 65, 70, TRUE, no
overcast, 64, 65, TRUE, yes
sunny, 72, 95, FALSE, no
sunny, 69, 70, FALSE, yes
sunny, 75, 70, TRUE, yes
overcast, 72, 90, TRUE, yes
overcast, 81, 75, FALSE, yes
rainy, 71, 91, TRUE, no

WEKA: Explorer
Pre-process: Choose and modify the data being acted on.
Classify: Train and test learning schemes that classify or
perform regression.
Cluster: Learn clusters for the data.
Associate: Learn association rules for the data.
Select attributes: Select the most relevant attributes in the
data.
Visualize: View an interactive 2D plot of the data.

Explorer: pre-processing the data
90
Data can be imported from a file in various formats: ARFF, CSV,
C4.5, binary
Data can also be read from a URL or from an SQL database
(using JDBC)
Pre-processing tools in WEKA are called “filters”
WEKA contains filters for:
Discretization, normalization, resampling, attribute selection,
transforming and combining attributes, …

@relation heart-disease-simplified
@attribute age numeric
@attribute sex { female,
male}
@attribute chest_pain_type { typ_angina, asympt, non_anginal,
atyp_angina} @attribute cholesterol numeric
@attribute exercise_induced_angina {
no, yes} @attribute class { present,
not_present}
@data
63,male,typ_angina,233,no,not_p
resent
67,male,asympt,286,yes,present
67,male,asympt,229,yes,present
38,female,non_anginal,?,no,not_
present
WEKA only deals with “flat” files

15 University of Waikato
01/07/13
load
filter
analyze

Explorer: building “classifiers”
Classifiers in WEKA are models for predicting
nominal or numeric quantities
Implemented learning schemes include:
Decision trees and lists, instance-based classifiers,
support vector machines, multi-layer perceptrons,
logistic regression, Bayes’ nets, …

age income student credit_rating buys_computer
<=30 high no fair no
<=30 high no excellent no
31…40 high no fair yes
>40 medium no fair yes
>40 low yes fair yes
>40 low yes excellent no
31…40 low yes excellent yes
<=30 medium no fair no
<=30 low yes fair yes
>40 medium yes fair yes
<=30 medium yes excellent yes
31…40 medium no excellent yes
31…40 high yes fair yes
>40 medium no excellent no
112
Decision Tree Induction: Training Dataset

age?
overcast
student? credit rating?
<=30 >40
no yes yes
yes
31..40
113
fairexcellentyesno
Output: A Decision Tree for “buys_computer”

 B a s i c algorithm (a greedy algorithm)
 T r e e is constructed in a top-down recursive divide-
and-conquer manner
 A t start, all the training examples are at the root
Attributes are categorical (if continuous-valued, they
are discretized in advance)
Examples are partitioned recursively based on
selected attributes
 Te s t attributes are selected on the basis of a
heuristic or statistical measure (e.g., information gain)
114
Algorithm for Decision Tree Induction

University of Waikato123 01/07/13

01/07/1
3
63
Explorer: clustering data
WEKA contains “clusterers” for finding groups of similar
instances in a dataset
Implemented schemes are:
k-Means, EM, Cobweb, X-means, FarthestFirst
Clusters can be visualized and compared to “true” clusters
(if given)
Evaluation based on loglikelihood if clustering scheme
produces a probability distribution

65
 G i v e n k, the k-means algorithm is implemented in four
steps:
Partition objects into k nonempty subsets
 Compute seed points as the centroids of the clusters of
the current partition (the centroid is the center, i.e., mean
point, of the cluster)
 Assign each object to the cluster with the nearest seed
point
 G o back to Step 2, stop when no more new assignment
The K-Means Clustering Method

66
Explorer: finding associations
 W E K A contains an implementation of the Apriori algorithm
for learning association rules
 Wo r k s only with discrete data
 C a n identify statistical dependencies between groups
of attributes:
 m i l k , butter  bread, eggs (with confidence 0.9 and support
2000)
Apriori can compute all rules that have a given minimum
support and exceed a given confidence

75
Explorer: attribute selection
Panel that can be used to investigate which
(subsets of) attributes are the most predictive
ones
Attribute selection methods contain two parts:
A search method: best-first, forward selection,
random, exhaustive, genetic algorithm,
ranking
An evaluation method: correlation-based,
wrapper, information gain, chi-squared, …
Very flexible: WEKA allows (almost) arbitrary
combinations of these two

