JAC : A Journal Of Composition Theory ISSN : 0731-6755

PATH ROUTING ON A MAP AND SIMULATION USING A STAR

ALGORITHM
Gannavarapu Datta Lohith1, C N Sujatha2, Challa Pamuleti3 , Karimolla Mahesh4 and Samala Sai
Charan5
1-5
ECE, Sreenidhi Institute of Science and Technology, Ghatkesar, Hyderabad, Telangana

ABSTRACT: Human life these days has become easier with new technologies developing with respect to our
needs. Path generating techniques is one such technology, there is a significant increase in its demand for the
last few years. In this area, the most accurate and faster search algorithm is A*(A STAR) algorithm. In this
paper we discuss about A*(A STAR) algorithm, it is a path searching technique, which is an update of Dijkstra's
algorithm to find the shortest distance between two given locations/points (starting and target). A*(A STAR)
algorithm main aim is to improve the running time using heuristics. It uses the fact that all nodes generated
from spatial domains have related coordinates, these coordinates help in finding how far the target node is from
the present node by using a distance formula. Up on finding the target node exact position at regular intervals
will make A*(A STAR) work much faster and accurately in finding paths when compared to Dijkstra's
algorithm. In this paper we will generate the shortest path between the starting position and the target position
on a given image input. Initially our input image is converted into a grayscale image using image processing
technique through MATLAB. This helps in splitting the available traversing pixels and pixels with obstacles on
the image. Upon assigning starting position and target position on non-obstacle pixels, the A*(A star) algorithm
will generate the shortest path along the available traversing pixels using distance formula. This distance
formula is also known as distance function. In computer science, A* (as “A star”) is a computer algorithm that
is widely used in pathfinding and graph traversal, the process of plotting an efficiently traversable path between
points, called nodes. The A*(A STAR) algorithm combines features of uniform-cost search and pure heuristic
search to effectively compute optimal solutions. Noted for its performance and accuracy, it enjoys widespread
use. Peter Hart, Nils Nilsson and Bertram Raphael first described the algorithm in 1968. It is an extension of
Edsger Dijkstra’s 1959 algorithm. A*(A STAR) achieves better performance (with respect to time) by using
heuristics. As A*(A STAR) traverses the graph, it follows a path of the lowest known cost, Keeping a sorted
priority queue of alternate path segments along the way. If, at any point, a segment of the path being traversed
has a higher cost than another encountered path segment, it abandons the higher-cost path segment and
traverses the lower-cost path segment instead. This process continues the goal is reached.
Keywords: A*(A STAR) algorithm, Grayscale conversion, Nodes, Heuristics, Simulation, Path routing.

I. INTRODUCTION
There was a problem of finding the shortest path between two locations in real-time. There is a need to travel for
every person and as time is limited to everyone, we need to travel with minimal time possible. This intensified
when there were more and more transportation and more use of vehicles.Then came path finding algorithms, to
find the shortest distance between the nodes. Google, one of the top company in the world is the main provider
of maps which help us travel in daily life for everyone. The efficiency of the algorithms is very high, that’s what
makes it readily usable by everyone. In google maps, lot of path finding algorithms are used. Some of the path
finding algorithms are Dijkstra algorithm, Grid algorithm, A* algorithm. Google Maps uses A* algorithm for

