Available via license: CC BY-NC 4.0
Content may be subject to copyright.
Automatic Control Based on Voice
Commands and Arduino
Alfiantin Noor Azhiimah 1,* Khusnul Khotimah1, Meini Sondang Sumbawati1,
Agus Budi Santosa1
1 Postgraduate program, Universitas Negerei Surabaya, Surabaya 60213, Indonesia
*Corresponding author. Email: alfiantinnoor.azhiimah@gmail.com
ABSTRACT
Control that is widely used today was control by voice commands. This study aimed to determine the application of
voice and Arduino-based control automation which had been developed from 2014-2020 based on 25 journals that
would be studied. Journal assessments were carried out by taking into account the similarities or differences in each
journal. The research method used in this research was literature review by reviewing 25 journals. The results obtained
from the journal review showed that the developed voice control consists of 2 types, namely voice recognition and
speech recognition. Voice recognition used an easy VR hardware device and a microcontroller, while speech recognition
used an application installed on Android, controlling can also be done remotely using Bluetooth and the internet. The
way the voice control system works was that the user given voice commands via a microphone, then the command
would be converted by the voice control module on the EasyVR module or the voice recognition module. The conversion
results were sent to the microprocessor and activate the controlled device. Whereas in remote control, the data resulting
from the conversion was also sent to the bluetooth module or NodeMCU module which was sent to the microprocessor
to activate the controlled device. The factors that affect the voice control system were clear pronunciation, pitch and
sound, microphone distance, sound source, intonation and noise level.
Keywords: Automation, Voice recognition, Speech recognition
1. INTRODUCTION
The rapid development of technology has made it
possible to control electronic devices using a variety of
control methods such as controlling electronic devices
automatically by utilizing sensors or by remote control.
The existence of this technological advancement makes
electronic device users not have to feel difficult in
controlling their electronic devices. Control that is
widely used today is control that is done by using sound.
Like controlling a smartphone to run applications
installed in it, controlling a smart home such as turning
off or turning on lights, opening or closing doors, and
controlling the robot. Control using voice is in great
demand because it has a security system that utilizes the
characteristic pattern of an individual by determining the
specific physiology of the individual so that only
individuals who meet predetermined characteristics can
control the electronic device. Control of this electronic
device is called a voice command.
Voice commands are divided into two, there is voice
recognition and speech recognition. Voice recognition is
a system to identifying someone through the voice
characteristics of that person. Characteristics can be the
intonation of the voice, the level of sound depth, and so
on. Voice recognition works by analyzing each
individual's voice features derived from anatomical and
behavioral patterns such as tone of voice, style and
speech accent. Thus, voice recognition can assist users in
matching previously validated voices, and verifying
voice match. Speech recognition is a technology that can
be used to convert voice into text. Users simply say the
word they want to write, then the system will convert the
voice into text. The use of this technology can be found
in instant messenger applications, google voice,
Microsoft dictation and so on.
Therefore, this writing is done with the aim of
knowing the application of voice-based and Arduino-
based control automation which has been developed from
Advances in Engineering Research, volume 196
International Joint Conference on Science and Engineering (IJCSE 2020)
Copyright © 2020 The Authors. Published by Atlantis Press B.V.
This is an open access article distributed under the CC BY-NC 4.0 license -http://creativecommons.org/licenses/by-nc/4.0/. 29
2014-2020 based on 25 journals to be studied. Journal
assessments are carried out by taking into account the
similarities or differences in each journal.
1.1. Literature Review
1.1.1. Voice Command
Voice command is control of electronic devices that
is carried out through commands given to the system
using the voice of the system user. Control using voice is
a control that is in great demand today in smart home
systems. This is because the operation of the system is
very easy and does not require a lot of power, which is
the main reason voice commands are suitable for smart
home systems. This voice command consists of voice
recognition and speech recognition with the speaker, for
example by using a series resistor. The exact value of the
resistor depends on the sensitivity of the headphones and
the desired output volume (usually in the order of 1-
10kΩ).
