Speech & Web based appliance control

International Journal of Science, Engineering and Technology Research (IJSETR), Volume 5, Issue 3, March
2016
1
ISSN: 2278 – 7798 All Rights Reserved © 2016 IJSETR
Abstract - Speech recognition is a familiar and
accessible concept to most people in western society. All
that is needed in order to use speech recognition is a
smart phone with some form of speech recognition
application. Home automation might not be as well-
known as speech recognition but it is still an area that
offers a lot of products and solutions for people that are
interested in a more comfortable or manageable
lifestyle. Most home systems come in complete solutions
with different kinds of features and possibilities. The
design and implementation of Secure Home
Automation using Raspberry Pi 2 for mobile devices
that leverage mobile technology to provide essential
security to our homes and associated control
operations is provided. Sound intelligence is added to
home automation for physically challenged people as a
broad perspective of the system. Raspberry Pi
manages operations on home appliances, such as
turning ON/OFF a television, lights and other
electronic/electrical devices. According to the received
texts, appliances can be controlled. The entire above
are implemented in a low cost Raspberry Pi 2 board.
Index Terms—Voicecommand, Web based
control, Raspberry Pi 2, Google API, NGINX, Wiring
Pi.
I. INTRODUCTION
Manually controlling various devices found in
one’s home might be tiresome. This system is able
to ease a user’s interaction with home equipment
by providing a new, easy to use, interface. The
initial phase of the development involved surveying
the current field by conducting studies on speech-
recognition software and the needs of potential
users. These studies provided a foundation for the
prototype requirements that acted as the goal for
the implementation phases that followed. The paper
also presents the finished prototype, which is a
system capable of interpreting speech as
commands. These commands can be formulated in
several different ways to make the interaction more
natural. After a command has been interpreted an
appropriate action is taken on the correct home
equipment. In order to control a number of
different equipment the control is achieved through
radio-controlled outlets or multi-channel relay
boards. The use of outlets enables the prototype to
control most equipment that is controlled with on
and off switches. A further development carried out
is to control the equipments through web as well.
II. Implementation Methodology
A. Software implementation
The below given figure shows the basic block
diagram of speech recognition system [11]. To
detect the speech signal electric microphone is
used. The microphone can be either a dedicated
USB microphone or a usb powered headset
microphone. The speech signal fed using the mic is
converted to a text signal. There are various speech
to text conversion engines available, but Google
STT (speech to text) proves to be the fastest and the
most accurate one. Hence the Google STT engine
is used. The conversion of this speech signal to text
signal is carried out online by the Google web
server. Once it is converted to text signal, it is
matched with a configuration file stored in
raspberry pi. This configuration file has a list of
commands that should be recognised properly and
to every command there is a function to be carried
out. If the speech signal was recognised and
converted properly into a text signal, it shall match
with the command stored in the configuration file
and the necessary function will be carried out For
example: If the user says “Turn ON the lights” then
this signal is first converted to text signal and then
assigned to specific task i.e. turning on the light. In
this way a dictionary of commands (configuration
file) is created which will be used to perform the
various automation tasks. If a command said by the
user is not present in the pre-defined dictionary
(configuration file), no operation will be carried out
and the user will be prompted “Command not
recognised” by raspberry pi. The environment
created is very user friendly where the hardware
also communicates with the user.
Figure.1 Block diagram of speech-recognition system [11]
Speech & Web Based Appliance Control
Sanjay Mirchandani, Mohit Khatri, Kinjal Vaghamshi, Aditya Nair, Sonali Talreja

2016
2
B. Hardware implementation
The heart of this project is its mini-computer i.e.
Raspberry Pi 2 [1]. The Raspberry Pi is a low
cost, credit-card sized computer that plugs into a
computer monitor or TV, and uses a standard
keyboard and mouse. It is a capable little device
that enables people of all ages to explore
computing, and to learn how to program in
languages like Scratch and Python. It’s capable of
doing everything one would expect a desktop
computer to do, from browsing the internet and
playing high-definition video, to making
spreadsheets, word-processing, and playing games.
What’s more, the Raspberry Pi has the ability to
interact with the outside world, and has been used
in a wide array of digital maker projects, from
music machines and parent detectors to weather
stations and tweeting birdhouses with infra-red
cameras.
Figure.2 Raspberry pi 2 [1]
A small relay circuit has been made to switch an
appliance ON/OFF [10]. It consists of a BC547
transistor, which acts like a switch. The transistor
triggers the relay when it receives a voltage at its
base from the GPIO pin of raspberry pi. A 1N4001
diode is used to protect the transistor from reverse
voltages created in the relay coil. The circuit can be
powered either using the 5V supply from the
raspberry pi or using an external battery. The GPIO
4 (pin 7) of raspberry pi should be connected to the
base of transistor and pin 6 (GND) of raspberry pi
should be grounded.
Figure 3 Relay Board [10]
Figure 4 Interfacing relay with raspberry pi [6]
(Note: The above connection shows a Raspberry pi
board. It remains the same for Raspberry pi 2 board)
C. Flow of the Project
1. Receive input speech using external MIC
connected to Raspberry pi
2. Convert the speech signal to text signal
using online Google API speech engine.
3. Process the text signal, match it with the
stored command file. If they match, send it
to GPIO (General Purpose Input Output)
pin on Raspberry pi
4. Perform the required operation as per the
text signal (Example: If one says Lights
ON, then the light connected to the
Raspberry pi will turn ON)