Volume XIV, Issue VIII, AUGUST 2021 Page No: 293

JAC : A Journal Of Composition Theory ISSN : 0731-6755

finding the shortest path and alternating routes in real time. A* algorithm is an advanced form of Breadth first
search. It avoids costly paths and chooses the most promising path. It is a very smart algorithm. There are a lot
more constraints present in the google maps. The types of constraints are like time constraint, distance
constrained vehicle type (it shows route based on the type of vehicle like 2-wheeler or 4-wheeler). Google maps
is very better for the human needs. The way google maps routes the map is user friendly for a daily routine user.
Based on the distance and time constraint it shows the shortest and easy path. Google maps also takes calculate
the traffic present in the route. If there is more traffic, then it enroots the map. Finally, it makes the user to travel
easily. It shows the exact directions of the location, and it also gives the voice assistance for the route based on
the directions. But the path routing which we are doing in our project does not evaluate all those constraints. It
mainly focuses on the shortest path finding by distance constraint. Time constraint is not applicable. This path
routing algorithm is helpful for the smart needs in the home e.g., robots for working (this algorithm shows the
path that is needed for the robot to travel). we need to convert all the obstacles present in the home into a
grayscale format. The obstacles are captured and converted into the grayscale format. By the calculation of the
obstacles and many other we can route a shortest path for the robot. There are many applications of our project
in the daily life. Even the project is used for military applications also. A map is taken from the desired initial
position to the final position.
A*(A STAR) algorithm was first proposed by Peter Hart, Nils Nilsson and Bertram Raphael in the year 1968 at
Stanford Research Institute. Even though it is an extension of Dijkstra's algorithm it does not change the
complete structure. It also generates the two sets as Dijkstra that is a frontier and an explored set and uses these
for search on graphs. In every loop it moves a node from the frontier set to the explored set. The main difference
is A*(A star) moves the nodes with the lowest F value, whereas Dijkstra moves the node with lowest G value.
H: shortest path distance to the goal node, G: shortest path distance to the starting node and F: addition of both
H and G. The F values help us to identify the ranking order of the nodes in the frontier set. Based on the F value,
we can understand the cost of the path. This indicates that nodes having least of all F value is one which leads us
directly to the target node. To have consistency it should make sure that, if an edge is present between two of the
nodes. The difference in values of H values between those nodes are maintained such that they never exceed the
edge weight of the nodes. There are few terminologies to be known about A*(A STAR) algorithm to understand
how it works.
Node: All potential position or stops with a unique identification
Transition: The process of moving in between the nodes.
Starting Node: Initial point to start search
Goal Node: Destination/target node where search ends.
Search Space: A collection of nodes, similar to all the elements involved in a board game
Cost: Distance between a node and another node along the search path.
g(v): Represents the Exact cost between Starting node and any node v along the searched path.
h(v): Represents the Estimated heuristic cost between a node to the target node.
f(v): least cost in the adjacent node.

Each time A*(A STAR) enters a node, calculates effectively the cost of the path or route, f(n)(n being the
neighbouring node, to traverse to each and every adjacently placed nodes, and then enters the node which has

Volume XIV, Issue VIII, AUGUST 2021 Page No: 294

JAC : A Journal Of Composition Theory ISSN : 0731-6755

the least value of f(n).

f(v)=h(v)+g(v)
Here h(v) is known as heuristic distance. A* Heuristic - Euclidean Distance is easy to calculate and formulate
the path. The Euclidean distance is the absolute distance between two nodes. It is the distance between two
points in free space. This heuristic is chosen because the map with obstacles will generate a long path. The
diagonal distance is one of the shortest distances. The Euclidean distance and the diagonal distance give us top-
quality paths. Minimum number of nodes are explored whilst the usage of the circulate distance, because it
comes toward efficiently estimating the best course duration to the target node. Based on those concerns we've
got selected to stay with the diagonal distance whilst A*(A STAR) is used in the grid global. This is the
heuristic function which is used in this system.
II. LITERATURE REVIEW
Leigh R, Zheng T: Path planning has become a prominent technique in the navigation field. It is employed in
most of the games, mobile robots and many more. The accuracy of the algorithms is a major factor, the best
result algorithm is always employed in the real world apart from others. One such algorithm for path planning is
A* (* spelled as star) algorithm. Mobile robots, uses A* algorithm, are the best example for path planning
applications which travel in the shortest path towards the target location. A* algorithm uses heuristic values this
makes the algorithm most effective, accurate and unique from all other algorithms. Mobile robots have the
analytical ability to generate paths in a real-world environment that allows the robots to reach their destination
without any ambiguity and collision. Robots use trajectory planning for navigation. This decides the movement
of the robot in generating the path to reach the destination while at rest and even in the motion. While travelling
it might come across many obstacles so it has to clearly identify, neglect the obstacles and navigate accordingly
towards the target without collision. Collision is the major aspect that has to be looked after in case of mobile
robots for its safe navigation. The safety of a mobile robot depends upon how much clearance area has been
found when it moves from the trajectory of a sharp or closer obstacle. To start over Richardo acatillo has
proposed a method to overcome the collision while travelling, using different terrain and obstacle encounters he
designed a modular robotic system unfortunately it doesn't work practically [1].

