NavEye: Smart Guide for Blind Students
Hamza A. AlAbri, Ahmed M. AlWesti, Mohammed A. AlMaawali, and Ali A. AlShidhani
Sultan Qaboos University, {u084055, u082176, u085152}@student.squ.edu.om, alily@squ.edu.om
1
Abstract - Blind students all over the world need guidance to
navigate around the city or university campus. It is estimated
that 374,400 visually impaired students attended colleges at
some point in the United States alone. With the rapid
technological advancement, new opportunities are emerging to
make life easier for blind students. Motivated by the problem,
the objective of this project is to design, prototype and evaluate
a smart guidance system to help blind students navigate from
and to classrooms. To design such a system, the design team
had to exercise several design steps typically involved in an
engineering design process. The steps include literature
research, requirements specification, solution brainstorming,
system functional decomposition, functional modeling,
prototype implementation and testing. The designed system
consists of two fully integrated modules; a mobile application
and a smart cane. The blind student interacts with the mobile
application using voice commands. The blind student can
instruct the mobile application to navigate him/her to a desired
destination (eg. classroom). On the other hand, the cane helps
the blind student to avoid obstacles and alerts him/her when
walking off track. The cane and the mobile application are
synchronized via Bluetooth technology. The system was
modularly designed, tested and finally integrated together.
Testing exhibited promising results and proved to be beneficial
to such an important sector of the society.
INTRODUCTION
It is well known that a blind person daily encounters many
difficulties such as walking on the road, finding right path
and taking a bus [1]. Escorting blinds is a social activity
closely tied to our Islamic teachings and traditions. Barriers
may harm the blind when they are walking on the street. The
World health organization [2] presented important
statistics related to blind people. 285 million people are
virtually impaired: 39 million blind and 246 million having
low vision. 90% of the visually impaired population lives in
developing countries. Blind people do eventually go to work
or school. Table I. illustrates some statistics extracted from
studies conducted by the national federation of the blind
with respect to blinds and education [2][3].
TABLE I
STATISTICS OF VISUALLY IMPAIRED STUDENTS IN THE US.
Less than High school Some college Bachelors
high school diploma education degree or
higher
4,232,100 1,061,600 939,700 374,400
Blind people may find it easy to move within a small area
such as a room but it is challenging to navigate outdoors.
Blinds all over the world need some help navigating from
978-1-4799-4836-9/14/$31.00 (c) 2014, IEEE
one place to another. Many traditional techniques are used to
guide a blind person. Some blind people prefer using guide
dogs and others prefer a real person to guide them instead.
People may not be available all the time to help the blind.
Hence, it is required to have a system that will help blind
people navigate their way without the need of a guiding
person or a pet. Another problem is to detect objects on the
blind persons walking path. Blind people usually hold a
white cane to make sure nothing obstructs the path. So,
detecting obstacles in the path is an important requirement.
The objective of this project is to design and prototype a
navigation system to help blind students navigate to and
from classrooms. The student should communicate with the
system through voice commands. Additionally, the system
should be safe to use.
I. Related Projects
Many alternative ways of guidance exists. The first one is
known as Seeing-Eye dog. A Seeing-Eye dog is a dog that
has been specially trained to help blind people navigate. The
dog must be well trained to guide the blind person. Some
blind people use a white cane to navigate their way around.
The cane works as an object detector. A blind person usually
swings the cane as he/she walks to help him/her find and
avoid any obstacles in the pathway.
Todays technological advances simplified peoples life.
Several researches and projects exist in the field of
navigation. Global Positioning System (GPS) added an
important value to navigation technology. Many systems
developed and tested to meet navigation requirements
specialized for blind people. Place Recognition for Indoor
Blind Navigation thesis [4] was done by S. Anjum from
Carnegie Mellon University. The thesis focused on indoor
navigation by exploring place recognition and navigation
algorithms accessible via handheld devices to guide blinds
inside unfamiliar buildings. The outcome of the work is a
camera-based indoor blind-navigation model that can
recognize previously mapped locations.
The authors in [5] published an article detailing a
system that consists of a pair of glasses equipped with
cameras and sensors. The system produces a 3D map of the
wearers surroundings and his/her position. The system is
able to generate up to 10 maps per second then send it to a
braille device to be displayed as a dynamic tactile map.
