AI CONTROLLED HOME APPLIANCE CONTROL

SYSTEM USING IoT

Team 2
Luis Gomes
UG student in Electronics and Telecommunication Engineering
Don Bosco College of Engineering, Goa, India
gomes.luis479@gmail.com

Abstract – Home Automation using internet of things is an water and reduces much of the wastage [1].
emerging technology and todays need. This technology offers It is impossible for someone to predict that IoT would
new and exciting opportunities to increase the connectivity of
appliances within the home for automation. Generally, Wi- come into existence in a very short period. IoT is one of the
Fi, Bluetooth, GSM modules or an Android app are used by promising technology used to control objects connected to
the existing methods for controlling various home appliances. internet through IP address. In Real Time Application such
But the expansion of internet in the last decade, gives one the
as smart home, smart cities, smart grids and intelligent
potential for controlling and automating various appliances
through it in a comfortable and secure way. On the other transportation, it is widely used. It has also expanded its
hand, Artificial Intelligence is evolving as a technology for features into medical fields such as instruments and many
developing automatic systems that can perceive the more. Smart home is one in which various electronic and
environment, learn from environment, and can make
decision using case based reasoning. The framework for electric appliances are wired up to a central computer
taking real-time decision and automation for IoT is provided control system so that they can either be switched on and
by artificial intelligence. AI has vision ability, knowledge off at certain times [2].
base, learning and decision-making ability, which makes it a Cloud computing is a rapidly growing technology where
better solution for automatic systems. This paper mainly
aims to provide the user the facility to control their home resources such as storage devices, platform and
appliances using internet and artificial intelligence, which applications are shared over the internet and is widely used
will help them saving the electric power and human energy. by multiple users in small and medium business. Cloud
The various appliances are connected to the sensor which is
connected using wireless network.
services can be provided and delivered remotely by
vendors such as Amazon or Microsoft as “public clouds”,
Keywords – Home Automation, Internet of Things, Artificial or the resources are designed, installed, monitored and
Intelligence controlled internally as “private clouds”. Cloud data
retrieval is an important service to be considered as certain
specific data files the users are interested during a given
I. INTRODUCTION
session must be retrieved in an efficient way and quickly
[3].
An automated and networked home is “An automated
In this project we are connecting the various sensors and
and networked home is one in which every appliance can
devices to Raspberry Pi which is connected with IFTTT
be remotely managed [4] from anywhere on the Internet
and the devices are operated as per the voice commands
with a simple Web browser” [5]. The general goal of the
given through Google Assistant.
automatic-home movement is to use networking
technology to integrate the devices, appliances and
services found in homes so that the entire domestic living
space can be controlled centrally or remotely.
Automation is a technique, method, or system of
operating or controlling a process by electronic devices
with reducing human involvement to a minimum. The
fundamental of building an automation system for an
office or home is increasing day-by-day with numerous
benefits. Industrialist and researchers are working to build
efficient and affordability automatic systems to monitor
Figure 1. Popularity of Smart home in market [1]
and control different machines like lights, fans, AC based
on the requirement. Automation makes not only an
efficient but also an economical use of the electricity and
 II. LITERATURE REVIEW transferred to the VM in the cloud. The user can use the PC
or Android phone to monitor or control through the
The Home automation system that uses Wi-Fi Internet to power-saving of the house.
technology. System consists of three main components; Arduino microcontroller to receive user commands to
web server (Particle photon and IFTTT), which presents execute through an Ethernet shield. Our house network
system core that controls, and monitors users’ home and used together both wireless ZigBee and wired X10
hardware interface module (Raspberry Pi or ESP8266), technologies. This system followed smart task scheduling
which provides appropriate interface to sensors and with a heuristic for the Resource-constrained-scheduling
actuator of home automation system. The System is better problem (RCPSP). The mobile device can be either wired
from the scalability and flexibility point of view than the to the central controller through USB cable or
commercially available home automation systems. The communicates with it wirelessly, within the scope of the
user may use the same technology to login to the server home. Arduino contains the web server application that
web based application. If server is connected to the communicates through the HTTP protocol with Web-based
internet, so remote users can access server web based Android application. The system is highly flexible and
application through the internet using compatible web scalable and expandable.
browser. The Fig.1, shows projected trends in the smart The home network which monitors the appliances and
home market in the coming year. sensors and transmits data to the cloud-based data server
The Google Assistant, by Google, has been developed which manages the information and provides services for
based on the android system. An interface card has been users by transmitting data and receiving user commands
developed to assure communication between the remote from mobile application. The proposed system has good
user, server, raspberry pi card and the home Appliances. modularity and configurability characteristics with very
The application has been installed on an android low power consumption in cost efficient way.
Smartphone, a web server, and a raspberry pi card to Application developed using the Android platform
control the shutter of windows. Android application on a controlled and monitored from a remote location using the
smartphone issue command to raspberry pi card. An smart home app and an Arduino Ethernet based micro web-
interface card has been realized to update signals between server. The sensors and actuators/relays are directly
the actuator sensors and the raspberry pi card. interfaced to the main controller. Proposed design offers
Cloud-based home appliance monitoring and controlling are the control of energy management systems such as
System. Design and implement a home gateway to collect lightings, heating, air conditioning, security, fire detection
metadata from home appliances and send to the cloud- and intrusion detection with siren and email notifications.
based data server to store on Particle Cloud. Raspberry Pi
proves to be a powerful, economic and efficient platform III. SYSTEM DEVELOPMENT
for implementing the smart home automation. Raspberry
pi based home automation is better than other home As shown in Fig.2 home appliance control system is
automation methods is several ways. For example, in implemented using IoT and AI.
home automation through DTMF (dual tone multi-
frequency), the call tariff is a huge disadvantage, which is
not the case in their proposed method. Also, in Web server
based home automation, the design of web server and the
memory space required is ejected by this method, because
it simply uses the already existing web server service
provided by G-mail. LEDs were used to indicate the
switching action. System is interactive, efficient and
flexible.
Shih-Pang Tseng et al. [7] proposed Smart House
Monitor & Manager (SHMM), based on the ZigBee, all
sensors and actuators are connected by a ZigBee wireless
network. They designed a simple smart socket, which can
remote control via ZigBee. PC host is used as a data
collector and the motion sensing, all sensing data are Figure 2. Block diagram of Proposed System
 Here we have used IFTTT with Google Assistant to give
the commands to a control system via Particle cloud. The
control system consists of Raspberry Pi with which all the
appliances are connected. In the Raspberry Pi we have
installed Particle core on which the firmware code is
flashed. This core allows for flashing of firmware code to
the Raspberry Pi over an online IDE linked to the
Raspberry Pi. Particle variables are used to store the
device statuses.
IFTTT links Google Assistant to the Particle cloud. It is
linked such that when the required voice message is
received on Google Assistant, a function is called on the
Figure 4. 6V Relay Module
device firmware with the appropriate arguments
(commands). This function processes the received C. Particle Cloud
command and controls the appliance. The appliances are
Particle powers IoT products from large enterprises,
relay driven. innovative start-ups, and everyone in between. Join a
community of 120,000 developers and engineers building
A. Raspberry Pi with Particle.