Software that can be used to connect the EasyVR
module to a PC is EasyVR Commander. There is also
SonicNet software that can be used to wirelessly connect
the EasyVR module to a PC.
1.1.2. Voice Recognition Module V3
Voice recognition module V3 is a voice recognition
module that can be used in voice recognition. This
module has the ability to accommodate a maximum of 80
voice commands with a maximum of 7 voice commands
that can be used at the same time. Each voice command
can be given in 1500ms, which is one or two words.
This module has a voltage of 4.5-5.5V with a current
of less than 40mA. The digital interface is a 5V TTL level
for the UART interface and the GPIO, while the analog
interface is a 3.5mm mono-channel microphone
connector and a microphone pin interface. The size of
this module is 31mm x 50mm with a speech recognition
accuracy of 99% when used in ideal conditions.
1.1.3. Arduino
Arduino is an electronic device designed to facilitate
the use of electronics in various fields in the form of an
open- source single-board micro controller. Arduino uno
is one of the Arduino products with an ATMega328
microcontroller. This tool is used to realize electronic
circuits ranging from simple to complex.
In its use, Arduino is combined with Integrated
Development Environment (IDE) software. IDE is a
program that can be used in designing or sketching
programs for Arduino boards. Arduino source code is
made using the C programming language called sketch.
Sketch that has been completed in the IDE can be
immediately decompiled to find out where the program
error is. If the program has no errors then the program
can be uploaded directly to the Arduino board.
2. METHOD
The research conducted was a literature research
based on the results of written works such as the results
of research that has been done. In library research,
literature search was not only for the first step in
preparing a research framework (research design) but at
the same time utilizing library sources to obtain research
data [1]. The data collection technique was carried out by
conducting a study of books and a number of articles on
controlling using voice commands and arduino. The data
analysis technique used was to describe data and
information in the form of easy-to- understand sentences
as an effort to get answers to the problems studied [2],
related to several studies on controlling using voice
commands and arduino.
The steps taken in conducting literature research were
divided into 3 steps. The first was to determine the title
of the research to be carried out, what literature sources
would be used, how much literature was used and so on.
Furthermore, the literature search was carried out in
accordance with the matter to be reviewed. Finally, a
content analysis of the easyvr 3.0
Easy vr 3.0 is a voice recognition module that can be
used easily and is compatible with almost all
applications. Easy vr 3.0 can be used with all uart
(universal asynchronous receiver-transmitter) at a
voltage of 3.3v- 5v.
Easyvr 3.0 audio output interface can be connected
directly to an 8 ohm speaker or connected with an
external audio amplifier for low impedance sound. In
addition, it can also be connected to high impedance
loads such as headphones, provided that the power output
is in accordance literature that has been obtained was
carried out related to how the relationship between the
literature obtained was. The block diagram of literature
research steps can be seen in “figure 1”.
Figure 1 Block diagram of literature research steps
The literature review in this study was conducted by
examining 25 journals from 2014-2020 which discussed
controlling using voice commands and arduino. Journal
assessments are carried out by paying attention to the
similarities or differences in each journal based on the
application of voice commands to the electronic device
to be controlled, the voice command used in the form of
voice recognition or speech recognition, the hardware
used, the tests carried out, the test results, the system
operation and the factors involved affected by system.
Advances in Engineering Research, volume 196
30
3. RESULT AND DISCUSSION
3.1. Implementation
Based on the results of the studies that have been
done, the research that many researchers have developed
was the development of a smart home prototype. The
control developed in a smart home was in the form of
controlling electronic devices so that they can be
controlled automatically by using voice commands from
the user or smart home owner. The development of
controlling using voice on this smart home could make it
easier for users to carry out daily activities. Sound-based
control of electronic devices in a smart home could be
applied to lights, fans, air conditioners, TVs, bells,
alarms, windows and doors as research conducted by
almost all researchers. The control was carried out on
lights, fans and air conditioners, namely by turning on or
off the lights, fans and air conditioners automatically by
giving commands according to predetermined codes. For
controlling that is carried out on the TV, namely by
controlling the TV turn on or off, turning off or turning
on the TV sound, displaying the TV menu, and making
the TV sound in mute mode. Control of the bell and alarm
was done by giving commands to the system to turn off
or turn on the bell and alarm. Control of windows and
doors was done by giving orders to open or close
windows and doors with a predefined code.
