UNMANNED SURFACE VEHICLE (USV) WITH OBSTACLE AVOIDANCE

SYSTEM

MAHMOOD MUTHANNA A.JABBAR AL-TEKREETI

September, 2019
UNMANNED SURFACE VEHICLE (USV) WITH OBSTACLE AVOIDANCE

SYSTEM

A THESIS SUBMITTED TO

THE GRADUATE SCHOOL OF NATURAL AND APPLIED

SCIENCES OF

ÇANKAYA UNIVERSITY

BY

MAHMOOD MUTHANNA A.JABBAR AL-TEKREETI

IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE

DEGREE OF

MASTER OF SCIENCE

IN

ELECTRONICS AND COMMUNICATION ENGINEERING

September, 2019
Title of the Thesis: UNMANNED SURFACE VEHICLE (USV) WITH

OBSTACLE AVOIDANCE SYSTEM

Submitted by: Mahmood Muthanna A.Jabbar AL-TEKREETI

Approval of the Graduate School of Natural and Applied Sciences, (ankaya

University.

Prof. Dr. Can COGUN

Director

certify that this thesis satisfies all the requirements as a thesis for the degree of
Master of Science.

Assoc. P�RA�
Head of Department

This is to certify that we have read this thesis and that in our opinion it is fully
adequate, in scope and quality, as a thesis for the degree of Master of Science.

zyer
Supervisor

Examination Date: 13 / 09/2019

Examining Committee Members:

Prof Dr. B~lent TA VL\ (TOBB ETU)

Dr. Instructor. Selma ~ZAYDIN (Cankaya Univ.)

Dr. Instructor. Sibel Tanryan Ozyer ((ankaya Univ.)

 STATEMENTOF NON-PLAGIARISM PAGE

I hereby declare that all information in this document has been obtained and
presented in accordance with academic rules and ethical conduct. I also declare that,
as required by these rules and conduct, I have fully cited and referenced all material
and results that are not original to this work.

Name, Last Name : Mahmood Muthanna AL-TEKREET

Signature

Date : 13/09/2019

111
 ABSTRACT

UNMANNED SURFACE VEHICLE (USV) WITH OBSTACLE AVOIDANCE

SYSTEM

AL-TEKREETI, Mahmood Muthanna

M.Sc., Electronic and Communication Engineering Department
Supervisor: Dr. Sibel Özyer

September 2019, 62 pages

Unmanned Surface Vehicles (USV) is one of the most widely used in

open water technique. It has attracted wide attention in recent decades as it tries to
integrate monitoring with communication.

In this thesis, investigation and implementation of a prototype of unmanned surface

vehicles (USV) has been done, that has been widely used in developed countries in
various fields. The USV has been protected from barriers that may obstruct the USV
or prevent it to reach the target. The protection system has been an assistant system
for the operation of the vehicle controller in case he fails to see the obstacle, whether
due to bad weather conditions or omission by the controller. The proposed system
has been worked by means of a set of sensors with multi techniques which has been
placed on the vehicle including Light Detection and Ranging (LIDAR) sensor,
LIDAR-Lite sensors and ultrasonic sensors that will work to send data to the crash
avoidance system (CAS) and the main computer for making a right decision to take
the vehicle out of danger area. This has been done by determining whether USV in
safe or not.

iv
The performance has been increased by increasing system accuracy to be able to
work faster and coincides with the speed of the vehicle. On the other hand, the
system has the ability to direct the vehicle to conduct manoeuvres with high
professionalism. Accuracy and speed in transmitting system data to the
microcontroller and driver have also been considered when building the system. In
this work, algorithms to help CAS to make the most appropriate decision and Python
language have been used to program the system in general.

Keywords: Unmanned Surface Vehicles (USV), Crash Avoidance System (CAS).

v
 ÖZ

ENGELDEN KAÇINMA SİSTEMİ İLE İNSANSIZ SUÜSTÜ ARACI (İSA)

AL-TEKREETI, Mahmood Muthanna

Yüksek Lisans, Elektronik ve Haberleşme Mühendisliği Bölümü

Danışman: Dr. Sibel Özyer

Eylül 2019, 62 sayfa

İnsansız Suüstü Araçları (İSA) açık su tekniğinde en yaygın kullanılanlardan biridir.

İzlemeyi iletişimle bütünleştirmeye çalıştığı son on yıllarda büyük ilgi görmüştür.
Bu tez çalışmasında, gelişmiş ülkelerde çeşitli alanlarda yaygın olarak kullanılan
insansız suüstü araçlarının (İSA) prototipinin araştırılması ve uygulanması
yapılmıştır. İSA, İSA' yı engelleyebilecek veya hedefe ulaşmasını engelleyebilecek
engellerden korunmuştur. Koruma sistemi, kötü hava koşulları ya da kontrolör
tarafından atlanmaması nedeniyle engelleri görmemesi durumunda araç kontrol
cihazının çalışması için yardımcı bir sistemdir. Önerilen sistemde, araca yerleştirilen
Işık Algılama ve Değişme (LIDAR) sensörü, LIDAR-Lite sensörleri ve çarpışmadan
kaçınmaya veri göndermek için çalışacak ultrasonik sensörler dahil olmak üzere
çoklu tekniklere sahip bir dizi sensör aracılığıyla çalışılmıştır. Sistem (EKS) ve ana
bilgisayar, aracı tehlike alanından çıkarmak için doğru kararı vermeye çalışmaktadır.
Bu İSA' nın güvenli olup olmadığını belirleyerek yapılmıştır.

vi
Daha hızlı çalışabilmek ve aracın hızıyla çakışabilmek için sistemin doğruluğunu
artırarak performans arttırılmıştır. Öte yandan, sistem aracı yüksek profesyonellikte
manevralar yapmaya yönlendirir. Sistem verilerinin mikrodenetleyiciye ve sürücüye
aktarılmasındaki doğruluk ve hız da sistem oluşturulurken dikkate alınmıştır. Bu
çalışmada, EKS'nin en uygun kararı vermesine yardımcı olacak algoritmalar ve
sistemi genel olarak programlamak için Python dili kullanılmıştır.
Anahtar Kelimeler: İnsansız Suüstü Araçları (İSA), Engelden Kaçınma Sistemi
(EKS).

vii
 ACKNOWLEDGMENTS

I would like to thank my Supervisor Dr. Instructor. Sibel Özyer without her helpful
advice, valuable comments and guidance this thesis could not be completed. Her
door was always open for me whenever I need her help. It is a pleasure to express my
special thanks to my family for supporting all the time.

This work is supported by the Turkish grants Kosgeb via project name “Fully
Automated Remote-Control Life-Saving Robot”, project code 2F8UV, as Research
and Development Project and also Çankaya University Technology Transfer Office
(TTO) project no 1/KSGB/ARGEI/2018/04/132-1.

Thanks also to Rakun company to make this project possible and for their unlimited
supporting from talented team members.

Lastly, I would like to thank all the staff in my department at Cankaya University
Who support me throughout academic courses.

viii
 TABLE OF CONTENTS

STATEMENT OF NON-PLAGIARISM PAGE …………………………...…...… iii

ABSTRACT …………………………………………………………..…..……….. iv

ÖZ …………………………………………………………………..…………..….. vi

ACKNOWLEDGMENTS ……………………………………………………...… viii

TABLE OF CONTENTS ……………………………………………………...…… ix

LIST OF FIGURES …………………………………………………………..…… xii

LIST OF TABLES …………………………………………………………..……. xiv

LIST OF ABBREVIATIONS ……………………………………………...……… xv

1. INTRODUCTION

1.1. Introduction …………………………………………………………………... 1

1.2. Motivation …………………………………………………………………..... 2

1.3. Problem Statement ............................................................................................ 3

1.4. Aims and Objectives ........................................................................................ 4

1.5. Thesis Structure ................................................................................................ 4

2. LITERATURE REVIEW AND REQUIREMENTS

2.1. Overview ………………………………………………………………….….. 6

2.2. History of USV ………………………………………………………………. 7

2.2.1 USV prototypes …………………………………………………..………. 8

2.2.2 Employment of USV …………………………………………….……… 11

2.3. Crash Avoidance System (CAS) ……………………………………………. 12

ix
2.4. The sensors ………………………………………………………………….. 12

2.4.1. Ultrasonic sensors …………………………………………………….... 13

2.4.1.1. Fundamentals of ultrasonic waves ……………...……...……….... 14

2.4.1.2. Ultrasonic measuring ………………………………..………….... 14

2.4.1.3. JSN SR04T …………………………………………….……….... 16

2.4.2. Light Detection and Ranging (LIDAR) sensors ……………………….. 17

2.4.2.1. Lidar-lite sensors ………………………….………………………. 18

2.4.2.2. The RPLiDAR ……………………………………………………. 18

2.5. The development boards ……………………………………………………. 19

2.5.1. Arduino …………………………………………...……………….…… 20

2.5.1.1. Arduino Uno ……………………………………………..……….. 21

2.5.1.2. Arduino Mega ………………………………………………….…. 22

2.5.2. Lattepanda Alpha (mainboard) ………………………………………… 24

2.6. Electrical Motors ……………………………………………………...…….. 25

2.6.1. Servo motor …………………………………………………………….. 26

3 ALGORITHMS & SYSTEM DEVELOPMENT

3.1. Introduction …………………………………………………………….…… 28

3.2. A* Algorithm ………………………………………………………….……. 28

3.3. Weighted Sum Concept ……………………………...………………..……. 30

3.3.1. Weighted Sum Model ………………………………………..………… 31

3.3.2. The WSM Method ……………………………………….…….………. 31

3.3.3. The Weighted Sum Implementation …………………………..……….. 32

3.4. Kalman Filter ……………………………………………………………….. 34

x
 3.4.1. Design the Kalman Filter …………………………………….………… 36

3.5. Simultaneous localization and mapping (SLAM) Algorithm …...………….. 37

3.6. Proposed Work ………………………………………..…………………….. 39

3.6.1. Python language ……………………………………………………….. 40

3.6.2. Hardware description ………………………………………………….. 41

4. IMPLEMENTED CIRCUIT AND SIMULATION

4.1. Introduction …………………………………………………………………. 46

4.2. Implemented circuit ……………………………………………………….... 46

4.3. Simulation of Obstacle Avoidance System on USV ……………………..…. 48

5. CONCLUSION AND FUTURE WORK

5.1. Conclusion ………………………………………………………………….. 57

5.2. Future work …………………………………………………...…………….. 58

REFERENCES ……………………………………………………….…………… 59

xi
 LIST OF FIGURES

Figar2.1: An overview of USV in the timeline [14] …………...……..……..……. 10

Figure 2.2: Acoustic sound ranges and frequency values [18] ………….....….….. 14

Figure 2.3: The things that affect of ultrasonic sensor measuring [19] ……….….. 16