2016
3
5. A relay board is connected to GPIO pins
to connect multiple devices to Raspberry
pi
Figure 5 Project Flow [14]
II. SETTING UP RASPBERRY PI
Raspberry pi works on Raspbian operating system
which works best for beginners. Below mentioned
are the steps for installing Raspbian on Raspberry
pi [2].
1. Format an SD card which is 4 GB or
greater with FAT 32 format
2. Download and extract files from the zip
file of operating system
3. Copy the extracted files onto the SD card
that you just formatted.
4. On the first boot the ‘Recovery FAT ’
partition will be automatically resized to a
minimum size
Now once the Raspberry pi is setup, various
dependencies will be needed to perform speech
recognition and controlling the GPIO pins.
Following dependencies will be required to start
with speech recognition and control the hardware:
Wiring Pi:-This will be required to control the
GPIO pins i.e. to turn high or low a particular pin
from the terminal window of raspberry pi [4].
Voicecommand:-Voicecommand is based on
Google speech to text engine. This dependency will
be required to recognise the speech commands,
convert them to text, and then perform the required
operations. A list of commands called configure
file is created so that the user is familiar about the
commands recognised by Raspberry pi [5]
A. Installing Wiring Pi
The java script that controls the electrical devices
relies on the Wiring Pi project to switch the GPIO
pins high and low. Run the following commands to
download and install Wiring Pi. Executing the
following commands in the terminal shell
(command window) will install Wiring Pi on
Raspberry Pi [4]:
1. sudo apt-get install git-core
2. git clone git://git.drogon.net/wiringPi
3. cd wiring Pi
4. ./build
B. Installing Voicecommand
Voice Command is a program that will run on the
Raspberry Pi and listen for a keyword for
addressing it. After hearing its keyword or name,
the Voice Command will prompt for an instruction
by default saying "Yes sir". It will then attempt to
process whatever is said depending on how it is
configured. Type the following commands in the
terminal shell (command window) to install
Voicecommand on Raspberry Pi [5]:
1 sudo apt-get install git-core
2 git clone
git://github.com/StevenHickson/PiAUISuite.git
3 cd PiAUISuite/Install/
4 ./InstallAUISuite.sh
III. FURTHER WORK
After the successful implementation of speech
based device control, raspberry pi has been made to
act as a web server so that it could be remotely
accessed. An HTML page is created and two
buttons are added to turn ON or OFF a device [10].
To make raspberry pi act as a web server, NGINX
application is used. NGINX (pronounced engine x)
[7] is a popular lightweight web server application
that can be installed on the Raspberry Pi to allow it
to serve web pages. Like Apache, NGINX can
serve HTML files over HTTP, and with additional
modules can serve dynamic web pages using
scripting languages such as PHP [8].
Figure .6 Web based control [10]

2016
4
The whole system is composed of two parts: Server
and Client. Here, the server is the web interface
consisting of buttons and UI (User Interface) that
will turn ON/OFF a device. It consists of PHP files,
Html files and a .txt file (to store data).The server
usually stores information regarding the button
press on the page (ON/OFF) on a .txt file. This is a
simple Html file consisting of two buttons. The
clicking of the buttons will trigger the execution of
a PHP file called button.php [10]. This program
serves as an API (Application Programming
Interface) to store data on to a text file called
buttonStatus.txt. The data is a string : “ON”, if ON
button is clicked and “OFF”, if OFF button is
clicked. Thus the current button press state is
recorded in the text file called buttonStatus.txt.
The client side consists of a Raspberry Pi with a
relay circuit connected to its GPIO pin. Raspberry
pi runs a program which is used to ‘Post’ a url link
using urllib2.That is, the raspberry pi constantly
reads the contents of a url link. Here, the url link is
another PHP file called buttonStatus.php. This PHP
file serves as an API to read the contents of the text
file buttonStatus.txt. After reading the data, the
program checks if the string obtained is “ON” /
“OFF” based on which it switches ON/OFF the
relay respectively via its GPIO pin.
Below given Figure 7 shows the screenshot of the
GUI for two devices. A single toggle button is used
to turn ON/OFF a device instead of using two
separate buttons. One button will turn ON/OFF the
lights in the room, whereas the other button will
turn ON/OFF the Television in the room. On
similar basis a GUI multiple devices can be
developed to control every device at home as
shown in Figure 8.
Figure 7 Screenshot of mobile web browser
Figure 8 GUI for multiple device control
IV. CONCLUSION
After carrying out logical analysis of the project,
certain conclusions were made. The time taken for
speech to text translation depends on the internet
connectivity speed. If the speed is slow, the time
required for the commands to get recognized and
the operation to be performed is high. Since this
conversion is carried out by Google speech to text
engine, this system will not work offline. The
accuracy of the system is high. The words get
properly recognized even is spoken very quickly.
The system can be made portable by using a wifi
dongle instead of an Ethernet cable. This systemis
made from easily available hardware which is quite
affordable. Features like security can be added. The
systemcan be made to respond and operate only on
voice of a single person which makes it user
dependent. A mobile application can be developed
to control the devices from home and also from
remote location. The possibilities are immense. The
field of IoT (Internet of Things) can be
incorporated along with speech recognition to
widen the scope of this project.
REFERENCES
[1].https://www.raspberrypi.org/help/noobs-setup/
[2].https://www.raspberrypi.org/help/quick-start-
guide/
[3].https://www.raspberrypi.org/documentation/set
up/
[4].http://computers.tutsplus.com/articles/how-to-
control-a-lamp-with-voice-commands-and-a-
raspberry-pi--mac-60825
[5].http://stevenhickson.blogspot.in/2013/06/installi