In addition to developing a prototype on a smart
home, there were researchers who develop it with the
aim of helping users with disabilities and smart hospital
rooms. Controls developed to help with disabilities were
controlling the wheelchair and controlling the printing of
documents using braille. Control on smart chairs for
physically disabled people who couldnot move a
wheelchair by hand developed by Aktar [3]. This
wheelchair was also equipped with a GPS module that
can be used to track and send information to smartphone
applications via firebase. Controlling the printing of
documents using braille is done with the aim of
developing a voice-controlled braille printer developed
by Manigiri [4]. This printer would help users with
disabilities print a spoken voice into a document in
braille format. For controlling the prototype smart
hospital room, the application was carried out in
controlling the patient bed in the hospital and controlling
the device to call nurses developed by Agustin [5]. The
patient's bed could be adjusted using voice commands in
the sitting, intriguing and sleeping position, while the
buzzer is used as an assist button to call the nurse. Thus
controlling the prototype smart hospital room would
make it easier for patients.
3.2 Voice Command
The differences in research that was conducted were
at the voice command used, namely using voice
recognition or speech recognition. Researchers used
voice commands in the form of voice recognition with
the hope that the smart home prototype being developed
can have a better level of security. This is because voice
recognition identifies the user's voice based on
intonation, accent, or speech style according to the stored
voice. So that if other users given orders to the smart
home, the system will not function like research
conducted in reference [3], [6], [7], [8], [9], [10], [11],
[12], [13], [14], [15]. Meanwhile, researchers in
reference [4] and [5] chose to used speech recognition
because researchers focused more on developing
research on the success of the system in capturing words
uttered by users due to each accent. different regions,
testing the sensitivity of the system better in which
language, or testing whether the system could recognize
words that do not match the structure of the language.
3.3 Hardware Used
The difference in the research conducted also were at
the choice of hardware used by researchers in voice
control or microcontrollers. The voice control module
that can be used was the EasyVR module or the voice
recognition module. This module was mostly chosen by
researchers. But there were researchers in reference [9],
[11] who used applications developed on Android
connected by bluetooth. As for the microcontroller, most
researchers used Arduino Uno and others used Arduino
Nano, Arduino raspberry pi3 (Arbaus and Aryani) and
wemos (Prakoso). Arduino Uno has been chosen because
it maked it very easy in rapid prototyping, trial and error,
the upload process is faster, and safer because it had an
internal regulator and fuse. Meanwhile, raspberry pi3 is
used because it can be used in projects involving high-
level software using the widely known C language, and
wemos is used because it had ESP-8266-based Wireless
Fidelity (Wifi) which makes it easier to develop devices
that used Wifi.
Remote voice control is done by combining software
that can be developed on Android and connected with a
Bluetooth module or a NodeMCU module. Control that
adds a bluetooth module can be carried out in an
intermediate distance with a maximum distance of 10
meters. The drawback was that if there is a barrier, it can
make the system unable to catch the command given.
Meanwhile, control that adds a NodeMCU module can be
done from anywhere with an internet connection. The
drawback was that if the internet speed was low, the
system couldnot catch the command given.
3.4 Testing
Testing on voice-based control systems consists of
voice command testing, program testing, circuit testing,
distance testing, environmental testing, response time
testing and delay testing. Voice command testing was
Advances in Engineering Research, volume 196
31
used to determine whether the system could recognize
voices and perform commands according to the voice
given. This test was done by giving voice commands
from the same person, other users, different accents, and
age ranges. For program testing, it was carried out to
determine whether the program was in accordance with
the planned system and whether there was an error.
Program testing was related to the programming
language and programming software used. Circuit testing
was carried out to test whether parts of the circuit, namely
inputs such as microphones, processes such as
microcontrollers, and outputs (related to actuators) can
operate properly or not. The distance test was carried out
to test whether the microphone can pick up sound at a
specified distance. Meanwhile, for remote control,
distance testing was carried out to test whether the
transmitter and receiver can communicate properly.