Figure 2.4: The shape of JSN SR04T ………………..………………...…...…….. 16

Figure 2.5: The mechanism of RPLiDAR [22] ………………...……………….….19

Figure 2.6: The specifications of Arduino UNO [23] ……………….……………. 21

Figure 2.7: The description of Arduino Mega [25] …………………......…..……. 23

Figure 2.8: Types of electrical motors ………………...…...…………….……….. 25

Figure 2.9: The types of servo motors ……………………......…..….…………… 26

Figure 3.1: Starting and Destination Points in A* Search Algorithm [32] …….… 29

Figure 3.2: Starting the A* Search Algorithm [32] …………………………....…. 29

Figure 3.3: Flowchart of the conventional A* algorithm ………...………………. 30

Figure 3.4: Kalman filter algorithm [33] ……………………………...….………. 35

Figure 3.5: System control units of USV ……………………………...……..…… 39

Figure 3.6: The (CAS) *Detailed Explanation of Figure 3.5 …..……………….… 40

Figure 3.7: The range of Ultrasonic sensors ……………………………...…...….. 41

Figure 3.8: Ultrasonic sensors connection …………………………...…...………. 42

Figure 3.9: The range of LIDAR-Lite sensors ………………………..….……….. 42

Figure 3.10: Lidar lite sensors connection …………………...………….…...…… 43

Figure 3.11: The range of RPLIDAR sensor ………………………...…………… 43

xii
Figure 3.12: The shape of RPLIDAR A3M1 ……………………….....………….. 44

Figure 3.13: The shape of LattePanda Alpha …………………………....……….. 44

Figure 3.14: Hardware connection of the proposed system ……………....………. 45

Figure 4.1: The components of the USV system ……………....…...….…...…….. 47

Figure 4.2: The components of the CAS system …………………...…………….. 48

Figure 4.3: USV (auto guard) initial position ……………...………...…...………. 49

Figure 4.4: USV (auto guard) next position ……………...……………..…..……. 49

Figure 4.5: USV route update …………………………………….………..….….. 50

Figure 4.6: USV with autopilot case ………………………………...…….…..….. 50

Figure 4.7: USV with autopilot case ……………………………………………… 51

Figure 4.8: The USV return to the control …………………………..………...….. 51

Figure 4.9: The USV discovered the new obstacles ……………….....……..……. 52

Figure 4.10: New information about obstacles ………………..…..……..……….. 52

Figure 4.11: New information about obstacles ………………...….……..……….. 53

Figure 4.12: USV returned back to the focused route …………….……………… 53

Figure 4.13: USV returned back to the focused route ……………….…...………. 54

Figure 4.14: The flowchart of the simulation …………………………………….. 56

xiii
 LIST OF TABLES

Table 2.1: The specifications of Arduino UNO ………………………………...… 22

Table 2.2: The description of Arduino Mega ………………………………….….. 23

Table 2.3: The description of those three kinds of LattePanda boards [26] …..….. 24

xiv
 LIST OF ABBREVIATIONS

CAS : Crash Avoidance System

cRIO : Compact RIO

DEM : Digital Elevation Models

DTM : Digital Terrain Models

EKF : Extended Kalman filter

GPS : Global Positioning System

IDE : Integrated Development Environment

IoT : Internet of Things

KF : Kalman Filter

LCD : Liquid Crystal Displays

LIDAR : Light detection and ranging

MCDA : Multi-Criteria Decision Analysis

MSD : Mine-Sweeper Drone

PCB : Printed Circuit Board

RFS : Random-Finite-Sets

SLAM : Simultaneous Localization and Mapping

SR : Sonic Ranger

UAV : Unmanned Aircraft Vehicles

USV : Unmanned Surface Vehicles

UUV : Unmanned Underwater Vehicles

WSM : Weighted Sum Model

xv
 INTRODUCTION

1.1 Introduction:

Water rescue is a difficult task. Lifeguards must adapt to their particular area’s
currents and landscape, and much of the response hinges on quick recognition of
persons in trouble. When conditions are particularly challenging, it is even more
difficult to retrieve an individual from the water [1].

The past decade of research in marine and field robotics has led to the establishment
of a new class of small-sized unmanned surface vehicles (USVs) [2].

The USV is a watercraft that operated without the crew. USV can be controlled
remotely by an operator from different locations such as another vehicle or ground.
Moreover, it can be used to explore and map the sea environment. In this way,
tracking of the sea animals, the rescue of the people and many operations in the sea
will be possible to research and industrial studies [2].

The work is done on a composite construction project that has the ability to operate
under all environmental circumstances. Furthermore, works to accomplish all the
tasks assigned to it in relation to the work of the USV. For instance, the survey and
provide specialists with the required information. In addition, it saves people who are
vulnerable to drowning. It will be the main work of the vehicle so as to reduce the
drowning conditions and access to people under different circumstances and at high
speed. In this research, USV will be studied and worked on a part of this project
which is to protect the vehicle from hazards and obstacles that may hinder the work
of the vehicle.

During the USV freely tours, it might be an exhibition of some of the obstacles that
may be stationary or moving. These obstacles may cause damage to the vehicle or
the obstruction of the movement. With the help of some equipment and the
information

1
obtained from the GPS and further the information stored in advance in the database
of the USV, it can avoid those barriers.

The Internet of Things (IoT) indicates the conception of extending Internet

connectivity beyond customary computing stages such as personal computers and
portable devices, and into any extent of customarily "dumb" or non-internet-enabled
physical gadgets and regular objects [3]. The sensors and components that contact
with the Internet and will be used in Unmanned Surface Vehicles (USV) are just
kinds of IoT.

A navigation algorithm supplying a robust localization technique is needed as the

speed of information and the position of the vehicle is to be decided with high
precision via the usage of faulty measurements [4].

The autopilot algorithm essentially renders the motion of a surface vehicle without
continuous human intervention possible. Moreover, it is required to existing complex
and agile maneuvers guarantying the steadiness of the vessel. However, the
procedure plan of the autopilot is quite difficult regarding the nonlinear and coupled
dynamics of the system [4].

1.2 Motivation:
The vehicle has been built to carry out humanitarian and scientific work together to
provide a lot of things, including material, the tasks were collected in one vehicle, as
well as provide the effort and energy
The vehicle will operate with a lower amount of energy and will be powered by clean
energy, which in turn will be an auxiliary factor to protect the environment from the
risk of pollution

a) Humanity:
It has been mentioned previously that the vehicle was built for humanitarian
work and this will be its main work as the main task of the vehicle is to save the
lives of citizens from drowning. There are many drowning cases occurring every
day and the whole world has noticed the events of 2016 that claimed the lives of
many innocent people. It is the responsibility of everyone to take care of the lives

2
 of the innocent. This is why the responsibility rests with the researchers. The
efforts have been directed towards contributing to finding a solution to this issue.
The shape and capabilities of the vehicle enable it to reach places that are
difficult to reach by the Coast Guard and are also lighter than ordinary boats,
which helps to get more quickly from the rescuers to the place required. On the
other hand, the vehicle is designed to work in various environmental conditions
and this is what distinguishes it from others, which is why the world today is
interested in such type of projects and work on developing them. The self-driving
or remote-controlled vehicle is easy to move and equipped with a number of
sensors and cameras, which the commander can easily control. This means that
they can reach shipwreck sites or fireplaces in the water and rescue people there.

b) Scientifically:
The construction of the vehicle was also designed to carry out research and
scientific studies. The body of the vehicle is equipped with many sensors and
measuring devices. It is equipped with a sonar device to measure the depth of
water and also helps to know the places of obstacles and provided the vehicle
with a temperature measuring device and many sensors and cameras. All these
devices collect information from the external station and send it to a central
calculator for the purpose of storage, and also upload a copy of the information to
the cloud through the Internet to see this data all those who need it.

1.3 Problem Statement:

The basic problem is basically humane. Many people have lost their loved ones
because the coast guard did not arrive in the right place at the right time or had
difficulty getting in because the weather conditions were not appropriate.

Previous research has built a vehicle for scientific purposes and a vehicle has not
been built to save people from drowning and exploring together.

In the case of the subject of rescue, the vehicle is based on the determination of the
points and be directed in a straight (linear) means that the vehicle is exposed to the
dangers of obstacles encountered in the way.

3
1.4 Aims and Objectives:

The applications of the unmanned vehicles in the previous few decades especially in
the army and civil sectors have seen a big tendency in developing. Additionally, the
tasks, which can't be carried out with required precision and effectivity through
human abilities or might also possibly endanger human life, are expected to be
primarily performed with the aid of these vehicles in the near future. Unmanned
surface vehicles (USV) that have been designed to cross in continuous contact with a
surface of the water is the substance of this study about in actuality concentrates
upon

The USV is a watercraft that operated without the crew. USV can be controlled
remotely by an operator from different locations such as another vehicle or ground.
Moreover, it can be used to explore and map the sea environment. In this way,
tracking of the sea animals, rescue of the people and many operations in the sea will
be possible to research and industrial studies [2].