Further Jong-hyun Park and UK-Youl Huh proposed SGPP(save global path planning) combination of classical
A* with PRM (probability roadmap) algorithms. This helps to overcome the collisions at the corners and
turnings (even after use of sensors) by determining the hazardous points, risk distance and velocity control. Path
is generated using SGPP along with A* algorithm with minimum cost.They explained with a simulation and the
SGPP method showed superior performance to the other considered algorithms in terms of the safety cost and
navigation time[2]. Further Borenstein and Koren came up with a new collision free technique VFH method, In
which a two-dimensional Cartesian Histogram Grid is updated continuously and in real-time with range data
sampled by the onboard range sensors. The data in the Histogram Grid is reduced to a one-dimensional Polar
Histogram that is constructed around the robot's momentary location. This Polar Histogram represents the polar
obstacle density around the robot A sector with low obstacle density which is close to the target direction is
selected, and the robot's steering is aligned with that direction. They implemented and tested the VFH method
on a commercially available mobile platform (Cybermation, 1987), with best result [3]. Next T. Arai & J. Ota
proposed KRMP (Kinodynamic Randomized Motion Planning): They introduced the concept of motion

Volume XIV, Issue VIII, AUGUST 2021 Page No: 295

JAC : A Journal Of Composition Theory ISSN : 0731-6755

planning within dynamic robot networks that coordinate centralized planning. Each robot will have their own
network, whenever two or more robots enter each-others network they establish links among themselves (LAN)
to communicate with others. Each generates a new centralized paths and share among themselves among them
the best paths are opted and conformed. Then they start motion. It planned with a high probability of success,
even in cluttered environments involving robots, stationary obstacles and moving obstacles. Planning times of
less than 100 ms allowed the robots to re-plan in real-time and react quickly to changes in the environment[4].
Hasan Mujtaba & Dr. Gajendra Singh proposed a new technique combination of classical A* and virtual
objects have given satisfying results. In this method they employed a backtracking method from goal node to
starting node and by putting virtual stationary objects in the environment on a grid map. Simultaneously from
front and back helped the robot to travel along the collision free path. Size of the virtual objects plays a vital
role, if the size increases 98% of the collisions with the obstacles as well as with the other robots can be
overcome[5]. German A.Vargas et.al has proposed A* path planning algorithm using webot. The paper defines
initial or goal positions and obstacle avoidance using simulation environment and diagonal movement of a
mobile robot on cell decomposition platform [6]. Yudi Pratama and Wasolim have developed a mobile robot for
logistic transport of material in indoor. Wall following algorithm for a mobile robot has been used for the
transportation of material using the corner or sides of a room in navigation, the combination of sensor on robot
give the exact location of initial or target points. But does not satisfy the obstacle avoidance methods of a
mobile robot [7]. Jur van den Berg, Pieter Abbeel and Ken Goldberg wanted to solve the problem of path
planning in robots. They developed a technique LQG-MP(linear-quadratic Gaussian motion planning) which
characterizes in advance planning, the priori probability states of the robot on the path it is about to travel. The
results are reported using random trees(RRT) in the work done in this paper [8].
Shaung Liu, Dong sun and Changan Zhu saw the possibility of making multiple robots move in coordination
with each other along designed paths for them. While moving, they need to satisfy the formation requirement to
avoid collisions and meet parameters in movement such as velocity, acceleration. To achieve this they have used
linear interactive and general optimizer. The results are verified in simulation and experiments of group of
robots [9]. Hajduk and Marek Sukop presented some points on multiple robotic systems mainly focused on the
point of view of control. In such systems, they presented about a dynamic exchange of action of robots and their
strategic positions. This is purely implemented for the purpose of robotic soccer. The results for dynamic change
of agents while playing was brought out by software agents exchange control of robots [10].