Several generations of Raspberry Pi’s have been

released. All models feature a Broadcom system on a
chip (SoC) with an integrated ARM compatible central
processing unit (CPU) and on-chip graphics processing
unit (GPU), Fig. 3.

Figure 5. Particle.io

D. UbiDots
It is a powerful combination of backend, frontend and
firmware tools to accelerate the IoT initiatives. It has
Figure 3. Raspberry Pi following features:
➢ Live Dashboards.
Processor speed ranges from 700 MHz to 1.2 GHz for ➢ Device Libraries available for more than 50
the Pi 3; on-board memory ranges from 256 MB to 1 GB devices.
RAM. Secure Digital (SD) cards are used to store the ➢ Time-series storage and Playback.
operating system and program memory in either SDHC or ➢ Trigger Events
MicroSDHC sizes. The boards have one to four USB ➢ Math and Statistical
ports. For video output, HDMI and composite video are Engine
supported, with a standard 3.5 mm phono jack for audio
output. Lower-level output is provided by a number of
GPIO pins which support common protocols like I²C. The
B-models have an 8P8C Ethernet port and the Pi 3 and Pi
Zero W have on-board Wi-Fi 802.11n and Bluetooth.
Prices range US$5 to $35 [8].
B. Relay Module
The relay module is shown in Fig. 4. The 6V relay is Figure. 5. Ubidots Logo

used to drive the various appliances.