The aim of the research is to protect the vehicle from the obstacles it faces. As the
vehicle can make the right decision when it reaches the stage of danger. That is, if
the controller is busy or not seeing obstacles, the central calculator in the vehicle
makes the most appropriate decision. By taking the most appropriate route based on
data from sensors and a positioning device that works simultaneously with the
movement of the vehicle.

That will be carried out a novel vehicle design with real-time obstacle avoidance
using IoT devices. To best of our knowledge, there is no previous study that
investigates the USV system by using RPLIDAR sensor, LIDAR-Lite sensors, and
Ultrasonic sensors via LattePanda Alpha, which is a single-board computer.

1.5 Thesis Structure:

In this study, new applications for the unmanned surface vehicle (USV) are studied
and implemented. A brief overview of the subject matter is given and the motives for
this work are presented in Chapter 1

4
Chapter 2 provides a review of critical literature, the history of USV platforms and a
brief explanation of the vehicle's operation and protection system (the crash
avoidance system CAS) that will protect the vehicle from hazards. This part of the
thesis will also examine the components and the software of the protection system
installation and how it works in general.
Different path mapping algorithms are the most common algorithms, and some of
these types will be talked about: A * algorithm, weighted sum model (WSM) also
there are Kalman filter (KF) and SLAM algorithm. There will also be a full
explanation of the parts of the system and how each part of these islands and how to
link. All this information will be mentioned in chapter 3.
The main results of this study, as well as the simulation results that were constructed
to obtain high accuracy results and ensure the proper functioning of the system are
summarized, will be in Chapter 4.
Chapter 5 concludes the thesis by providing a brief summary of the study and
incorporating recommendations for further possible research work.

5
 LITERATURE REVIEW AND REQUIREMENTS

2.1. Overview:

In this section, some similar studies are identified in the past and some methods and
principles of work are explored. Similar research studies are presented below:
In [5] the author simulated and implemented a boat. To control it autonomously, and
he used a fusion algorithm based on EKF is developed. Model is derived based
totally on experimental tests, the algorithm used to be evaluated in different
situations with different values of environmental disturbances (wave, wind, and
current) as long as different obstacles.
In [6] the author used SLAM-robot, based on the concept of random-finite-sets
(RFS). An overview of the RFS-based SLAM algorithm was presented showing its
advantage when using low-cost high-error sensors. This robotic platform is remotely
managed to move and scan unknown environments using a differential algorithm of
drive system, sending immediate function feedback to the faraway operator. The
platform was able to successfully explore and map surrounding indoor environments
providing reasonably high accuracy with the estimated paths also very accurate.
In [7] the author proposed sensing system for vehicles based on lidar. The ability of
the lidars is measuring distance up to 40 meters. The lidars were mounted onto the
stepper motors. The Lidars have been installed on the composite engine where the
contact with the lidars and movement of vehicles controlled by a Compact RIO
(cRIO). The Compact RIO (cRIO) managed the communication with lidars and
motion of the motors, additionally, the algorithm ran on the real-time section of the
cRIO. The algorithm is constructed on the idea of the nearest neighbour method and
indicates promising consequences on the synthetic data.
In [8] This paper indicates the use of less costly sensors fusion for obstacle detection
and avoidance: Sonic ranger (SR) and Light detection and ranging (LIDAR). Further

6
sensors are also set up on the UAS. Experimental statistics results indicate that
sensors integration is very beneficial when climate conditions come to be critical, in
precise when the temperature increases.
In [9], the author used in this paper uses the Global Positioning System (GPS)
capabilities with fixed, linear, and exponential scale configurations used to slow
down USV when the victim's site USV approaching.
The USV controlled by using LIDAR sensor, LIDAR-Lite sensors and ultrasonic
sensors. Ultrasonic sensors principle is to use ultrasound in the known air velocity
and measure the time when sound waves reach obstacles and reflect back, then
calculate the actual distance [10].
The Lidar-Lite measures the distance of an obstacle by sending a Near-Infrared laser
signal and when the signal reflected by an obstacle. It calculates time delay between
the transmission and reception of the signal. This translates into the distance using
the speed of light [11].
RPLIDAR is an improved generation of the LIDAR-Lite. It uses the same principles
for calculations.
All operation will be carried out using LattePanda Alpha which is a single-board
computer that runs a full version of Windows 10 [12].

2.2. History of USV:

The Second World War witnessed the first experience with the USVs. Canadians
developed the concept of the COMOX torpedo in 1944 as a pre-Normandy the USV
before the invasion of designed to smoke during the invasion as an alternative to the
plane. At the same time, the US Navy and showed several types of "Demolition
Rocket Craft" earmarked for demining and obstacles in the surf area [13].

Expanded USVs applications after the war, where USN unmanned boats used to
collect samples of radioactive water after the explosions of atomic bombs Able and
Baker at Bikini Atoll in 1946. The draft “Drone" is of the navy of the defence of the
mines in the United States which dates back to the era of the 1950s, the experience
and testing demining boat for after 1954 [13].

By the 1960s of the last centuries, the Navy used unmanned boats "aviation rescue"
rely on remote control boats after the exercise of launching missiles and use Ryan

7
Firefish boat target in the Artillery Training. As unmanned aircraft UAV continued
development and use of USV and evolved over the years [13].

Continued interest in USV as unmanned surveillance aircraft and minesweeping and

other dangerous tasks in growth after the 1950s for obvious reasons and ensure
further development of the American navy the small "Unmanned vehicle"-a 15 feet
USV to deploy unmanned ammunition, that was developed quickly and deployed in
10 sets a vehicle fleet in 1965 during the Vietnam War. As developed the USVs great
mine-sweeper drone (MSD) and published in Vietnam in the late 1960s [13].

Many countries have recognized the value of recruitment systems unmanned mine-
clearance and the development and dissemination systems. For example, Danish
STANFLEX and Troika groups in Germany (the ship manned three unmanned
planes), unmanned aircraft in the Netherlands, RIM drones in the United Kingdom,
the SAM II ACV drones in Sweden, and SAM ACV Japanese worked by
Hatsushima Class MCM vessels [13].

Since the 1990s, expanding the use of USV with a number of important research and
development programs, for example, the OWL MK II and Roboski. Nevertheless, in
the last decade years, USV has an attractive role in different activities, such as the
exploration of the ocean and the coastal lines and control the environment. To track
and control the USV to complete the task. The system monitors the subprogram for
navigation of the traffic data in real-time and gets from the various maritime
navigation sensors to determine the location, speed, and path of the USV. The sub-
system is responsible to initiate safe routes, in accordance with the specific task,
using the information that is provided from the sensors sub-system by the movement
[14].

2.2.1 USV prototypes:

Figure 2.1 illustrates the main model USV models and overview in the timeline.
Focusing on the control strategies and development of navigation should be noted.
Follow this platform path in advance and lacks the ability to plan the route, including
the avoidance of a collision. Since 2005, detect and avoid obstacles has gained
prominence in the development of USV systems. The USVs audit found that have

8
developed in the past two decades, they are used in a wide range of educational,
military tasks and civil [14].

9
Figar2.1: an overview of USV in the timeline [14]

10
 2.2.2 Employment of USV

The governments have been greatly expanded it used the USV, where it has recently
worked in various fields, including Environmental missions, Scientific research,
Military uses, Ocean resource exploration, and other applications [15].

• Scientific Research: The USV recently entered into many scientific

subjects, to reduce risks to humans and decrease effort and time to complete
the task. On the other hand, it reduces the material cost of labour. Some
scientific research by using USV: the bathymetric survey, changes and
migration in major ecosystems and ocean activities, the biological
phenomena of the ocean, cooperation of multi-vehicle (like cooperative work
between air, ground, underwater, surface waters vehicles); operating and
propulsion systems, and experimental platforms for the purpose of test
structures, also the communication and sensor equipment’s, as well as
control strategy [15].
• Environmental missions: The USV contains many sensors to collect data,
which can predict the notification and reporting of cases before they occur
based on that information. some environmental tasks actions of the USV:
they are monitoring, evaluation, and sampling; catastrophes (such as tsunami,
the eruption of a submarine volcano, hurricane) helped predictability and
management, and the response of emergency; measurements of pollution and
cleaning [15].
• Ocean resource exploration: Some types of USV specializes in research and
collection of information for the exploration of the resources of the oceans,
such as oil, gas and my explorations; or the tasks of building the platform,
building pipeline and maintenance abroad [15].
• The military uses: The basis for building the USV was to carry out military
actions to protect the armies during the battles and normal cases. Therefore,
considerable interest by developed countries in this type of vehicles, which
carried out the following: the harbour, port, the coastal surveillance,
patrolling, and reconnaissance, rescue and search; mine Countermeasures;
anti-terrorism force protection, weapons platform remotely, Unmanned boat
target [15].

11
 • Other applications: Recently the USV has been used in most applications
that make life comfortable and less cost. The most important applications that
the USV provide it is Transportation; The relays of mobile communications,
the platform to refuel the vehicles of USVs, unmanned underwater vehicles
(UUVs), unmanned air vehicles (UAVs), and other populated [15].

2.3. Crash Avoidance System:

The Crash Avoidance System (CAS), also known as a collision mitigation system, is
the Auto Safety System in vehicles that designed to reduce or prevent the collision
severity. The CAS is using the radar (in all weather conditions) and sometimes laser
(LIDAR), the camera (which are used to recognize the pictures) to detect an
imminent accident. GPS sensing devices can detect fixed dangers through a database
on site.