85
Explorer: data visualization
Visualization very useful in practice: e.g. helps
to determine difficulty of the learning problem
 W E K A can visualize single attributes (1-d)
and pairs of attributes (2-d)
 T o do: rotating 3-d visualizations (Xgobi-style)
Color-coded class values
“Jitter” option to deal with nominal attributes
(and to detect“hidden” data points)
“Zoom-in” function

References and Resources
References:
 W E K A website:
http://www.cs.waikato.ac.nz/~ml/weka/index.html
 W E K A Tutorial:
 Machine Learning with WEKA: A presentation demonstrating all graphical user
interfaces (GUI) in Weka.
 A presentation which explains how to use Weka for exploratory data mining.
 W E K A Data Mining Book:
 I a n H. Witten and Eibe Frank, Data Mining: Practical Machine Learning
Tools and Techniques (Second Edition)
 W E K A Wiki:
http://weka.sourceforge.net/wiki/index.php/Main_Page
Others:
 J ia we i Han and Micheline Kamber, Data Mining: Concepts and
Techniques, 2nd ed.

“anytime, any place,
anything in connectivity.”
IOT

Introduction to IOT
Internet of Things (IoT) is the networking of physical objects that
contain electronics embedded within their architecture in order to
communicate and sense interactions amongst each other or with
respect to the external environment.
In the upcoming years, IoT-based technology will offer advanced
levels of services and practically change the way people lead their
daily lives.
Advancements in medicine, power, gene therapies, agriculture, smart
cities, and smart homes are just a very few of the categorical
examples where IoT is strongly established.

Over 9 billion ‘Things’ (physical objects) are
currently connected to the Internet, as of now.
In the near future, this number is expected to
rise to a whopping 20 billion.
According to Business Insider, there will be
more than 64 billion IoT devices by 2025

Four main components used in IoT:
Low-power embedded systems
Cloud computing
Availability of big data
Networking connection

Low-power embedded systems
Less battery consumption, high performance are the
inverse factors play a significant role during the design of
electronic systems. looking for another alternative naming
system to represent each physical object.

Cloud computing
Data collected through IoT devices is massive and this data
has to be stored on a reliable storage server. This is where
cloud computing comes into play. The data is processed
and learned, giving more room for us to discover where
things like electrical faults/errors are within the system.

Availability of big data
We know that IoT relies heavily on sensors, especially
real-time. As these electronic devices spread throughout
every field, their usage is going to trigger a massive flux of
big data.

Networking connection
In order to communicate, internet connectivity is a must
where each physical object is represented by an IP
address. However, there are only a limited number of
addresses available according to the IP naming.

How it Works?
The entire IOT process starts with the devices themselves
like smartphones, smartwatches, electronic appliances like
TV, Washing Machine which helps you to communicate with
the IOT platform.

Fundamental components of an IoT system
1. Sensors/Devices:
2. Connectivity:
3. Data Processing:
4. User Interface:

1. Sensors/Devices:
Sensors or devices are a key component that helps you to
collect live data from the surrounding environment. All this
data may have various levels of complexities. It could be a
simple temperature monitoring sensor, or it may be in the
form of the video feed.
A device may have various types of sensors which performs
multiple tasks apart from sensing.
Sensors can be either standalone devices or are embedded in
ordinary objects or machines to make them smart.

Types of Sensors
Temperature sensors
Moisture IoT sensors
Light IoT sensors
Acoustic & noise IoT sensors
Water level IoT sensors
Presence & proximity IoT sensors
Motion IoT sensors
Gyroscope IoT sensors
Chemical IoT sensors
Image IoT sensors

Temperature sensors
This most basic type of sensor finds its application in every
kind of IoT use case where keeping track of thermal
conditions of air, work environment, machines or other
objects is vital.
Temperature sensors are particularly useful in manufacturing
plants, warehouses, weather reporting systems and
agriculture, where soil temperature is monitored to provide
balanced and maximised growth.

Moisture IoT sensors
While their most obvious and widespread use is in
meteorology stations to report and forecast weather, quite
surprisingly, moisture and humidity sensors are also being
extensively employed in agriculture, environment
monitoring, food supply chain, HVAC and health monitoring.