OnTheBus [6] is an Android application designed to help
blind users reach their destination by taking the bus. The
application guides them from their present location to the
closest bus stop. The app informs the blind user on how
141
much time to wait before the bus arrives. When blind users mobile application. Each alternative has its own set of
take the bus the system informs them on how many stops are advantages and drawbacks. However, the latter alternative
left and alert them when they arrive. After getting out of the was selected because the required I/O components (such as
bus, the app resumes giving directions to guide the blind Camera, microphone, etc) are integrated and accessible.
user outdoors only.
IV. Functional Decomposition of the Design
Most of the previous projects depend on smart phones
to gain access to its features such as GPS, accelerometer and Functional Decomposition describes the system functionality
Internet connection. We noticed that handheld devices (ex. or behavior. Our system consists of several inputs such as;
Smartphones) are being used in blind navigation systems. users voice, ON/OFF button, QR code and sensors. Voice
An important property of smartphones is that they are instruction and vibration will be the output. QR stands for
portable and lightweight which makes it easy for blind Quick Response Code, which is a matrix barcode that
person to use. Smartphones provides the best became popular recently and it can be read by smartphones
communication results with a greater functionality and dedicated QR reading devices. QR can encode different
performance. Capability of sharing information over a types of data compared to the classical barcode. QR consists
network such as the Internet makes life easier and capitalizes of a pattern of black dots arranged in a pattern of white
on smartphones capabilities. background. The QR Code is used to solve the problem of
determine the current location of the user and to check if
II. Engineering Requirements of the Proposed System he/she is in the correct destination. Figure 1 shows a bird
The proposed system should meet the following engineering view of the system functionality. Figure 2 shows the
requirements: interconnection between different modules and sub-modules
1. As a measure of performance, the response time in the system.
should be less than 10 seconds.
2. The system should be available upon demand.
3. Due to the importance of the users location, accuracy
should be within (+/-) 5 meters of the target.
4. The system should be able to operate for 10 hours.
5. Because blind people will use the system, the system
should be easy to learn and should not take more than
an hour to learn how to operate and use.
6. System should be designed to permit upgradability in FIGURE 1
the future. BIRD EYE VIEW OF SYSTEM FUNCTIONALITY
7. A physical button should be used to switch on the
system.
III. Engineering Constraints
Engineering Constraints are special type of requirements and
should be considered to achieve successful and realistic
design. Design constraints are:
Economic constraint: The cost of the system should not
exceed 20 OMR.
Health and Safety constraint: The system should detect
stationary objects that may harm the blind person.
Ethical constraint: Communications between the device
and the blind person must be private.
Social and Cultural constraint: The system should
support Arabic language.
FIGURE 2
FUNCTIONAL DECOMPOSITION OF THE SYSTEM
SYSTEM DESIGN
In functional decomposition level 1, the system consists of
Designing a system involves extensive research to find the
two main modules which are the smart handheld device and
best solution that matches the requirements. Many design
the cane as shown in Figure 2. Users voice, ON/OFF button
alternatives exist, but constraints force designers to a
and QR code are the inputs to the smart handheld device
specific design direction. During the course of the project,
module while the sensors are the input to the cane module.
we thought of three different design alternatives. Those
Smart handheld device will have an intermediate connection
alternatives are: web application based system, development
with the cane. The output of the smart handheld device will
board based system and a hardware system interfaced with a

142
be the voice guidance and specific vibration will be the
output of the cane module as shown in Table II and Table
III.
TABLE
SMART HANDHELD DEVICE MODULE
Module Smart Handheld Device
Inputs - User Voice
- ON/OFF Button
- QR Code.
- Power
Outputs - Voice Instruction
- Bluetooth signal
Functionality The module will accept users voice instruction,
ON/OFF button and QR code as inputs. Then it
will process and it sends voice guidance as FIGURE 3
output. In addition, it will send some information MOBILE APPLICATION MODULE FLOWCHART
to Cane module through Bluetooth.

TABLE
CANE MODULE
Module Cane FIGURE 4
Inputs - Sensor QR MODULE FLOWCHART
- Bluetooth Signal
- Switch Voice Recognition Sequence Diagram
- Power Internet-based voice recognition engine is used to convert
Outputs - Vibration voice to text. The handheld device will receive and record
Functionality The Cane will receive signals from the sensor
and smart handheld device. Depending on the the voice instruction. Then, the recorded audio will be send
coming signal, the cane will produce a specific to Google server and the reply will return as a text
vibration. The switch is used to turn ON/OFF the conversion of the voice command. Figure 5 illustrates the
Cane.
operation of the voice recognition module.