Once the collision is detected, these systems provide a warning to the driver. When a
collision is imminent, they acted independently without any action from the driver
(through the brake or the router, or both). Such a case of high speed’s vehicle has to
consist appropriate braking, while at low speeds vehicle (for example, less than 50
km/h (31 mph)) the avoiding crash by router might be more appropriate than higher
vehicle speeds if the passages are clear.

2.4. The sensors:

A device that detects and reacts to some input type from the physical environment is
called a sensor. The particular input may be heat, light, moisture, motion, pressure, or
any of a great number of environmental phenomena. in general, the output is a signal
that is transformed into a human-readable display at the location of the sensor or
transmitted electronically over further processing or a network for reading.

There are so many kinds of sensors as said before but, this thesis will be talked about
the movement sensors.

12
Motion sensors in various systems including crash avoidance, automatic doors, home
safety lights, and etc., typically they issued some types of energy, like ultrasonic
waves, microwaves, or light/laser beams and detect when the flow of energy is
discontinuous by entering something its path.

2.4.1. Ultrasonic sensors:

Ultrasonic sensors are regularly used in automation tasks to level measurement,

position changes, measure distance, such as the detection of the presence or in
exclusive applications of detection sensor to discover obstacles is an ultrasonic
sensor. This kind of sensor depends on the principle of the ’time-of-flight’. The
transmitter periodically sends ultrasonic pulses (usually forty kHz). If there is an
object within range, then this object will reflect sound waves and that will be
obtained through the receiver. Because of the velocity of sound is finite, there will be
a measurable difference in time between the transmitting and the receiving of the
pulses. Knowing the velocity of sound and the difference in time, and then can be
calculated the distance to this object [16].

These sensors are inexpensive sensors and that is the benefit to the users and can
operate in different environmental conditions where other sensors would fail also
they have low strength consumption [17].

One of the limitations of Ultrasonic, if there are many ultrasonic sensors used close
to each other that may happen an interference and the signals that receive to the
receiver become from another sensor, one of other limitation of ultrasonic sensors is
that they generally cannot see obstacles which have their normal at an angle larger
than forty-five degrees to the transmitter. That is because the sound will deflect from
this obstacle and will not reflect. Therefore, the pulses that sent out of the sensor will
not receive. The absorbing materials certainly will not reflect the sound waves also,
so these will be invisible also to the ultrasonic sensor [17].

As the name indicates, by using ultrasonic waves the ultrasonic sensors measure
distance.

13
 2.4.1.1. Fundamentals of ultrasonic waves:

The propagation of waves in a frequency range above twenty kHz, where the
frequency is inaudible conventionally to human hearing that is considered known as
Ultrasound. The travels of ultrasound wave from one material to another depending
on the medium mechanical deformation. The power transmitted is in a systematic
movement as kinetic energy within the particles of the medium move.

Ultrasonic strategies have been utilized to measurement and monitoring in

engineering and another field such as medical sciences, materials, and construction
since the late 1990s. It is a non-harmful and non-intrusive technique that constantly
measures the ideal properties of the material interface [18]. Figure 2.2 shows the
frequency range and the number of kinds of ultrasound through it.

Figure 2.2: Acoustic sound ranges and frequency values

The first range is known as infrasound, from 0 - 20Hz. This is followed by the
frequency range of 20 Hz - 20 kHz than known as the acoustic sounds. Finally, the
last frequency ranging from 20 kHz to 200 MHz that is the ultrasound, Therefore, the
frequency is the number of the accomplished cycles of the sound of a wave per unit
time and is measured in cycle per seconds (cps) or Hertz (Hz) [18].

2.4.1.2. Ultrasonic measuring:

The principle of Ultrasonic ranging is to use ultrasound in the velocity of the air as
the known and measurement of the time when the waves of sound reach obstacles
and reflect back after that the actual distance can be calculated [35].

14
The time that taken by the pulse is actually the time of the travel of ultrasonic
signals, however, we need only half of this. Therefore, the time is taken as T/2.

𝐷𝑖𝑠𝑡𝑎𝑛𝑐𝑒 = 𝑆𝑝𝑒𝑒𝑑 ∗ 𝑇𝑖𝑚𝑒/2

The speed of sound in the materials depends on the temperature. The speed in the air
is approximately 345 m/s, in a bar of steel is 5000 m/s and in water is 1500 m/s.

To measure the distance of a sound sign transmitted, it should be reflected. This

sound sign is a longitudinal sound wave that strikes a flat surface. The sound is then
reflected, provided that the dimension of the reflective surface is large compared to
the wavelength of the sound [19].

• Surface

The perfect target surface is smooth and hard. This surface reflects a greater signal
amount than a soft surface or rough surface. A weak echo is the conclusion of a soft
or small object. This minimizes from the operating distance of the ultrasonic sensors
and reduction its accuracy [19].

• Distance

The shorter the range from the ultrasonic sensor to an object, the returning resonance
is stronger. Therefore, as the distance raise, the object requires higher reflective traits
to return an adequate resonance [19].

• Size

A massive object has a greater surface to reflect the sign than a small one. The
surface location recognized as the goal is usually the location closest to the sensor
[19].

• Angle

The objects surface inclination facing the ultrasonic sensor influences how the object
reflects. The element perpendicular to the sensor returns the resonance. If the whole
object is at a larger angle, the sign is then reflected away from the sensor and no
resonance is detected [19]. Figure 2.3 is to explain all these four cases.

15
 Figure 2.3: the things that affect of ultrasonic sensor measuring.[19]

2.4.1.3. JSN SR04T:

The transducer is a component used to convert an electrical signal into an acoustic

signal (on the sending side) or convert an acoustic signal into electricity (at the
receiving end). as shown in figure 2.4.

Figure 2.4: the shape of JSN SR04T.

This sensor is an ultrasonic distance measuring module with a non-contact distance

detection function it provides 2-400cm non-contact measurement function with the
ranging accuracy can reach up to 0.3 cm. The modules include ultrasonic
transmitters, receiver and control circuit. This product adopts an industrial-grade
integrated, ultrasonic design with waterproof type, and has stable performance.

The specifications of this sensor are measuring an angle is 75-degree, Small size,
easy to use -Low voltage, low power consumption, High accuracy and Strong anti-
jamming.

16
The distance d follows Eq.

𝑑 = (𝑇. 𝑎) / 2

Where T is high-level time and a is the speed of sound (340 m/s in dry air at 20°C).
If no obstacle is detected, in a fixed time delay, the output pin gives a 38 MS high-
level signal [18].

2.4.2. Light Detection and Ranging (LIDAR) sensors:

LiDAR is a remote sensing technology that uses a laser pulse to collect measures,
which can then be used to create three-dimensional models and environments and
maps of the objects. LiDAR is a shortcut of Light Detection and Ranging. LiDAR
operates in a manner similar to the Sonar and Radar, but it uses alternative than
sound or radio waves, the light waves from a laser.[20]

A LiDAR System calculates the duration of the Light to reach an object or surface
and reflected once again on the scanner. Then the distance is calculated by using the
speed of light (The speed or velocity of light is 299,792,458 m/s). these
measurements are known as the ‘Time of Flight’. LiDAR systems can about
1,000,000 pulses/s. [20]

The LiDAR systems are used commonly in the functions of the survey. In view of
their capacity to collect 3D measurements, the scanning systems of laser are common
to clear the built environment (like buildings, railways, and roads) as well as the
establishment of Digital Terrain Models (DTM) also the Digital Elevation Models
(DEMs) of the particular landscape.[20]

The lidar a measures distance by calculating the time delay between the transmission
of a Near-Infrared laser signal and its reception after reflecting off of a target. This
translates into the distance using the speed of light.

Due to the shorter wavelength of light, LiDAR’s can be a lot more accurate than
RADARs in detecting obstacles

17
 2.4.2.1. Lidar-lite sensors:

Lidar-Lite targets the need for very compact optic distance measurement, and high-
performance sensors for applications like UAV’s, drones, and robotics where a high
performance, low-power, very small, reduced cost sensor and it is desirable [21].

The single-chip processing solution in combination with minimal supporting

hardware enables a new class of optical distance measurement sensors. Lidar-Lite’s
goal is to make an easily configurable sensor module with the availability of
technology that may be used as the basic block building for sensor applications in
projects [21].

In fact, the Lidar-Lite has a wide distance detection. There are many factors that may
have an effect on the sensor’s maximum range. The angle of the sensor and the target
reflectivity are the two that have the most influence [21].

There are ways to increase the speed of measurement. However, the sensor will lose
the sensitivity of the laser and possibly decreasing the maximum range.

2.4.2.2 The RPLidar:

RPLIDAR is a 2D Laser scanner (LIDAR) Low Cost and 360 degrees was developed
by RoboPeak. The system can survey the 360 degrees in different ranges started from
6 m to one km. The producing two-dimensional cloud bitmap data can be used in
localization, mapping, and modelling of object/environment [22].

RPLIDAR's frequency scanning arrived at 5.5 Hz when taking in each round 360
points. Can be configured with a maximum of 20 Hz. RPLIDAR is the laser
triangulation measure system in basic. Also, it can work well in all types of the
internal environment and the external environment without the rays of the sun [22].

RPLIDAR has a system of a range scanner and system of a motor. After Playing
Each sub-system, start RPLIDAR in recycling and scan in a clockwise direction. The
user can obtain the survey data through the contact interface (serial port or USB)
[22].

18
 • Mechanism:

RPLIDAR depends on the principle of the laser triangulation ranging and used in the
acquisition and processing of high-speed vision hardware developed by RoboPeak.
the distance data that measures by the system are more than 2 thousand times per
second and with the output of high-precision distance (contains an RPLIDAR
development unit standard (A1M1-R1). The engine control unit logical RPLIDAR
integrated, for the command (3.3 volts) which can be used to configure the scanning
frequency by adjusting the engine speed, the developer can also choose to turn off
the engine for the purpose of providing energy. The mechanism of RPLiDAR is
shown in Figure 2.5 [22].