Light IoT sensors
Depending on ambient light intensity, smart TVs, mobile
phones or computer screens are able to adjust their
brightness thanks to light sensors, yet sensors for detecting
ambient light are not only commonplace in consumer
electronics, but also smart city applications. They are
increasingly used for adapting street lights or urban lighting
levels for increased economy.

Acoustic & noise IoT sensors
Smart acoustic sensors enable to monitor the level of noise
in a given environment.
Being able to measure and provide data to help noise
pollution prevention, acoustic IoT sensor systems are gaining
ground in smart city solutions.

Water level IoT sensors
To prevent natural disasters, data gathered by the water
level monitoring sensors can be used in flood warning
systems for analytics and prediction.
Apart from environmental protection, this sensor finds its
use in a variety of industrial applications to control and
optimise manufacturing processes.

Presence & proximity IoT sensors
By emitting an electromagnetic radiation beam, this type of
sensor is capable of sensing its target object presence and
determining the distance that separates both.
With their high reliability and long life, it is no wonder that
they have quickly made it into so many IoT sectors, such as
smart cars, robotics, manufacturing, machines, aviation, and
even smart parking solutions.

Motion IoT sensors
A smart building system is probably having IoT application
for the motion sensor to imagine.
While this obviousness holds largely true, apart from helping
to monitor private or public spaces from intrusion and
burglary, the use of motion sensors use is extending to
energy management solutions, smart cameras, automated
devices and many others.

Gyroscope IoT sensors
The task of this type of sensor is to detect rotation and
measure angular velocity, which makes it perfect for
navigation systems, robotics, consumer electronics and
manufacturing processes involving rotation.
For a more day-to-day IoT application, gyroscope sensors are
increasingly installed in IoT devices used by athletes for
accurate measurements of body movements to analyse and
improve their sports performance.

Chemical IoT sensors
Sensors to detect chemical compounds (solids, liquids, and
gases) are indispensable elements in industrial security
systems, environmental protection solutions, and, quite
obviously, scientific research.
Moreover, they have already gained a foothold in IoT-
supported air quality monitoring which helps cities and
states fight harmful impact of air and water pollution.

Image IoT sensors
Converting optical data to electrical impulses, an image
sensor enables the connected object to view the
environment around it and act upon it using the intelligence
obtained from the analysis of data provided.
Image sensors are used whenever there is a need for the
smart device to ‘see’ its immediate surroundings, which
includes smart vehicles, security systems, military
equipment like radars and sonars, medical imaging devices
and, of course ____________.

Actuators
Linear actuators—these are used to enable motion of object
or element in a straight line.
Motors—they enable precise rotational movements of device
components or whole objects.
Relays—this category includes electromagnet-based
actuators to operate power switches in lamps, heaters or
even smart vehicles.
Solenoids—most widely used in home appliances as part of
locking or triggering mechanisms, they also act as controllers
in IoT-based gas and water leak monitoring systems.

Satellite Sensors
Mechanical structure
Propulsion subsystem
Thermal control subsystem
Power supply subsystem
Telemetry, tracking and command (TT&C) subsystem
Attitude and orbit control subsystem
Payload subsystem
Antenna subsystem

Connectivity:
All the collected data is sent to a cloud infrastructure.
The sensors should be connected to the cloud using various
mediums of communications.
These communication mediums include mobile or satellite
networks, Bluetooth, WI-FI, WAN, etc.

Data Processing:
Once that data is collected, and it gets to the cloud, the
software performs processing on the gathered data.
This process can be just checking the temperature, reading
on devices like AC or heaters.
However, it can sometimes also be very complex like
identifying objects, using computer vision on video.

Input
For any processing to occur, input must be available. The
data collected may be in the form of images, QR codes,
text, numbers, or even videos. All these data must be
converted into machine readable form before they can be
sent for processing.

Process
This is the phase where the actual data processing happens.
Different techniques like classification, sorting, calculations,
etc. are used to get meaningful information from the data
received.

Output
Although the information is produced in the processing
phase itself, it is rendered into human readable format in
the output stage. This output maybe in the form of text,
graphs, tables, audio, video, etc.
Output may also be stored as data for further processing at
a later date. This is essential because comparison of current
information with historical data can produce useful insights
into the overall functioning of a system. ( This comparison
can also be used to predict future behaviour ~ ML )

Best Tools for IoT Data Processing
Here are some of the best tools and platforms being used for IoT
data processing.
 Google cloud
 IBM Watson IoT
 Amazon AWS IoT Core
 Microsoft Azure IoT suite
 Oracle IoT
 Cisco IoT Cloud Connect

IOT Software’s
IoT software encompasses a wide range of software and
programming languages from general-purpose languages
like C++ and Java to embedded-specific choices like
Google’s Go language or Parasail.