Smart Handheld Device Module
Functional decomposition level 2 for the smart handheld
device is shown in Figure 2. The Activate module is
responsible for starting the voice recognition module which
will convert user voice to text and forward it to the
application module. The QR code module is used to detect
and capture the QR image and transforms it to text to the
application module.
Flowchart of Mobile Application Module
The mobile application waits for users instruction and voice
command. When the user gives the "Navigation command,
the system will ask the user to supply its current location FIGURE 5
either by voice or scanning a QR code. We assume the VOICE RECOGNITION SEQUENCE DIAGRAM

existence of a QR code tags on classroom doors containing Cane Module

the classroom number. Then, the application checks if the
starting location is included in the map or not. In case it does
not exist, an error voice message will respond otherwise the
application will store the current location and ask the user to
give the desired destination. Following that, the system
checks the destination. If the distention is known, the
navigation process will start and if it is unknown, an error
voice message will respond forcing the system to go back to
initial state. The flowchart is shown in Figure 3.
QR Code Reader Flowchart
When the QR reader starts, the camera will open and it will
try to capture the QR Code image. Later this image will be
processed and a result text will be stored. The flowchart is
shown in Figure 4.
The Cane detects any objects around the blind person using
distance-measuring sensor. Then it alerts the user in case an
obstacle is found. If the user deviates to the wrong direction,
the smart handheld device will send a signal over Bluetooth
to the cane to alert the user using vibrations. Figure 2 shows
functional decomposition level 2 of the cane module. The
microcontroller module accepts the signals from sensors or
Bluetooth, and will process them and determine the pattern
of turning on and off the motor to generate vibrations.
Communications Technology Flowchart
A communication link is required between the mobile
application and the cane. The mobile application uses a
compass to detect if the user goes out of the prescribed path
and immediately alerts the cane. Figure 6 shows the

143
importance of the communication link to guide the user in
the correct path.
FIGURE 6
COMMUNICATION LINK FLOWCHART
V. System Circuit Design
After describing the functionalities of major system
modules, this section sheds some light on the selection
criteria of the components constituting the system.
Components Selection
System requirements and constraints such as cost, accuracy
and functionality should be considered when selecting
components. To do that, alternatives of every component are
compared against each other to select the best option that
will satisfy the requirements and constraints.
Microcontroller Selection
Many alternatives are available such as; AVR Atmega8 or
Atmega tiny from Atmega Family. Atmega tiny has few
ports not sufficient to be used in our system. In addition, it
does not support serial communication such as RS232 and
SPI. Atmega 16 and 32 supports additional functionality and
extended ports. So, Atmel Atmega8 is a suitable choice.
Sensor Selection
There are two distance sensors alternative: Ultrasonic and
InfraRed (IR) sensor. IR sensors are affected by the
surrounding lighting and its range is limited compared to
ultrasonic sensors. In addition, ultrasonic sensors are more
accurate. Therefore, ultrasonic sensor is a suitable option for
the project needs.
Motor Selection
DC motor or stepper motor is needed to generate vibrations.
Stepper motor can rotate clock-wise or counter clock-wise in
a step of 18 degrees. Stepper motor needs 12 DC V which is
much larger than the voltage needed for DC motor (5-7
Volts). Stepper Motor provides functionalities that we dont
need and consumes power greater than DC motor.
Therefore, DC motor provides sufficient functionalities for
the objectives of the project therefore it was chosen.
Communications Technology Selection
There are several options for the intermediate connection
between the cane and the smart handheld device. Examples
include, direct cable connection, Bluetooth and ZigBee. The
cable is not easy to use because of impeding and controlling
the movement of the blind person and both ends of cable
must be connected and the blind doesnt enjoy vision to do
that easily which fell short with the ease of system usage
requirement. On the other hand, ZigBee and Bluetooth give
the user the freedom to move around compared to cable
connections. However, Bluetooth is more compatible with
144
smart handheld devices such as mobile phones. Therefore,
Bluetooth is the selected communication method.