 IV. PROPOSED SYSTEM

A. Cloud Integration

The Particle cloud service along with Particle core on

Raspberry Pi offer the required functioning. The firmware
code for appliance control is fed to the Raspberry Pi
through the Particle core. The firmware code contains
functions specific to the Particle cloud. These functions
control the appliances based on the arguments passed to it.
Since this is a cloud function, it can be called by external
sources with the appropriate authentication and arguments. Figure 7. Experimental results
The system developed uses the IFTTT to link Google
Assistant on the user’s device. Thus, IFTTT allows the VI. FUTURE SCOPE
commands received from Google Assistant to be passed to The proposed system can be further integrated with
the required function on the Particle Cloud linked different sensors and voice feedback from different
firmware code enabling appliance applications. The application can be developed that can
control. take voice commands, process them, show real time status
and give the voice feedback to the user.
The system can also be integrated to monitor various
parameters in the house and also control them so that the
complete house can be automated.

VII. CONCLUSION
In this highly developing era, where directly or
indirectly, everything is dependent on computation and
information technology, Raspberry Pi proves to be a smart,
economic and efficient platform for implementing the
Figure 6. Proposed System for Home appliance control based on AI home automation. This paper provides a basic application
using IoT. of home automation using Raspberry Pi which can be
easily implemented and used efficiently. The code
B. Appliance Status Visualization provided is generic and flexible in a user friendly manner
and can be extended for any future applications like power
Ubidots Cloud service is used for status visualization. control, surveillance, etc, easily. Moreover, this technique
The firmware code uses Webhooks on Particle Cloud to is better than other home automation methods is several
make REST API calls to Ubidots. Variables are created on ways.
the Ubidots Cloud to log the appliance statuses. REST
calls are made with the appropriate authentication tokens, REFERENCES
device labels and variable labels. A REST call is made to
[1] Vaishnavi S. Gunge and Pratibha S. Yalagi,“Smart Home Automation:
the required variable every time a device status is
A Literature Review”, International Journal of Computer Applications
changed. The call also contains the current status of the (0975 – 8887) National Seminar on Recent Trends in Data Mining
device in a JSON object which is used by Ubidots to (RTDM 2016) (accessed on: 23rd December 2017)
update the variable (device status) value. [2] Chandra SekarB, Nikhil K S, Raju K N, Sanjay M, Chandrappa D N,
“Internet of Things based Automation using Artificial Intelligence”,
International Journal of Emerging Research in Management &Technology
V. Experimental Results ISSN: 2278-9359 (Volume-6, Issue-7) (accessed on: 23rd December 2017)
[3] Sunil S MTech Scholar, A Ananda Shankar,”Fast and Efficient Cloud
Fig. 7, shows the experimental result of LED ON/OFF Data Utilization with Deduplication”, International Journal of Emerging
Research in Management &Technology ISSN: 2278-9359 (Volume-6,
and the output of motion sensor.
Issue-6) (accessed on: 24th December 2017)
[4] I. Han, H.-S. Park, Y.-K. Jeong, and K.-R. Park, “An integrated home
server for communication, broadcast reception, and home automation,”
IEEE Transactions on Consumer Electronics, vol. 52, no. 1, February
2006.
[5] P. M. Corcoran and J. Desbonnet, “Browser-style interfaces to a
home automation network,” IEEE Transactions on Consumer
Electronics, vol. 43, no. 4, November 1997.
[6] Sandeep Kumar and Mohammed Abdul Qadeer, “Application of AI
in Home Automation”, IACSIT International Journal of Engineering and
Technology, Vol. 4, No. 6, December 2012 (accessed on: 25 th December
2017)
[7] Shih-Pang Tseng, Bo-Rong Li, Jun-Long Pan, and Chia- International
Journal of Computer Applications (0975 – 8887) National Seminar on
Recent Trends in Data Mining (RTDM 2016) 10 Ju Lin,”An Application
of Internet of Things with Motion Sensing on Smart House“, 978-1-
4799-6284- 6/14 c ⃝2014 IEEE (accessed on: 25th December 2017)
[8]” Raspberry pi”, https://en.wikipedia.org/wiki/Raspberry_Pi ,
(accessed on 28th December 2017).