C & C++
The C programming language has its roots in embedded
systems—it even got its start for programming telephone
switches.
It can be used almost everywhere and many programmers
already know it. C++ is the object-oriented version of C,
which is a language popular for both the Linux OS and
Arduino embedded IoT software systems.
These languages were basically written for the hardware
systems which makes them so easy to use.

Java
While C and C++ are hardware specific, the code in JAVA is
more portable. It is more like a write once and read
anywhere language, where we can install libraries, invests
time in writing codes.

Python
There has been a recent surge in the number of python users
and has now become one of the best languages in Web
development.
Its use is slowly spreading to the embedded control and IoT
world—specially the Raspberry Pi processor.
Python is an interpreted language, which is, easy to read,
quick to learn and quick to write. Also, it’s a powerhouse for
serving data-heavy applications.

User Interface:
The information needs to be available to the end-user in
some way which can be achieved by triggering alarms on
their phones or sending them notification through email or
text message.
The user sometimes might need an interface which actively
checks their IOT system.
For example, the user has a camera installed in his home.
He wants to access video recording and all the feeds with
the help of a web server.

Choosing the IOT Platform
Choosing an IoT platform is a pre-requisite for
beginning the development of an end-to-end IoT
solution.

Arduino
Arduino is a microcontroller board that is used for dedicated
applications;
for example:
Actuating small devices like motors, sensors, and lights.
The Arduino UNO runs on an 8-bit ATmega328 chip at only 20
MHz.
www.arduino.cc

Raspberry Pi
Raspberry Pi has a microcontroller, HDMI ports, and RAM.
Which means that; with basic coding knowledge.
we can configure an OS on Raspberry Pi and use it as a media
streaming device, running a web server, or VPNs.
Raspberry Pi runs on an ARM 11 CPU with 700 MHz
www.raspberrypi.org

Comparison
Pi comes complete with ports, like USB, RJ45, HDMI
and an SD card reader.
But Arduino depends more on external interfaces to
provide the necessary connections.
Pi is essentially a mini computer so it is more
expensive than Arduino

Is it possible to work together?
For a smart home applications we can choose Raspberry PI to
be a central server, in charge of communication, data
collection and storage from the Arduino, dealing with
massive data workload (such as media processing), and
handling data from mobile apps to make it more convenient
to control applications.
Raspberry Pi can work with Arduino Ethernet and Zigbee on
data transmission.

Endless Possibilities of combining AI with IOT

Dataset
http://www.timeseriesclassification.com/dataset.php

a. Smart Home
The estimated amount of funding for Smart Home startups
exceeds $2.5bn and is ever growing.
The list of startups includes prominent startup company
names such as AlertMe or Nest as well as a number of
multinational corporations like Philips, Haier, or Belkin etc.

b. Wearable devices
It make our life easier and healthier

c. Smart City
IoT solutions offered in the Smart City area solve various
city-related problems comprising of traffic, reduce air and
noise pollution and help make cities safer.

d. Smart Grids
Electricity suppliers in an automated fashion in order to
improve the efficiency, economics, and reliability of
electricity distribution.
41,000 monthly Google searches is a testament to this
concept’s popularity.

e. Industrial Internet of Things
One way to think of the Industrial Internet is, as connecting
machines and devices in industries such as power
generation, oil, gas, and healthcare.

f. Connected Car
Connected car technology is a vast and an extensive network
of multiple sensors, antennas, embedded software, and
technologies that assist in communication to navigate in our
complex world.

g. Connected Health (Digital Health/Tele-
health/Telemedicine)
IoT has various applications in healthcare, which are from
remote monitoring equipment to advance & smart sensors to
equipment integration.
It has the potential to improve how physicians deliver care
and also keep patients safe and healthy

h. Smart Retail
Retailers have started adopting IoT solutions and using IoT
embedded systems across a number of applications that
improve store operations such as increasing purchases,
reducing theft, enabling inventory management, and
enhancing the consumer’s shopping experience.

i. Smart Supply Chain
With an IoT enabled system, factory equipment that contains
embedded sensors communicate data about different
parameters such as pressure, temperature, and utilization of
the machine.