VI. Constraints Effects on System Design
Previously mentioned constrains forced the design towards
specific directions. Health and safety constraints forced the
designers to add a cane to the system to detect obstacles.
Cultural and social constrains resulted in adding extra
program functions to support Arabic language. Economic
constraints limited the selection of some components
because the objective was to design a system affordable to
students.
SYSTEM IMPLEMENTATION AND TESTING
NavEye is the name of the designed blind guiding system.
It consists of two main modules which are; the mobile
application and a cane. This section will describe NavEye
implementation and Testing.
VII. System Implementation
NavEye consists of two main modules; a mobile application
on smart handheld device and a cane. Each module is
separately implemented and tested. Bluetooth
communication technology is used to link both modules.
Figure 7 shows the user interface of NavEye. Footpath
library is used to provide indoor navigation. The library was
modified and extended to meet the project requirements.
FIGURE 7
NAVEYE USER INTERFACE
System Developing Tools
Eclipse: is a development tool based on Java. Used in
multiple programming languages and it is one of the
common tools available to develop android mobile
application.
JOSM (Java OpenStreet Map Editor): FootPath function
using an OpenStreet maps which is an open source map.
JOSM was used to draw and configure the paths for the
navigation process.
Voice Recognition Implementation
Android operating system supports voice recognition online.
Offline voice recognition could be a desirable alternative
because it doesnt require an Internet connection. By testing
the offline voice recognition we found that the accuracy is
low compared to online voice recognition. Therefore,
Google online voice recognition is used to achieve fast and
accurate voice to text recognition. An additional Class has representing the cane and simulated its operation using
been added to record the voice of the user and handle the Proteus simulation software and we observed the circuits
reply. The application depends on the resulted text to behaviour. Figure 11 shows the circuit layout and the
identify the users request. Therefore, we have tried to connection between different components in the cane.
recognize the same voice command and came up with
several possibilities and tried to match them with single
instruction to increase the reliability of the application.
QR Code Implementation
Many QR Code Reader implementations exists. Zxing
library is one of those implementations with the ability to be
embedded with other applications. QR Code reader is used
in two different scenarios: Reading QR code image of the FIGURE 8 FIGURE 9
starting point of the journey and checking the users current SNAPSHOT OF RECEIVING SMS SNAPSHOT OF CURRENT LOCATION
location by reading the nearest QR code. AT SENDER SIDE
Bluetooth Communication Implementation
Bluetooth is needed to establish a communication channel
between the smart handheld device and the cane. An extra
class has been created to handle the Bluetooth connection. It
contains three sub functions. Firstly, BTturn and it is used to
check whether the Bluetooth device of the smart handheld
device is ON or OFF. Secondly, BTConnect and it is used to
establish a connection with the cane. Finally, BTSend and it
is used to handle sending data process.
FIGURE 10
Emergency Situation
SNAPSHOTS OF COMMAND MATCHES
The worst case that may happen to a blind student is when
he/she gets lost while navigating from one point to another. Cane Implementation
The problem is to detect the current location during the The Cane is a central module in NavEye. In order to
navigation. The developed application can determine the implement the cane module, we must have general
starting location coordinates (latitude, longitude) and it can information about the components to be used and the
also keep track and update the location coordinates during connections. In this section, more details are shown on how
navigation. Therefore, we have added an extra feature as an the cane module was implemented. We designed the circuit
emergency solution. In case if the user is lost, he/she can representing the cane and simulated its operation using
supply help voice command which will initiate a phone Proteus simulation software and we observed the circuits
call and at the same time send SMS that contains the current behaviour. Figure 11 shows the circuit layout and the
location to a predefined phone number. Figure 8 shows a connection between different components in the cane.
snapshot of the received SMS that contains a link to Google The system contains sub modules that will be discussed
map to show the exact location of the user. Figure 9 shows below:
the map and pointer that point to the users location.
Atmel Atmega8 Microcontroller : Processing unit
Multi Language Support
Bluetooth Module : Receiver module
One of the cultural constraints is to support Arabic language.
3 Voltage Regulator: Voltage Divider.
In addition, the voice recognition results are not consistent
due to the errors occurring in the voice recognition process. DC motor: Vibration generator.