From the literature survey, by observing all the problems and respective approaches to find a solution, showed
us to locate the problem first then to find a perfect solution for it. The main problem is finding the obstacles, as
we see to understand the working of the A*(A STAR), the algorithm is test run on simple 2D matrices of
different sizes and the results are observed. The inputs are the obstacles, vehicle position, target position. The
maximum time consumed is due to the time taken to give inputs to the program graphically with minimal human
error. Human error constitutes of errors such as wasting the time, unintentionally or intentionally, or delay in
understanding the instructions or delay in giving inputs to the code perfectly. This process is not so effective
when considered for bigger matrices or bigger problems.
To overcome this problem, there is an automatic obstacle detection equipped in the code, then and only then the
obstacles are found at faster rates that is at the rate of CPU. The CPU processing rate is must faster than the rate

Volume XIV, Issue VIII, AUGUST 2021 Page No: 296

JAC : A Journal Of Composition Theory ISSN : 0731-6755

at which humans give obstacles as inputs. The need for simulation of the path carried by the vehicle in the map
given as input can give a clear understanding of the implementation of A*(A STAR) algorithm in real life maps.
The inputs are real life maps and with the simulation, greater and precise results are achieved.

III. PRESENT WORK

Figure 1: Block diagram

In the block diagram shown in figure 1, the program is written in matlab which fully implements a star
algorithm to find the optimal path. The image of the map is converted to grayscale format to detect obstacles
and understand the available space for the path planning. The locations are selected on the map graphically (the
nodes are selected). The path is formulated by calculating the shortest distance from the starting node to the
ending node. Output path is generated and the data is sent to unity. The whole program is simulated in unity 3d
by reading the data received from MATLAB.
MATLAB software is used for processing and formation of the path on the map using A*(A STAR) algorithm.
The MATLAB 2018b is used for execution of the code in the project. There are other versions as well, but
2018b offers stability and a chance to implement Simulink if needed. The software is used to manipulate the
map image and then use it with A*(A STAR) algorithm to find the shortest path between two locations. This
generates various variables and visually represented graphics which are quite helpful for the output of the
proposed system. When a map image is taken, it is of the form of RGB image. Detection of obstacles is slightly
difficult, so we convert the image to grayscale. When you open google maps, the maps shows different places
with names on it. These are called labels. These labels if contained in the image may arise logical errors of the
working of the project, so we need an image of the map without labels. This is possible from special websites
such as snazzymaps.com. From this website, we can get image of the maps without labels. As the input is in the
image format it is a RGB image. A*(A STAR) algorithm directly does not work on this RGB image, it would be
complex to find a path on it. So, to make our work simple we use an image processing technique i.e, Gray scale
conversion. This helps us to find a path on the image. Grayscale images are much easier to work within a variety
of tasks like in many morphological operation and image segmentation problems, it is easier to work with a
single layered image (Grayscale image) than a three-layered image (RGB colour image). It is also easier to
distinguish features of an image when we deal with a single layered image.
To convert RGB to grey there is a function called rgb2gray() in MATLAB. The three colours on a pixel are now
turns into a single value colour by calculating the weighted addition of all the three colour components. In figure
2, we can see that the conversion of the map image to grayscale, which is used to find the obstacles in the map.

Volume XIV, Issue VIII, AUGUST 2021 Page No: 297

JAC : A Journal Of Composition Theory ISSN : 0731-6755

Figure 2:: Illustration of Gray Scale conversion and imtool function

In figure 3, the left image is the Map1.j
Map1.jpg
pg downloaded from snazzy maps website. The image is RGB image.
The right image is the imtool window of the converted grayscale image. This is done by the utilization of a
special function called Pixel_finder.m. The pixel value of the pixel on the map is kknown. The pixel value of the
pixel at (8,1) is 236. The value 236 is the intensity of the pixel in the range of [0
[0-255].
255]. This is crucial as to find
out the pixel value of the pixels which represent as obstacles on the map. Now, based on the pixel values, the
t
obstacles are found for a condition l<240 and l!=238, where I denotes the Intensity of the pixel.