Environmental testing was carried out to measure the
noise level whether it affects the system or not. Testing
of response time and delay was carried out to test the
response of the system when given a voice command,
whether the system would immediately execute the
command or there was a pause to carry out the command.
3.5 Result
The results were obtained from the testing process, in
the voice command test, the following results were
obtained when voice testing with the same person had
high results of more than 80%, while for testing other
users had low results of less than 25%, while for testing
different accents and age obtained accents of kelantan,
trengganu and kedah are the worst accents and age 60
years and over give a lot of error commands. The results
of program testing as a whole did not experience
problems, this can be attributed to all research that shows
the system could work in accordance with the
instructions given. For the results of the circuit testing as
a whole there were also no problems, because the circuit
could work according to the instructions given, for
example in research using the L298N motor driver, in this
study it was known that the motor driver could function
properly. The results of the distance test showed that the
best distance between the sound source and the
microphone was 3 cm - 10 cm, while for remote testing
using bluetooth and wi-fi, the best distance was between
1 - 20 m. Apart from distance, the remote control system
was also affected by the baud rate and internet speed. The
higher the baud rate, the lower the results, and the higher
the internet speed, the higher the results. The results of
environmental testing show that a noisy environment
would give many errors. This can be seen from an
example of a study which states that the room noise level
of 34.5 dB gives a success of 86%, while the 62 dB room
noise level gives a success of 79%, another example was
in another study which states that the best environment
for testing was in the library and the worst environment
was located at the cafe. In the test results, the response
time and delay were influenced by the distance between
the sound source and the microphone, the farther the
distance between the two would make the response of the
system longer.
3.6. System Procedure
The way the voice-based control system works when
using voice recognition was that the user records their
voice to be stored in the database as a sample to be used
in verifying the voice input later. Input in the form of
voice commands given through the microphone, was
forwarded to Easy VR as voice recognition which will
match the voice sensor output data that had been
sampled. The microcontroller was Arduino which will
process the output from EasyVR and activate the
controlled device as in “Fig. 2”. If the voice command
used was in the form of speech recognition, the system
works by utilizing speech to text on the Web Speech API,
the input was then converted into text. Users must
connect to the internet for intermediary data exchange in
order to access the Speech API Web server. The input
that was processed will be responded by the server by
sending a switch on or switch off command. The
microcontroller would then run the relay according to the
command received.
Figure 2 Block diagram of how the system works
Remote control can be done using bluetooth and
NodeMCU. Controlling using bluetooth, the system was
run by opening an application that has been made on
Android. Then given the command to activate the
controlled device. This command would then be
converted and sent via bluetooth to android using a
voice and speech detector. The next command was
processed by the Arduino Uno microcontroller and will
be executed to turn on or turn off electronic devices and
send feedback to Android as in “Fig.3”. In controlling
using NodeMCU, the system worked by giving voice
commands as input through an application that has been
designed on Android which would then be converted
and processed in Arduino and the data was sent to the
NodeMCU module.
Advances in Engineering Research, volume 196
32
Figure 3 Block diagram of how the system works
using Bluetooth
3.7 Factors
Factors that could affect the sound-based control
system include clear pronunciation, sound pitch,
microphone distance, sound source, intonation and noise
level. If there was a change in the sound, the system
would not work. This was because the voice is not
identified in the database that had been created. Whereas
in remote control, things that needed to be considered
when using bluetooth are distance, obstacles, and signal
strength. If you used an internet network, you needed to
pay attention to the internet speed level. The greater the
internet speed, the greater the success rate.
4. CONCLUSION
The results of the study conducted by reviewing
journals as described can be concluded that the design
and implementation that has been done using voice
commands in the form of voice recognition and speech
recognition can facilitate human activities in everyday
life. Control automation can be applied to a smart home
by controlling electronic devices such as lights, fans,
TVs, air conditioners and so on which are based on
voice commands. It can also be applied as a security
system such as automatic door control. In addition, the
application of voice commands can also be applied to
smart hospital rooms such as controlling the patient's
bed in the hospital and assisting in calling for nurses. It
can also help users with disabilities such as moving a
wheelchair and printing documents converted to braille.