Therefore, there is a need for an extra feature to solve this The ultrasonic sensor HC-SR04 is connected to the
issue. Our solution is to add a dynamic dictionary that maps microcontroller through serial port. The microcontroller
the original command with other possibilities from voice triggers the sensor to send a pulse and run a timer at the
recognition and also match the original command to other same time. When an echo is received signal reflected by an
language words. Figure 10 shows a screenshot of commands object the timer will stop and time interval is obtained.
translated from English to Arabic. The user can issue a voice The Distance is calculated by knowing the time interval
command from any language he/she prefers. and using special formula (1) to calculate the distance as
Cane Implementation shown below:
The Cane is a central module in NavEye. In order to
implement the cane module, we must have general (1)
information about the components to be used and the
connections. In this section, more details are shown on how
the cane module was implemented. We designed the circuit

145

FIGURE 11
CANE MODULE
FIGURE 12
ACTUAL IMPLEMENTATION OF THE CANE
The microcontroller is programmed to decide whether
there is an object in front of the user or not and if an object
exists in a range below 1 meter the microcontroller will send
continues pulses to trigger the motor to rotate. When the user
deviates to the left or to the right, the smart handheld device
sends a character through Bluetooth Either R or L to the
microcontroller. The microcontroller decides whether user
should deviate to left or right. Figure 12 shows the actual
implementation of the cane.
VIII. System Testing
Several tests have been conducted to check the
functionality of the proposed system. Testing was conducted
on different levels starting with unit testing followed by
integration testing and finally acceptance testing. Unit tests
checks the functionality of each unit individually. Examples
of such tests include testing the voice to text module, QR to
text module, Bluetooth module, testing sensors and motors.
Integration testing was conducted to verify the integration of
different modules such as the communication between the
cane and the smartphone. Acceptance testing is used to
verify that the system meets design requirements listed in
section I.B. All tests were successfully carried out and
Figure 13 shows a blind student trying to use NavEye.
CONCLUSIONS AND F UTURE WORK
Blind people are facing difficulties every day; they need
assistance to help them move from one place to another.
146
There are many traditional methods to help a blind person to
navigate. Today, technological methods are being used to
guide blind people. The objective of this project is to design
and prototype a navigation system to help blind student
navigate from and to classrooms. Different design
alternatives exist; however, an application on a handheld
device was the design choice that met our pre-defined
requirements and existing engineering constraints. The
system communicates with the user through voice command
and it can read QR Code to identify the current location. A
cane was developed to help the blind student avoid obstacles
and stay in the prescribed path. NavEye can be extended
in the future to permit synchronizing the application with the
student study timetable. At the beginning of the semester,
NavEye will identify the classes that have to be visited daily
by the blind student and automatically generate the
navigation path. We anticipate that NavEye will have a
positive impact on the society and will help blind students
navigate safely.
FIGURE 13
A BLIND STUDENT TRYING NAVEYE
REFERENCES
[1] Dong, B. The Research on Blind Navigation System Based on
RFID.http://ieeexplore.ieee.org/xpl/login.jsp?reload=true&tp=&arn
umber =4340289. Accessed: October 10, 2012.
[2] Statistical Facts about Blindness in the United States.
https://nfb.org/factsaboutblindnessintheus. Accessed: October 10,
2012.
[3] World Health organization, Visual impairment and blindness.
http://www.who.int/mediacentre/factsheets/fs282/en/. Accessed:
October 10, 2012.
[4] Anjum, S. Place Recognition for Indoor Blind Navigation.
http://www.cs.cmu.edu/afs/cs/user/mjs/ftp/thesis-
program/2011/theses/qatar-anjum.pdf . Accessed: October 25, 2012.
[5] Knight, H. Robot sensing and smartphones help blind navigate.
http://www.newscientist.com/article/mg21428625.700-robot-
sensing-and-smartphones-help-blind-navigate.html. Accessed:
October 2, 2012.
[6] Coxworth, B. GPS-enabled app helps the blind take the bus.
http://www.gizmag.com/onthebus-blind-bus-navigation-app/23054/.
Accessed: October 10, 2012..
AUTHOR INFORMATION
Ali A. Al Shidhani is an Assistant Professor, Department of
Electrical and Computer Engineering, Sultan Qaboos
University, Sultanate of Oman. The rest of the authors
graduated from Sultan Qaboos University with Bachelors of
Electrical and Computer Engineering in 2013.