Figure 3: Use of Pixel_Finder.m function

Unity is a 3D environment development software which works with the help of unreal engine to understand and
work with the physics of the materials. In unity, we can actually create a 3D visualized environment fair
understanding of the moment of Rachel when a map is given as an input under starting and ending locations ask
also given. The first step in understanding unity is that we need to understand the view of the environment. The
view is a 3D perspective view, we can change it to orthogonal view or else 2D view based on the requirement.
There are basically two main important windows viz. Scene and game. Unity software is used for simulation
purpose as it is very easy to use and understand. The connection between MATLAB and unity is not through
servers but through files. The environment is setup for the simulation part of the project. The scene window is
where we do the working, the modification of environment and the establishment of the working space. The
game is simply the output window where we can see the execution off the scene which we have created earlier.
A 3D object is created in the scene name as simulation as shown in figure 4. We name each and every game
object in the scene, so that we can easily identify what game object we are using are modifying.
modifying. The hierarchy
in the scene is always from the top to down. In the scene “Simulation”, the main game object of our project is
the “Vehicle”. It is created by creating 3D object of the shape of the capsule for easy understanding. The main
camera and thee secondary camera of the environment are converted into the Childs of the game object

Volume XIV, Issue VIII, AUGUST 2021 Page No: 298

JAC : A Journal Of Composition Theory ISSN : 0731-6755

“Vehicle.”

Figure 4: Scene 'Simulation' and its Game Objects

In the figure 5, we can see that there are two cameras displaying over a single screen, the secondary camera is
displaying on the left side, the primary camera is displaying over the right side. The primary camera’s vision is
of TPP, so the vehicle is looking closely. The secondary camera is far from the vehicle and is of the orthogonal
perspective. It appears so small from the point of view of the secondary camera. The whole environment is
empty except for the capsule, as the other objects are created dynamically after the optimal path is found.

Figure 5: Camera views of the primary and secondary in game view

Visual studio 2019 is used for coding the scripts in Unity. The unity is controlled by c# scripts. These c# scripts
are edited, debugged and are written in Visual Studio 209 editor. The main use of this editor is because it
supports IntelliJ. IntelliJ helps code faster and easier. It gives suggestions of the code snippets while writing, so
even a rookie can understand the code well and write according to the needs. A Script named Gridmaker.cs is
responsible for the creation of the 3D map in unity. The script reads the data from the file and creates obstacle
object and white space object. These objects are made and stored as prefabs for easy accessing. These objects
are very simple cubes with black colour for obstacle and white for available space. The Map is created and
based on the optimal path locations, the vehicle is moved, and it can be visualized in the simulation. These
scripts are created and attached to the game objects in the scene, so that they are interactable. The attachment of
the scripts to game objects is very crucial for proper execution of the simulation. Gridmaker script is attached to
Grid game object and Movement script is attached to Vehicle game object.

Volume XIV, Issue VIII, AUGUST 2021 Page No: 299

JAC : A Journal Of Composition Theory ISSN : 0731-6755

IV. ALGORITHM OF THE PROJECT WITH FLOWCHART

In figure 6, the flow of the project is depicted in the form of flow chart. The flow chart is divided into two
blocks, MATLAB and UNITY AND VISUAL STUDIO. The whole functionality resides inside these two
blocks. Following steps are followed in the MATLAB block
1. Initialization of the Map image – The map image is taken from online and stored it in the project
folder. The map is given as input.
2. Detection of obstacles – The map image is converted to grayscale and then the obstacles are found
based on the intensity of the pixel values( each pixel considered as a node). The values are updated to
the MAP array.
3. Inputs from the user – Manual input of the vehicle initial position and the final target position is given.
4. Formation of the shortest path – The A*(A STAR) algorithm runs based on these inputs and finds the
optimal path between the two nodes( initial and target). The path is drawn graphically.
5. Storing the Data – The MAP array is stored as a .csv file in the project folder. The file contains the
values of the map depicting the obstacles, whitespace, initial position and final position.