2016
5
ng-and-updating-piauisuite-and.html
[6].http://www.instructables.com/id/Controlling-
AC-light-using-Arduino-with-relay-
modu/step2/Connections/
[7].https://www.raspberrypi.org/documentation/re
mote-access/web-server/nginx.md
[8].https://www.youtube.com/watch?v=EAMLwbS
hFFQ
[9].https://www.youtube.com/watch?v=xt-
rrH04O2g
[10].http://diyhacking.com/raspberry-pi-home-
automation/
[11].http://diyhacking.com/best-voice-recognition-
software-for-raspberry-pi/
[12]Dhawan S. Thakur and Aditi Sharma”Voice
recognition wireless home automation system
based on zigbee” Journal of Electronics and
Communication Engineering research volume
6,issue 8, May-june2013.
[13].S.Suresh and Y.Sindhuja Rao “Modelling of
Secured Voice Recognition Based Automatic
Control System” Journal of Emerging Technology
in Computer Science & Electronics, Volume 13,
Issue 2, March 2015
[14].Thoraya Obaid, Haliemah Rashed, Ali Abu El
Nour, Muhammad Rehan, Mussab Muhammad
Saleh, and Mohammed Tarique,” ZIGBEE BASED
VOICE CONTROLLED WIRELESS SMART
HOME SYSTEM” International Journal of
Wireless & Mobile Networks (IJWMN) Vol. 6, No.
1, February 2014 [3].
[15].Y.Usha Devi 1, II M.Tech, Wireless Home
Automation System Using ZigBee, International
Journal of Scientific & Engineering Research
Volume 3, Issue 8, August-2012
[16]Textbook on “Getting started with Raspberry
Pi” by Matt Richardson & Shaw Wallace
Mr. SanjayMirchandani is workingas
an Assistant Professor in Electronics &
Telecommunication Department in
V.E.S. Institute of Technology,
Chembur, Mumbai from July 2013. He
has passed B.E. in 2008 and M.E. in
2010. He has published one paper in
IEEE International Conference andthree
papers at the National level Conferences.
His field of interest is System Design,
Development & Integration of
Microwave Electronic based systems which includes RF &
Digital Signal Processing
Mr. Mohit Khatri is currentlypursuing
B.E in Electronics &
Telecommunication from V.E.S.
Institute of Technology, Mumbai and
will be completing the course in May
2016. He has published a technical
paper on “Bionic Eye- A Vision of
Hope” in the International Journal of
Scientific EngineeringandTechnology
Research (IJSETR), ISSN 2319-8885,
Vol 4 Issue 36, September 2015. His
area of interest is Embedded Systems & Networking. He has
been an Executive Committee member in the international
student body namedIEEE for the year 2014-15.He has received
as special prize forthe project “Paraplegic Wheelchair” made
under the Research & Development (R&D) Department of
college.
Ms. Kinjal Vaghamshi is currently
pursuing B.E in Electronics &
Telecommunication from V.E.S.
Institute of Technology, Mumbai and
will be completing the course in May
2016. Her area of interest is Networking
& Switching. She has been the secretary
of a technical student body named
I.S.T.E. for the year 2012-13. She is
certified as an Embedded System
programmer by Maximus Technolabs. She has also received
academic excellence award for being the topper of EXTC
department in college in the year 2014-15
Mr. Aditya Nair is currently pursuing
B.E in Electronics & Telecommunication
Department from V.E.S. Institute of
Technology, Mumbai and will be
completingthe course in May 2016. His
area of interest is Microwave & Radar
Engineering. He also has keen interest in
the technology of MiG & Sukhoi series
aircrafts He has been a council member in
the international student body namedIEEEforthe year 2013-14.
Ms. Sonali Talrejais currently pursuing
B.E in Electronics & Telecommunication
Department from V.E.S. Institute of
Technology, Mumbai and will be
completingthe course in May2016. Her
area of interest is Wireless
communication and also growing
knowledge in the field of Mobile
communication. She has been a state
level fencing player.

Speech & Web based appliance control

More Related Content

What's hot (7)

Viewers also liked (18)

Similar to Speech & Web based appliance control (20)

Recently uploaded (20)

Speech & Web based appliance control