Figure 2.5: The mechanism of RPLiDAR [22]

• RPLIDAR can be used in some applications：

• Localization and General robot navigation and (Home cleaning robot/service robot)

• Obstacle avoidance in smart toy’s

• 3D re-modelling and Environment scanning

• Generally Simultaneous Localization and Mapping (SLAM)

2.5. The development boards:

To develop integrated, they need two things: the development board and the
integrated development environment (IDE). The printed circuit board (PCB) with the

19
circuits and devices designed to facilitate experimentation with some of the features
of the microcontroller boards all these are under the name of the microcontroller
development board.[23]

The development board is a single printed circuit board that the microcontroller is
mainly entered into board manufacture. This board has all of the necessary circuitry
for a helpful control task such as processor, chipset, memory, and accessories
peripherals such as liquid crystal displays (LCD), keyboard, serial port, USBs, slot
card of SD, Ethernet, etc. with features of debugging. The intent is that the board will
be immediately beneficial to a developer of the applications, without necessary to
spend effort and time to develop the hardware controller. Also, this will keep them
far away from tampering with being in touch with the boards and the jumper
wires.[23]

2.5.1. Arduino:

The Arduino is the most popular electronics prototyping platform that is the open-
source based on a development environment and a simple Input/output board that
implements the Processing and Wiring language. used to create electronic interactive
applications. The free open-source IDE can be downloaded for Windows, Mac OS X,
and Linux [24].

To download the program from the computer via USB, it creates an interface of
serial communication. For the microcontroller, there is in the operating requirements
such as the oscillation circuit, connecting the external circuits by connecting the
board's pins to other pins and operating power. The Arduino works on a simplified
version of the programming language (C language) [24].

The Arduino microcontroller contains a feature that is an important one. that Arduino
feature is containing the universal serial port (USB) is used to connect to the
computer for the purpose of interchange of programming or data. Because of that,
there is no need for an external programmer which reduces the time and cost [24].

20
2.5.1.1. Arduino Uno:

"Uno" means in the Italian (one). The Arduino UNO board contains all the necessary
support to the microcontroller. The Arduino UNO microcontroller board is
absolutely very familiar for experts and beginners. UNO Should be considered as one
of the first development boards that work microcontroller-based. [23]

The Arduino UNO is the simplest and most powerful prototyping accurate based on
the ATmega328P microcontroller. It has 14 digital i/o pins (6 of them can be used as
outputs pins), and other 6 of them as analogue inputs, a USB connection, a power
jack, a reset button, and an ICSP header. It contains everything necessary to support
the microcontroller. The specifications of the Arduino UNO are shown in Figure 2.6
and Table (2.1).[23]

Figure 2.6: The specifications of Arduino UNO [23]

21
 Table 2.1: The specifications of Arduino UNO

2.5.1.2. Arduino Mega:

The Arduino Mega is a board based on the ATmega2560 microcontroller and it is

designed for more complex projects than Arduino UNO. It has 54 digital I/O pins (14
can be used as PWM outputs of which), (16 inputs analog), four as a UARTs
(hardware serial ports), (a 16 MHz crystal oscillator), a power jack, a USB
connection, a reset button, and an ICSP header. To support the microcontroller, the
Arduino Mega includes everything needed. simply connected it to a computer by
using the USB cable or power it by using an AC-to-DC converter or by using a
battery to get started. Figure 2.7 shows the …. And table 2.2 shows the description of
Arduino Mega [25].

22
Figure 2.7: The description of Arduino Mega [25]

Table 2.2: The description of Arduino Mega

23
 2.5.2. LattePanda Alpha (mainboard)

Back at the tail end of 2015, a single-board computer, and it had a Quad-Core Atom
processor, that was under the name of LattePanda

A LattePanda is a single board computer with complete Windows 10. It can do

anything that a systematic PC does and has everything an ordinary PC has. The
LattePanda is compatible with almost every device known such as printers, cameras,
joysticks, and more. Any peripheral devices that work on the PC will work on it [26].

A LattePanda arrives pre-installed with a (Windows 10 Home Edition), that means

can run powerful apparatus such as Java, Visual Studio, NodeJS, Processing, and
more applications. With present APIs, the developer can develop their own software
and hardware projects on a LattePanda as they would on a PC by using C#,
JavaScript, Python etc [26].

The company is back two years later (in 2017) with new boards, the LattePanda
Alpha and the LattePanda Delta.they became more powerful than LattePanda. Table
2.3 below shows the description of those three kinds of boards [26].

Table 2.3: The description of those three kinds of LattePanda boards [26]

24
2.6. Electrical Motors

An electrical motor is a machine that has introduced one of the largest developments
in the fields of engineering and technology ever since the invention of electricity. A
motor is nothing however an electromechanical system that converts electrical power
into mechanical energy. The following chart is the kinds of electrical motors as
shown in Figure 2.8 [27].

Shunt Motor

Series Motor
DC Motors
Compound Motor

Permanent
Magnet Motor

Induction Motor
Electrical Motors AC Motors
Synchronous
Motor

Stepper Motor

Servo Motor
Special Motors
Universal Motor

DC Brushless
Motor

Figure 2.8: Types of electrical motors.

In this thesis will choose Servo motor for its advantages.

25
 2.6.1. Servo motor

A servo is a small device which has an output shaft which positions on the coded
signal. It is a rotary or linear actuator that allows for precise control of the angular or
linear position, velocity and acceleration. Figure 2.9 shows the types of servo motors
[28].

Advantages:

• High efficiency.
• High output power is relative to motor size and power.
• High-speed operation is possible.
Disadvantages:

• Complex, require an encoder.

• High cost.
• The motor can be damaged by sustained overload.

Servo
Motors

AC Servo DC Servo
Motors Motors

Positional Continuous
Linear
Rotation Rotation

Figure 2.9: The types of servo motors.

26
All of the components listed above have been approved with the selected algorithm
in addition to the CAS system. The details will be mentioned in the next stages and
will mention the way each part works, in addition to some types of algorithms that
help facilitate the work of the system with high accuracy to predict the risks and to
protect the vehicle from obstacles.

27
 ALGORITHMS & SYSTEM DEVELOPMENT

3.1. Introduction:

This chapter presents the core development concepts, including engineering issues
that have been identified and diagnosed to build the model of unmined service
vehicle (USV). The primary concern of this part of the work is to adapt the algorithm
system to the required model and to develop the system to work in real-time.
Concisely, the journey from theory to prototype, and the evolution of novel
requirements which required to be handled are explained in detail. The first part of
this chapter state of the art for algorithms that already used within USV.

The section discusses the motion model and the observation model developed for the
system. The next part discusses the need for calibration of the sensor with respect to
the robot and the development of a calibration system to be able to transform the
coordinates in the sensor frame of reference to the robot’s frame of reference.

3.2. A* Algorithm

A* algorithm is one of the search algorithms for discovering the minimum cost path
via a given map to signify uniformed dimension cells. It has been widely used in
many real-time method games and is likely the most famous route-finding algorithm.

It offers values to the nodes in many ways. Each value node is the sum of the actual
cost to that node from the start and the heuristic estimate of the final value from the
node to the goal [32].

This algorithm makes the most efficient use of the heuristic characteristic of the
search, as an alternative than some other algorithm that uses the same heuristic that
product fewer nodes and finds an ideal path.

28
The optimality of A* algorithm represents the smallest quantity of nodes (cells)
located by using the heuristic search that ensures to locate the path. Due to this
optimality and the reason for producing the direction with the least distance cost
[32].

A* algorithm uses a starting point and a destination point to produce the desired path,

Figure 3.1: Starting and Destination Points in A* Search Algorithm [32]

A* algorithm starts to execute by looking at the starting node first and then
expanding to the surrounding nodes (Figure 3.2).

This operation continues until the destination node is found.

Figure 3.2: Starting the A* Search Algorithm [32]

f(n), is the combination of two scores for calculating the values of the algorithm:

𝑓(𝑛) = 𝑔(𝑛) + ℎ(𝑛)

Where g(n) is the cost of the path from the starting node to any node n, h(n) is the
heuristic estimated cost from any node n to the goal.

29
The complexity of time for A* algorithm depends on the number of nodes that have
been explored from the searching map.

The flowchart of the conventional A* algorithm is shown in Figure 3.3.

Figure 3.3: Flowchart of the conventional A* algorithm

3.3. Weighted Sum Concept:

The weighted sum model (WSM)is the first-class acknowledged, most commonly
used in decision-making techniques and simplest multi-criteria decision analysis
(MCDA) / multi-criteria decision-making technique for evaluating a number of
alternatives in phrases of a range of decision criteria. It combines the several goals
and weights corresponding to these targets to create a single rating for every choice
to make them comparable.

30
 3.3.1. Weighted Sum Model:

The Weighted Sum Model (WSM) is the most widely used technique in the problems
of the optimization in multi-objective. Because it tries to combine a lot of goals and
weights referring to these objects to make a single rating for every choice that could
be compared among these weights. The following equations used in WSM model as
shown.

𝐴𝑊𝑆𝑀
𝑖 = ∑𝑛𝑗=1 𝑊𝑗 𝑎𝑖𝑗 …Eq 3.1

𝐴𝑊𝑆𝑀
∗ = 𝑀𝐴𝑋𝑖 ∑𝑛𝑗=1 𝑊𝑗 𝑎𝑖𝑗 …Eq 3.2

In these equations, the WSM score for an alternative values Ai denoted as

𝐴𝑖𝑊𝑆𝑀−𝑠𝑐𝑜𝑟𝑒 can be computed by adding the products of a weight 𝑤𝑗 with its
referring parameter, where 𝑎𝑖𝑗 is the value of the objective. For instance, the cost of
the weight which has to be spent to perform the desired. The maximum WSM score
is the best alternative value chosen as the one which has the maximum WSM score
for given parameters. The multi types of objectives can be assumed to be positive:
the higher the score, is the better the alternative. Assuming that objective to be
negative (in case of cost models), the best alternative has equivalently the lowest
score [31].

