Abstract— An evolution of the existing home automation is home robotics. This
paper presents an approach to home robot using odroid XU4.This is specially
designed for paralyzed patients who are not able to move properly, which will
helpful to family of that patient, it only work through authorized persons
only. This is a small portable monitoring system, designed for home &
office. This includes the various processes such as recognition of voice, face
detection, localization of object, pick and place, home automation. All this is
controlled by Odroid XU4.To avoid frontal obstacle pair of webcams are used.
From this robot we can control lighting system of home or office through voice
as well as through mobile app which is operated through Bluetooth. The final
system can identify & follow target object (over 10 feet).
The odroid board XU4 is 2 GHz, Octa core processor
& multiple i/o interfaces. We use OpenCV libraries to process images &
POCKETSPHINX for speech recognition.
Robot, Odroid XU4, OpenCV, Stereo Vision, Pocketsphinax, Degrees of freedom,
Arduino mega, Bluetooth.
Paper Goal :
Nowadays it is very difficult to
get the manpower for household works. Works like shifting or moving of small objects
from one place to other. Look after a patient in the home when all people in
the home are working people. The robot in the paper is capable to overcome all
the problems. The robot is totally mobile machine, which could pick up the
object from one place and will place at particular place. The robot will work
only on instructions of authorized person.
Robotics plays very vital role in
different fields of our society such as security, industries, business etc.
From small houses to huge scale industries, now home robot is essential and
plays very vital role to fulfill our day to day aspects in many ways. Nowadays
keeping care taker for elder people or servant, domestic robots are useful.
Domestic robots are of two types – 1) Outdoor Domestic Robots, 2) Indoor Domestic
Robots. The home robot comes under Indoor domestic robots. Without any
complaints & getting tired this mobile machine, could pick up the object
from one place and deliver at particular place. The number of human interactions
for the control of basic human function can be replaced by using home robot
which can be considered as an act of using electronic systemdevices and its
programming. Actions of the system are controlled by using processing component
of the system which is heart of the system. In this proposed system, the Odroid
XU4 is a really powerful microcontroller and it can be used as a normal use
computer. Hence there is no need of PC and this system will work in standalone
mode. To avoid obstacles pair of webcams is used and are able to implement a
basic frontal obstacle avoidance system. The two images captured between two
cameras called disparity. Disparity means difference. Disparity map is
measurement of difference between two images.
Fig1: Major components of
Computer vision is one of the most demanding
areas in a robotic field. To shift materials, parts, tools or specialized
device through variable programmed motion for performance of variety of tasks
robot act as a programmed motion for performance of variety of tasks. This
proposed system is entirely controlled by voice and for this POCKETSPHINX
software is used. POCKETSPHINX is a
part of the CMU SPHINX Open Source Toolkit For Speech
III. LITERATURE SURVEY:
In the 18th century
scientists made one doll like machine which was stationary and used to help to
pick up the object from one and deliver. As it was stationary it could not
deliver the objects for far distances so there was need for mobile machine,
which could move and deliver at particular place. The proposed system robot
will overcomes all the problems.
Voice-Controlled Personal Assistant Robot: In 2015
International Conference on Industrial instrumentation and control (ICIC),
college of engineering Pune, India.
Voice controlled assistant robot was small in
size which does not pick up the object which are on the table. This drawback is
removed in proposed system.
ASIAN JOURNAL OF SCIENTIFIC RESEARCH
2015, in this paper the Hand Gesture based control of robotic hand, using Raspberry
Manipulation Through An Assistive Home Robot: 2012 IEEE
International conference on Intelligent Robots and Systems.
This Mobile manipulation assistive home robot is operated through remote and this
drawback is removed in proposed system, the proposed system works on the voice.
a Personal Reaction Robot: Introduced by Naongdech
Keeratippranon and Nuengong Tuaycharoen, Dhurkji Pundit University, Bangkok, Thailand.
journal of science, engineering and technology research (IJSETR) April
2015, the paper published the design of
Raspberry Pi vehicle with stereo vision
IV. Proposed system:
The proposed system in home robot using odroid XU4 and
OpenCV. Main objective of the system is to design a portable monitoring device
for home and office. When specific
instructions are given, the robot initializes the camera and checks for the
authentication of the person. Once the person is authorized, instructions will
be accepted by the processer. Where processers will give signal to the stereo
camera to detect the object in surrounding. After localization of the object
the processer will give the signal to pick up the object from its source
location. Once the object is picked up the instructions will be given to
deliver it at its destination location. For this the program is written in
python environment in software implementation.
From this robot we also control our home or office lights
through voice as well as through mobile for this we use Arduino mega and
2: Mechanical assembly structure
Frame structure of the Robot is shown in fig 2.
It is fully made up of Aluminum.
Aluminum is light in weight and having high tensile strength which is
sufficient for Robot. It has size of 35cm*35cm*80cm with weight of 22kg with
max speed 3 ft.sec.
angle = 10 degrees
= d= 4.37 in; r = 2.18in
Acceleration a = 1.72 ms^2
Fig 3: Inclined angle of bot
Torque = 22kg (1.72ms^2) + (9.81ms^2*sin (10)*0.0553
T = 4.162 N-m
Convert it into kg-cm, we get 16.75 kg-cm
We use two drive motors which having the torque 35 kg per
motor and having speed of 100 RPM, and two Omni wheels are used for proper
balancing the robot. Size of the wheels is 4 inches.
Robotic Arm: To pick and
place the object we design a one robotic arm with two fingers as
Fig. 4: Robotic Arm
gripper. This arm is made up of aluminum and cardboard. As
shown in a Fig 4 the link 1 is of aluminum and rest of all made up of
cardboard. All these links are connected through actuators i.e. servo motors.
This is operated through PWM signal given by ODROID XU4 when object is detected
then the arm starts his working to pick the object in his fingers and place
these object in authorized persons hand.
Here, link 1 is of 21.1 cm and 119 gm., link 2 is 22.1 cm
which is made of the glass fiber and of 70 gm. and link 3 of 24 cm with weight
80 gm. These arms carry the object of weight 120 gm. efficiently. The robotic
arm having the five degrees of freedom to pick and place operation.
Block Diagram of the robotic arm
There are two steps which can be used to implement the
proposed work –
1. Hardware Implementation