REFERENCES
[1] M. Zed, Metode penelitian kepustakaan, Jakarta:
Yayasan Obar Indonesia, 2014.
[2] Sugiyono, Metode penelitian kuantitatif, kualitatif
dan R&D, Bandung: Alfabeta, 2017.
[3] N. Aktar, I. Jahan, and B. Lala, “Voice recognition
based intelligent wheelchair and GPS tracking
system,” in International Conference on Electrical,
Computer and Communication Engineering,
February 7-9, 2019.
[4] B. Manigiri, D. Maladewa, R. Panchal, S. Hakke,
and S. Dicholkar, “Voice controlled braille
emboser,” International Journal for Research in
Apllied Science and Engineering Technology, vol.
8, no. 1, pp. 415-417, January 2020.
[5] E. I. Agustin, T. R. Yunardi, and A. A. Firdaus,
“Voice recognition sistem for controlling electrical
appliances in smart hospital room,” Telkomnika,
vol. 17, no. 2, pp. 965-972, April 2019.
[6] A. Seppiawan, Nurussa’adah, and P. Siwindarto,
“Sistem keamanan pintu pagar otomatis
menggunakan voice recognition,” Jurnal mahasiswa
TEUB, vol. 2, no. 6, pp. 1-6, 2014.
[7] Y. Mittal, P. Toshniwal, S. Sharma, D. Singhal, R.
Gupta, and V. K. Mittal, “A voice-controlled multi-
funcional smart home automation system,” in
Annual IEEE India Conference, 17-20 December,
2015.
[8] D. Arbaus, D. A. Prasetya, and A. P. Sari,
“Kecerdasan buatan pada sistem pintu otomatis
menggunakan voice recognition berbasis raspberry
PI,” Jurnal ilmu-ilmu teknik-sistem, vol. 12, no. 3,
pp. 186- 196, 2016.
[9] C. S. Tyagi, M. Agarwal, and R. Gola, “Home
automation using voice recognition and arduino,”
Journal of Recent Trends in Engineering &
Research, vol 2, no. 7, pp. 1-6, July 2016.
[10] A. Asni, A. F. S. Rahman, and M. Mursid, “Rancang
bangun buka tutup pintu otomatis menggunakan
pengenalan isyarat tutur,” in Seminar Nasional
Inovasi Teknologi Terapan, 2017.
[11] D. Ariyani, D. Iskandar, and F. Indriyani,
“Perancangan smart door lock menggunakan voice
recognition berbasis raspberry PI 3,” Jurnal cerita,
vol. 4, no. 2, pp. 180-189, August 2018.
[12] S. Ariyanti, S. S. Adi, and S. Purbawanto, “Sistem
buka tutup pintu otomatis berbasis suara manusia,”
Electronics, Informatics, and Vocational Education,
vol. 3, no. 1, pp. 83-91, May 2018.
[13] L. D. W. Raj, K. Santhosh, S. Subash, C. Sujin, and
P. Tharun, “Voice controlled door lock system using
matlab and arduino,” in International conference on
system computation automation and networking,
29-30 March, 2019.
[14] D. Eridani, I. Sanusi, and E. D. Widianto,
“Pengenalan dan analisis ucapan pada sistem
control perangkat listrik menggunakan arduino
uno,” Jurnal Teknologi dan Sistem Komputer, vol.
6, no. 1, pp. 18-24, January 2018.
Advances in Engineering Research, volume 196
33
[15] L. A. Dzulfikar, E. Haryatmi, and T. A. Riyadi,
“Rancang bangun purwarupa sistem pengunci
lemari dengan pengenalan suara,” Jurnal Ilmiah
Teknologi dan Rekayasa, vol. 24, no. 3, pp. 216-
225, Desember
Advances in Engineering Research, volume 196
34