3.3.2. The WSM Method:

The weighted sum model (WSM) is probably the most commonly used approach,
especially in single-dimensional problems. If there are 𝑚 alternatives and n criteria
then, the best alternative is the one that satisfies (in the maximization case) the
following expression

𝐴𝑊𝑆𝑀
∗ = 𝑀𝐴𝑋𝑖 ∑ 𝑊𝑗 𝑎𝑖𝑗 , 𝑓𝑜𝑟 𝑖 = 1, 2, 3, . . . , 𝑚.
𝑗=1

31
Where: A* WSM is the WSM score of the best alternative, n is the number of
decision criteria, aij is the actual value of the i-th alternative in terms of the j-th
criterion, and Wj is the weight of importance of the j-th criterion.

The idea that governs this model is the additive application assumption. However,
the entire fee of alternative is identical to the sum of the products as proven in the
subsequent phase. In one-dimensional, all units are the same (e.g., money, meter), the
WSM can be used reliability. In the other side, the most difficult part of this method
is emerging when it is implemented to multidimensional MCDM problems.
Nevertheless, in combining one of a kind dimensions, and therefore exceptional
gadgets, the additive software assumption is violated and the result is equivalent to
"adding apples and oranges"[29][31].

3.3.3. The Weighted Sum Implementation:

Example I

Suppose that an MCDM problem involves four criteria, which are expressed in
exactly the same unit, and three alternatives. The relative weights of the four criteria
were determined to be:” W1 = 0.20, W2 = 0.15, W 3 = 0.40, and W 4 = 0.25 “. Also,
the performance values of the three alternatives in terms of the four decision criteria
are assumed to be as follows [29]:

Nevertheless, the following decision matrix shows some data for MCDM problem
that can be summarized as follows:

32
When equation (2-1) is implemented in the previous example data the scores of the
alternatives could be:

𝑨𝟏 𝑾𝑺𝑴 − 𝒔𝒄𝒐𝒓𝒆 = 25 𝑥𝑂. 20 + 20𝑥𝑂. 15 + 15 𝑥 0.40 + 30𝑥𝑂. 25

= 21.50.

likewise, we can get:

𝑨𝟐 𝑾𝑺𝑴 − 𝒔𝒄𝒐𝒓𝒆 = 22.00,

𝑨 𝟑 𝑾𝑺𝑴 − 𝒔𝒄𝒐𝒓𝒆 = 20.00.

To sum up, the best alternative value (maximum value of the given data) is
alternative A2, due to the value has the highest WSM score among the others which
is 22.00.

Example II

The table below consist of many parameters of alternative and values of criteria for
choosing the best value of alternative among these A1, A2, A3, A4 and A5. In the
other side, the criteria values that determine the picking process is K1, K2 and K3.

33
Refer to Table 1, the value of criteria K1 is known and given by 0.3 or 30%, the
value of criterion K2 is 0.4 or 40% while the criterion value K3 is 0.3 or 30%. To
compute the value of each alternative by using WSM formula eq. 1. Refer to
equation 1, A4 is chosen as the best choice, due to the magnitude of A4 is the highest
value of all the given alternatives. Table 2 shows the result as follow [30]:

3.4. Kalman Filter

Kalman filter (KF) is a mathematical equation set to produce effective calculations

offering options for the least-squares method. KF has many effective aspects: it helps
estimations of the previous, present, and even subsequent states, KF could be
estimated the variables of a variety of processes. In mathematical terms, the Kalman
filter estimates the states of a linear system [33].

The fundamental operation is achieved by using the KF that generate estimates of the
preferred values, first off predict a value, then estimating the uncertainty of the above
value and discovering a weighted common of each the expected and measured
values. The least uncertainty weighted is given to the value [33].

The Kalman filter has been notably used in estimating the country situation of a
process. It has been broadly applied in a number of fields like navigation, monitoring
object, control systems, robotic movement planning, laptop vision and many extras.

In this thesis aims to take a more teaching-based strategy to current the Kalman and
Extended Kalman Filters from a practical usage perspective. One of the principal

34
differences between this work and the contemporary state of the art Kalman filtering
tutorial is that the statistical concept.

Figure 3.4: Kalman filter algorithm [33]

Figure 3.4 depicts the everyday step of the Kalman filter algorithm. It consists of
several iterative processes including the prediction of country and error covariance,
dimension updates with computation of Kalman reap and an estimation process, and
lastly, there is computation of error covariance which shows how correct estimates
are. The Kalman filter structure has one measurement enter 𝑍𝑘 and one estimation
output 𝑋̂𝑘. There are 4 system model A, H, Q, and R. A is the state transition matrix,
H is the state to measurement matrix, Q is the covariance matrix of transition noise,
and R is the covariance matrix of dimension noise. In step III, 𝐾𝑘 is the Kalman gain

35
which depends on 𝑃𝑘 error covariance. Error covariance indicates the difference
between Kalman filter estimation and the real value [33].

3.4.1. Design the Kalman Filter:

The Kalman filter is a statistical algorithm which can estimate the state of the process
given noisy data. The go with the flow of Kalman filter iteratively predicts and
estimate the prediction with entering dimension until some criteria are met. For
indoor positioning, the Kalman filter method consists of a vector of moving object X
in (x,y) coordinates and pace with the coordinates as the sole measurement. The state
and dimension model is given below:

State model:

𝑋𝑘 = 𝐴𝑋𝑘 − 1 + 𝑊

Measurement model:

𝑍𝑘 = 𝐻𝑋𝑘 + 𝑉

Matrix A is the state transition matrix that describes the movement of the machine
which in this case the equation relationship of position and velocity. Matrix H
indicates the relationship between the measurement and state variable [33].

Q and R are the process noise covariance matrix and dimension noise covariance
matrix, respectively. These two parameters will have an effect on the measurement
and prediction of the Kalman filter process. Between these parameters, process noise

36
covariance is difficult to determine, which is decided via experiment or experience. J.
Yim et al. [33] have highlighted primarily based on their experiment the ratio
between Q and R that actually consequences the performance of the Kalman filter.
By following the step, it was discovered that Q equals to 0.00001 gives greatest
results. The initial circumstance as follows was once setup:

3.5. Simultaneous localization and mapping (SLAM) Algorithm

It is the computational issue of building or constructing a map of an ambiguous

situation while at the same time monitoring an operator's area inside it. In another
word, it is learning a map and trying to locate the robot with each other.

SLAM algorithm tries to estimate the position of a certain robot with the same time
finds the map of the circumference. However, the map used by SLAM is required for
localization and to obtain a good position estimator is required for mapping.

37
The SLAM algorithm has many significant properties that direct the development of
vulnerability in both the map and vehicle estimates. These properties incorporate the
convergence of the guide state assesses, the need to keep up consistency in the
estimation procedure and the computational multifaceted nature engaged with
keeping up the covariance framework. three important convergence properties of the
SLAM filter. The three key results can be obtained as follows [34].

1. covariance matrix decreases gradually in the observed data for the determinant of
any submatrix in the map.

2. In the point of confinement, as the number of perceptions builds, the milestone

assessments become completely corresponded.

3. the accuracy of the limited map play a function for initial vehicle uncertainty when
the first landmark is observed [34].

These properties of the SLAM algorithm have significant ramifications for its
utilization of true frameworks. The confirmations successfully exhibit that the
uncertainty in the guide appraisals will in general diminishing to some lower bound
and that the connections between highlights or tourist spots will turn out to be
impeccably known. Given two mean and covariance estimates,

and , the covariance intersect update,

is computed as:

where 0 ≤ ω ≤ 1 and ω can be computed to minimize a certain measure such that

determinant and maximum eigenvalue, etc. of the covariance matrix. This approach
provides a computationally efficient mechanism for performing SLAM. Furthermore,
for the most part, permits enhancements in the vehicle and guide appraisals to be
spread all through the guide when a notable component is watched and is the way to
the combined properties of the calculation [34].

38
The algorithms mentioned earlier are a simple number of algorithms that help to
choose the most suitable route for vehicles. WSM has been adopted in the simulation
program which will be mentioned in Chapter 4 which is a small part of the real
system. The simulation program is built only to ensure the effectiveness of the
system and to choose the most suitable algorithm.

3.6. PROPOSED WORK

In this thesis, Unmanned Surface Vehicle (USV) has a discrete controlled computer
system to control the USV, sense the around of the USV and process data to avoid
obstacle collisions. This computer system’s schematic has been explained as shown
below:

Figure 3.5: System control units of USV

There is a CAS (Crash Avoidance System) Instruction which gets data from LIDAR
sensors and other sensors and sends this datum to CAS main board computer to
process this datum. After the processes of sensor datum, CAS sends analyzed sensor
datum to route creator computer to create available free routes. Route Creator
Computer decides the most secure route to avoid any crash and sends new route
information to the main computer. Finally, the main computer calculates needed
parameters and sends appropriate signals to the Motor Control Units and other digital
systems to move. All these algorithms and instruction systems manage by
LattePanda Alpha main computer which is located into the USV as shown in Figure
3.6 [36].

39
 Figure 3.6: The (CAS) *Detailed Explanation of Figure 3.5
The USV will be controlled to avoid obstacles by using Ultrasonic sensors, Lidar-lite
and Lidar. All steps in the flowchart of (CAS) had been implemented based on
python language which explained below:

3.6.1. Python language:

Python is a high-level programming language, object-oriented with integrated
dynamic semantics primarily for app development and web.
Python is pretty simple and easy to learn because it requires a special syntax that
focuses on readability. Developers can read and translate Python code a lot easier
than other languages.
Additionally, Python supports the use of modules and packages, which ability that
programs can be designed in a modular fashion and code can be reused throughout a
variety of projects. Also, one of the most promising benefits of Python is that both
the popular library and the interpreter are accessible free of charge, in both binary
and supply form.
That makes Python reachable to almost anyone. If you have the time to learn, you
can create some top-notch matters with the language.