Outcomes
we learned how IoT works and an entire IOT
system functions.
Also, we discussed some real-life examples
where we can use AI for IoT. We will be learning
more about IOT and AI in detail by start doing
projects.

My research works on IOT and AI
Current Projects:
1. Got fund Rs. 9.98 L for Mixed Reality application in Medical Education at SRM Medical
College hospital and Research Centre, Kattankulathur.
2. Got fund Rs. 3.14 L for IOT Game simulator for paralysed patients by Department of
Science and Technology through SIIC.
3. Executive Member in SRM IOT Centre of Excellence.
4. Core committee member of Industry Institution Summit, IET India.
Completed Projects
1. AR – IOT – Smart Home Automation (Patent Published)
2. IOT Speed Breaker for Indian Roads (Patent Published)
3. IOT Dental Chair with Machine Learning (Patent Applied)
4. Smart Horn System (Patent Applied)
5. Keypod (Patent Granted | KCG )

Publications in AI and IOT
International Journal Publications ( Scopus / SCI )
R. Rajkumar, V. Ganapathy, BIO-INSPIRING LEARNING STYLE CHATBOT INVENTORY USING BRAIN COMPUTING
INTERFACE TO INCREASE THE EFFICIENCY OF E-LEARNING, IEEE – ACCESS. ACCEPTED AT MARCH 20, 2020, ID ACCESS-
2020-13266.
R. Rajkumar, V. Ganapathy, CUSTOMIZED SPORTS E-LEARNING PLATFORM: FOR MAKING TEACHING AND LEARNING
MOBILE BASED APPLICATION, International Journal of Pure and Applied Mathematics Volume 119 No. 16 2018, 3635-
3643.
R. Rajkumar, V. Ganapathy, BRAIN COMPUTER INTERFACE TO IDENTIFY THE STATE OF MIND OF THE LEARNERS,
International Journal of Pure and Applied Mathematics Volume 119 No. 16 2018, 3645-3654.
R. Rajkumar, V. Ganapathy, DETECTION OF PANIC AND RECOVERY FROM PANIC USING BRAIN COMPUTER INTERFACE,
International Journal of Pure and Applied Mathematics Volume 119 No. 16 2018, 3655-3662.
R. Rajkumar, V. Ganapathy, VIRTUAL REALITY MULTIPLE QUESTIONNAIRES EXAMINATION PLATFORM, Dubai
International education Conclave, Curtin University, Dubai, 2018.
R. Rajkumar, V. Ganapathy, BIO INSPIRED BLOOD GROUP PREDICTION, International Journal of Control Theory and
Applications. Vol 2. 13. 2017.
R. Rajkumar, N. Vivekananthamoorthy, DETERMINANT FACTORS ON STUDENT EMPOWERMENT AND ROLE OF SOCIAL
MEDIA AND EWOM COMMUNICATION: MULTIVARIATE ANALYSIS ON LINKEDIN USAGE. Indian Journal of Science and
Technology, Vol 9(25), DOI: 10.17485/ijst/2016/v9i25/95318, July 2016
R. Rajkumar, N. Vivekananthamoorthy, THE ROLE OF SOCIAL NETWORKING SITES AND EWOM COMMUNICATION IN
ENHANCING STUDENT ENGAGEMENT IN CURRENT LEARNING SCENARIOS, 2015.
R. Rajkumar, M. Karthikeyan, N. Vivekantha Moorthy, COGNITIVE SEARCH E-LEARNING FRAMEWORK: AN EFFECTIVE
ASSESSMENT BASED LEARNING SYSTEM. International Journal of Computer Applications (0975 – 8887), 2015.

Thank you..!
Feedback & Questions?
R. Rajkumar
rajkumar03r@gmail.com
+91 9894808403.
https://www.linkedin.com/in/raj-kumar-1251a553/

Webinar on AI in IoT applications KCG Connect Alumni Digital Series by Rajkumar

More Related Content

What's hot (19)

Similar to Webinar on AI in IoT applications KCG Connect Alumni Digital Series by Rajkumar (20)

More from Rajkumar R (13)

Recently uploaded (20)

Webinar on AI in IoT applications KCG Connect Alumni Digital Series by Rajkumar