40
 3.6.2. Hardware description:
the proposed system that was explained before in Figure 3.6 consist of sets of
different elements which can be described as follow:
1. 12 of Ultrasonic sensors is used on both sides that mean six Ultrasonic sensors
for the left side and six Ultrasonic sensors for the right side of the USV.
Ultrasonic protect the USV from the nearest obstacles which floating on the sea
within range 0.25 to 5 m with a field of view between 15 to 70 degrees
depending on the sensors company. In another side, the number of sensors may
be increased or decreased depending on the field of view. The rays are emitted
from these sensors and if an obstacle found in the sea the echo wave is received
by the ultrasonic sensor and computing of distance between the obstacle and
USV in order to change the USV path. Figure 3.7 shows the range of Ultrasonic sensors on

USV prototype. All connection of Ultrasonic sensors can be explained as shown in

Figure 3.8.

Figure 3.7: The range of Ultrasonic sensors

41
 Figure 3.8: Ultrasonic sensors connection

2. A LIDAR-Lite sensor is used and located in front of the boat and another one at
the end of the USV. They move in 80 degrees as a field of view by using a servo
motor. That allows to detect the objects or obstacles that may hit or cause
damage for the USV, its operation in somehow likely the sonar but the LIDAR
rays could reach up to 15 m distance away from the USV [8]. These sensors are
installed on the Servo motor because the LIDAR sensors laser beams have
directly puled. Figure 3.9 shows the range of LIDAR-Lite sensors on USV prototype, and lidar
lite sensors connection can be explained in figure3.10.

Figure 3.9: The range of LIDAR-Lite sensors

42
 Figure 3.10: Lidar lite sensors connection

3. RPLIDAR is the next generation of laser range scanner developed by

SLAMTEC (A Software and Hardware Company That Uses the SLAM
Algorithm Specially). It works at high turnover speeds and with a range distance
more than 16 meters currently for A2 and 25 meters for A3. It makes available
to 360-degree laser range scanning. Furthermore, the RPLIDAR is equipped
with SLAMTEC OPTMAG technology that works to reduce the lifespan of the
traditional LIDAR system to work steadily for a longer period. The RPLIDAR is
an ideal sensor for working in cost-sensitive projects because it is a cost low
device [10]. Figure 3.11 shows the range of RPLIDAR sensor on USV
prototype.

Figure 3.11: The range of RPLIDAR sensor

43
 RPLIDAR A3M1 is a new generation of low-cost laser scanner developed by
SLAMTEC. It can take up to 16000 samples of a laser can be released per
second with high rotation speed. The device will be the most dependent on the
vehicle [22]. Figure 3.12 shows the shape of RPLIDAR A3M1.

Figure 3.12: The shape of RPLIDAR A3M1

4. A LattePanda Alpha is the first small and advanced panel that can run a full
version of Windows 10. It is equipped with an Intel Quad-core processor and has
excellent connectivity with three USB ports, built-in Wi-Fi and Bluetooth. It also
includes a common processor will be used as a central processor to process and
issue orders in USV [26]. Figure 3.13 shows the shape of LattePanda Alpha
[11]. of Arduino that does all the Arduino works with high quality [37].

Figure 3.13: The shape of LattePanda Alpha

44
5. Arduino UNO is used as a control board that controlled the servo motors and
check them and their movement. The board collects the information and send
that information to the main computer and controller.

6. Arduino Mega is used to connect to the ultrasonic sensors and lidar-lite sensors
because of the kind of microcontroller that makes that board is faster than
(UNO). In another side, the number of pins that the board have is also important
because there are many sensors connecting to the board.

The connection of all hardware components can be explained in Figure 3.14.

Figure 3.14: hardware connection of the proposed system

45
 IMPLEMENTED CIRCUIT AND SIMULATION

4.1. Introduction

In chapter four, the implemented circuit and the findings which are obtained from the
simulation of the USV are presented and discussed. In another side, the problems that
happened in a work also are discussed.

4.2. Integrated circuit:

In this part of the thesis will include the final form of the components and how to
link them and the sequence of work. Figure 4.1 shows that the USV system, when it
starts up, is the first thing that the main controller works as well as those controlling
other parts such as the camera and the sensor. Thereafter the action of temperature
and weather sensors and other sensors, which in turn begins to take the values and
study the ocean surrounding the vehicle. At the same time, the GPS function starts,
which locates the vehicle on the map and gives the controller information about the
area. This is followed by all the work of the CAS system, including the movement
sensors and calculates the distance of the obstacles surrounding the vehicle, if any.
This part of the system is the subject of research, which has been studied parts as
well as the algorithm. Which has been built a simulation program for some parts to
make sure the accuracy of the results, the components of which are illustrated in
Figure 4.2. The last part of the figure shows the lighting of the vehicle, which
contains the headlights of the vehicle as well as lighting alerts for marine navigation.
All these components combined are the integrated circuit of the USV.

46
Figure 4.1: The components of the USV system

47
 Figure 4.2: The components of the CAS system

4.3. Simulation of Obstacle Avoidance System on USV

In this section, Obstacle Avoidance System has been mentioned with the initial
simulation of some sensors that protect the USV.

Simulation of this process prepared with Python development tools, Python language
and libraries. Simulation backend recovers an autonomous route calculation
algorithm to find the fastest route on the sea. The simulation will be able to perform
better calculations with the contributions of TensorFlow machine learning in each
new route request by continuously collecting the generated route data.

Simulation has been used to recognize USV’s movements towards the obstacles. The
purpose of this simulation is to calculate the route from point A to B, without
crashing the obstacles, and to calculate the distance of the vehicle to the obstacles.
The sensors which are around the vehicle, continuously measure the distance of these
obstacles during the simulation of the vehicle's motion. After these measurements,
the simulation delivers the vehicle to the fastest way from point A to point B. This
tells us how the sensors interact with obstacles.

As shown in Figure 4.3, and after the target is determined, the USV is started to work
on the information that stored from the database and showed that there were
maritime obstructions in the vehicle's road that had already been stored into the cloud
from the GPS.

48
 Figure 4.3: USV (auto guard) initial position

The USV is worked on a new survey of the area to verify the information stored and
to update that information as shown in Figure 4.4.

Figure 4.4: USV (auto guard) next position

The USV will also avoid these obstacles and send the alarm to the controller system
to change the path of USV by choosing an alternative safe way to reach the target as
shown in Figure 4.5. The blue line is going straight to the target and the white line is
the new way that was chosen by the algorithm depend on the sensor information.

49
 Figure 4.5: USV route update

The system is sending the new information time by time and waiting for the response
from the driver, but when the USV become in the danger zone the system makes the
decision and change the USV to the autopilot case as shown in Figure 4.6 and Figure
4.7. In this initial time, the CAS works to calculate and get action to avoid crash to
the obstacle.

Figure 4.6: USV with autopilot case

50
 Figure 4.7: USV with autopilot case

After the boat is out of danger zone, CAS will be off and the USV returned on track
and return to the control of the driver of the vehicle as shown in Figure 4.8.

Figure 4.8: The USV return to the control

If the sensors have discovered any new obstacle in that way and make them as new
information in that area, the new information is stored in the database of USV and on
thecloud as shown in Figure 4.9.

51
 Figure 4.9: The USV discovered the new obstacles

The sensors checked the way by time and when they received the new information
about obstacle they send their information to the appropriate systems (Main
Controller, Main Computer and CAS) and it will work with those obstacles like
before (the system is sending the new information time by time and waiting for the
response from the driver, but when the USV become in the danger zone the system
makes the decision and change the USV to the autopilot case) as shown in Figure
4.10 and Figure 4.11.

Figure 4.10: New information about obstacles

52
 Figure 4.11: New information about obstacles

After getting out of the danger zone, the USV returned back to the focused route as
shown in Figure 4.12 and Figure 4.13.

Figure 4.12: USV returned back to the focused route

53
 Figure 4.13: USV returned back to the focused route

The operation for the simulated system can be illustrated with the following steps:

• When the USV starts working the speed is increasing from 0 to the maximum
level (35 km/h) and the system sends this message to the driver:

no alert no diversion
path clear. ignoring recommendation

• When the sensors received any signal that means there is an obstacle in the
way they will send that information to the system which will send that to the
driver:

Rec index: 3.0

turn recommended
path clear. ignoring recommendation

The first line sends the distance between the USV and the obstacle, the
second one is to give the suggestion to the driver, and the third one is
checking the USV if there is any problem.
• Also, when the sensors become closer to the obstacle the system will send
this message when the distance becomes closer:

Rec index: 2.0

54
 turn recommended
path not clear. following recommendation

• But when the USV does not receive anything from the driver and it becomes
in a danger area the USV will turn on the CAS and go to the autopilot case
and send this message:

Rec index: 0.0

alert with no alteration

• After the USV getting out from the danger zone the USV will return back to
the normal case under the driver and work as normal (send a warning
message to the driver)

Rec index: 8
turn recommended
path not clear. following recommendation

• The USV finds a new obstacle it will send a warning to the driver with the
distance:

Rec index: 6.0

turn recommended
path not clear. following recommendation
all the last steps will return back until the USV arrive at the target.
Figure 4.14 is showing the flowchart of the simulation.

55
Figure 4.14: The flowchart of the simulation

56
 CONCLUSION AND FUTURE WORK

5.1. Conclusion

Current and past studies and experiments were analyzed worldwide for the Crash
Avoidance System (CAS). Attention to this subject and try to develop this system
because of the concern to increase the efficiency and accuracy of this car. The
objective of this project is primarily to implement a prototype that detects obstacles
easily and accurately to protect the USV if there is a risk of impacting the USV path.
The following steps have been implemented:

• After the studied (A* algorithm, WSM, Kalman filter algorithm, and SLAM
algorithm) was found that the USV system works more efficiently if different
algorithms are used in different parts where they can be collected inside the
USV system for better results.
• A miniature system has been built to simulate the real system. To this end, the
performance of the weighted sum model (WSM) approach is evaluated in that
simulation as well as the algorithm is evaluated using the hardware concept.

The use of multiple sensors has proven to play a vital role in protecting the vehicle as
the algorithm was evaluated according to different values of environmental obstacles.
The simulation results show that the proposed system is able to achieve accurate
steering results, as speed and direction of movement are easily controlled while
sensors begin to detect the distance between the USV and the obstruction surface
within 25 meters and the CAS will begin to protect the USV between 5 and 2 meters
depending on the obstruction type, while the distance in some previous work is less
efficient than this or more expensive.

57
5.2. Future work

When talking about the USV, it means a lot of things to do in the future because the
subject is under development and some countries are striving to develop unmanned
vehicles. There are some points that can be added in the future to develop the vehicle
and get better results

1 - Add hardware parts to improve the performance of the vehicle such as an addition
of a high-precision camera or thermal camera that makes the work of the vehicle
more accurately. Addition of 3D LIDAR which detects the surrounding environment
with high efficiency and can be dispensed with the sensors that used before in the
case of using this type of LIDAR.

2 - Work to improve the work of algorithms by combining more than one algorithm
or reformulation of the algorithm to be more accurate to reduce the error rate if any.

3 - You can also add image processing to this project if you add the camera and using
the Kalman filter algorithm or other image processing algorithms.

58
 REFERENCES

[1]. Schofield, Rebecca T., Grant A. Wilde, and Robin R. Murphy. "Potential
Field Implementation for Move-to-Victim Behavior for a Lifeguard Assistant
Unmanned Surface Vehicle." 2018 IEEE International Symposium on Safety,
Security, and Rescue Robotics (SSRR). IEEE, 2018.
[2]. Bovcon, Borja, Janez Perš, and Matej Kristan. "Stereo obstacle detection for
unmanned surface vehicles by IMU-assisted semantic segmentation."
Robotics and Autonomous Systems 104 (2018): 1-13.
[3]. Basil, Elizabeth, and S. D. Sawant. "IoT based traffic light control system
using Raspberry Pi." 2017 International Conference on Energy,
Communication, Data Analytics and Soft Computing (ICECDS). IEEE, 2017.
[4]. Kumru, M. U. R. A. T. "Navigation and control of an unmanned sea surface
vehicle." Dept. Elec. and Electronics Eng., Middle East Technical Uni.
(2015).
[5]. Mousazadeh, Hossein, et al. "Developing a navigation, guidance and obstacle
avoidance algorithm for an Unmanned Surface Vehicle (USV) by algorithms
fusion." Ocean Engineering 159 (2018): 56-65.
[6]. Hampton, Brian, et al. "RFS-SLAM robot: An experimental platform for RFS
based occupancy-grid SLAM." Information Fusion (Fusion), 2017 20th
International Conference on. IEEE, 2017.
[7]. Pavelka, Marek, and Václav Jirovský. "Lidar based object detection near
vehicle." Smart City Symposium Prague (SCSP), 2017. IEEE, 2017.
[8]. [8] Papa, Umberto, Gennaro Ariante, and Giuseppe Del Core. "UAS Aided
Landing and Obstacle Detection Through LIDAR-Sonar data." 2018 5th
IEEE International Workshop on Metrology for AeroSpace
(MetroAeroSpace). IEEE, 2018.
[9]. Schofield, Rebecca T., Grant A. Wilde, and Robin R. Murphy. "Potential
Field Implementation for Move-to-Victim Behavior for a Lifeguard Assistant
Unmanned Surface Vehicle." 2018 IEEE International Symposium on Safety,
Security, and Rescue Robotics (SSRR). IEEE, 2018.

59
[10]. Rejab, Khaleda Sh, and Sara Mazin Naji Abd-Al Hussain. "Design and
Implementation of Intelligent Mobile Robot based on Microcontroller by
using Three Ultrasonic Sensors." (2017).
[11]. Catapang, Angelo Nikko, and Manuel Ramos. "Obstacle detection using a 2D
LIDAR system for an Autonomous Vehicle." 2016 6th IEEE International
Conference on Control System, Computing and Engineering (ICCSCE).
IEEE, 2016.
[12]. http://www.slamtec.com/en/lidar/A3 [ Accessed 15/03/2019 ] (10:30am).
[13]. Bertram, Volker. "Unmanned surface vehicles-a survey." Skibsteknisk
Selskab, Copenhagen, Denmark 1 (2008): 1-14.
[14]. Song, Rui. Path Planning and Collision Avoidance of Unmanned Surface
Vehicles in the Marine Environment. Diss. UCL (University College
London), 2018.
[15]. [15] Liu, Zhixiang, et al. "Unmanned surface vehicles: An overview of
developments and challenges." Annual Reviews in Control 41 (2016): 71-93.
[16]. Buis and Jan Maarten. “Obstacle and Cliff Detection For Robotics
Applications Using Miniaturized Sonar and IR Distance Triangulation
ELECTRICAL ENGINEERING PROGRAMME”2017.
[17]. Gibbs, Gerard, Huamin Jia, and Irfan Madani. "Obstacle Detection with
Ultrasonic Sensors and Signal Analysis Metrics." Transportation Research
Procedia 28 (2017): 173-182.
[18]. Adeyemi, Gbenga Joshua. Development of an External Ultrasonic Sensor
Technique to Measure Interface Conditions in Metal Rolling. Diss. University
of Sheffield, 2017.
[19]. Palma, Luis Puebla, and R. T. A. C. Americas. "Ultrasonic Distance
Measurer." Freescale Semiconductor, application note (2008).
[20]. https://www.3dlasermapping.com/what-is-lidar-and-how-does-it-work/
[Accessed 19/05/2019 ] (5:30pm).
[21]. https://cdn-shop.adafruit.com/product-files/3978/3978_manual_SJ-PM-
TFmini-T-01_A03ProductManual_EN.pdf [Accessed 17/06/2019 ] (4:30pm).
[22]. file:///G:/rakun%20master%20thesis/thesis%20word&pdf/New%20folder/usv
/datasheet-rplidar.pdf [Accessed 20/06/2019 ] (2:20pm).
[23]. https://www.elprocus.com/different-types-of-microcontroller-boards/
[Accessed 24/07/2019 ] (7:30pm).

60
[24]. https://www.arduino.cc/en/guide/introduction [Accessed 20/07/2019 ]
(3:20pm).
[25]. https://www.amazon.co.uk/SunFounder-ATmega2560-16AU-Board-
compatible-Arduino/dp/B00D9NA4CY [Accessed 25/07/2019 ] (9:20pm).
[26]. https://www.lattepanda.com/blog-2446.html [Accessed 28/07/2019 ]
(9:30pm).
[27]. https://www.electrical4u.com/electrical-motor-types-classification-and-
history-of-motor/ [Accessed 22/08/2019 ] (5:10pm).
[28]. Eom, Hong Duck, and Jae Wook Jeon. "Environment map building using
low-cost IR sensors and a servo motor for mobile robot." The 18th IEEE
International Symposium on Consumer Electronics (ISCE 2014). IEEE, 2014.
[29]. Triantaphyllou, Evangelos. "Multi-criteria decision making methods." Multi-
criteria decision making methods: A comparative study. Springer, Boston,
MA, 2000. 5-21.
[30]. Moestopo, Universitas Prof Dr, and S. T. I. K. O. M.
TunasBangsaPematangsiantar. "Comparison of Weighted Sum Model and
Multi Attribute Decision Making Weighted Product Methods in Selecting the
Best Elementary School in Indonesia." International Journal of Software
Engineering and Its Applications 11.4 (2017): 69-90.
[31]. Helff, Florian, Le Gruenwald, and Laurent d'Orazio. "Weighted Sum Model
for Multi-Objective Query Optimization for Mobile-Cloud Database
Environments." EDBT/ICDT Workshops. 2016.
[32]. Kilinçarslan, Mustafa. Implementation of a path finding algorithm for the
navigation of visually impaired people. Diss. MS thesis in Computer
Engineering, Atilim University, Ankara, 2007.
[33]. Abu Hasim, Ahmad Shukri (2016) Shunt Active Power Filter Employing
Kalman Filter Estimator For Harmonics Reduction. PhD thesis, Universiti
Teknikal Malaysia Melaka.
[34]. Dissanayake, MWM Gamini, et al. "An experimental and theoretical
investigation into simultaneous localisation and map building." Experimental
robotics VI. Springer, London, 2000. 265-274.
[35]. Abbas, Eyad I., Khaleda Sh Rijab, and Anas Fuad Ahmed. "Optimal Wavelet
Filter for De-noising Surface Electromyographic Signal Captured From

61
 Biceps Brachii Muscle." Engineering and Technology Journal 33.1 Part (A)
Engineering (2015): 198-207.
[36]. Al-Tekreeti M., Tarıyan Özyer S., Özdemir C. and Karadağ A., Al-Dakheel
S., “Unmanned Surface Vehicle with Obstacle Avoidance System”,
International Conference on Big data, IoT, and Cloud Computing (ICBICC),
2019, in press.
[37]. Al-Dakheel S., Tarıyan Özyer S., Özdemir C. and Karadağ A., Al-Tekreeti
M., “Unmanned Surface Vehicle with Drown Map System”, International
Symposium on Multidisciplinary Studies and Innovative Technologies
(ISMSIT), 2019, in press.

62