Figure 6:Flow Chart of the project

Following are the steps in the UNITY AND VISUAL STUDIO block:
1. Get data from MATLAB – The .csv file is accessed from the project folder by the Gridmaker.cs script
in unity when the code is run.
2. Process data and create 3D environment – The .csv file contains the MAP array, based on the values of

Volume XIV, Issue VIII, AUGUST 2021 Page No: 300

JAC : A Journal Of Composition Theory ISSN : 0731-6755

the elements in this array, the obstacles and

a whitespace is created in a 3D environment.
3. Simulate the vehicle in the 3D environment – The vehicle is moved according to the shortest path in
unity with the help of Movement.cs script.
V. EXPERIMENTAL RESULTS AND CONCLUSION
After the successful execution of A star algorithm in MATLAB, It has been found the results of the working
of the algorithm as shown in figure 7 and 8.
8. The functioning of the A star algorithm is that it finds the shortest
distance path from two nodes, which we give as an input during th
thee execution of the program. In this
proposed system, instead of simple grids of fewer number of nodes, real images of maps are taken into
consideration for the execution. The map images taken from online, are RGB images and after processing the
image, the obstacles are automatically detected in the code itself instead of manual inputs of the obstacles.
The obstacles are detected and stored accordingly in a MAP array of a large size equal to the size of the Map
image. The time taken to give the input of the obstacles has reduced drastically as the obstacles are detected
automatically.

Figure 7:Output
:Output after execution of the code written in MATLAB

Figure 8:: Output after execution of the code written in MATLAB with different inputs

After the execution in MATLAB, the data is sent to unity. The data received is used for developing the
environment. The speed and the location of the vehicle are controlled from the code itself. The speed can be
altered during the execution of the scene, i.e in the game window. The secondary camera facing left depicts

Volume XIV, Issue VIII, AUGUST 2021 Page No: 301

JAC : A Journal Of Composition Theory ISSN : 0731-6755

the visuals of the shape of map provided as input. T

The
he movement of the vehicle can be visualized
orthogonally. The primary camera on the other hand, clearly shows how the obstacles are avoided to reach the
final destination from the starting point of travel. The figure 10 depicts that the map is generated ffrom the
MAP variable after the execution of program in MATLAB. The screenshot is taken in an orthogonal view of
the map in scene window.

Figure 9:: Simulation result of the above map in Unity in Game window with two perspectives

Figure 10:: Simulation result of the above map in Unity in Scene Window

VI. CONCLUSION
We can conclude from the above observations and results, which are graphically represented, which are
collected from MATLAB by the execution of the program. It is clear that the A star algorithm works perfectly
for grid maps which are of considerable sizes, that is the total number of nodes are visually understandable to
the user to be able to give the inputs graphically to form a path which is optimal.
optimal. The case where the number
of nodes in the matrix is far greater, manual effort increases. We can observe from the tables in the results of
the previous system that the time taken for the total execution increases with increasing number of obstacles.
The maximum time consumed is due to the time taken to give inputs to the program graphically with minimal
human error. Human error constitutes of errors such as wasting the time, unintentionally or intentionally, or
delay in understanding the instructions or delay in giving inputs to the code perfectly. These kinds of human
errors are terminated, and the time taken to give the inputs is considered which is the total time taken to

Volume XIV, Issue VIII, AUGUST 2021 Page No: 302

JAC : A Journal Of Composition Theory ISSN : 0731-6755

complete the process of finding the optimal path. It is observed that this time increases with increasing
number of obstacles and with increase in number of nodes in the map and only due to manual input. The
manual input from the mouse or joystick or keyboard is completely dependent on the person handling the
program, if human error is to be considered then the code takes a long time to complete. If the human makes a
mistake in giving an obstacle at a wrong position, the code must be run again. This wastes a lot of time. This
process is not so effective when considered for bigger matrices or bigger problems. As there is an automatic
obstacle detection equipped in the code, then and only then the obstacles are found at faster rates that is at the
rate of CPU. The CPU processing rate is must faster than the rate at which humans give obstacles as inputs.
This problem is overthrown in the proposed system and many other functionalities are added which make the
new system ready for implementation in real life projects such as Automatic Vehicle for different purposes.
The Simulation connection between MATLAB and unity is remarkably simple and not involving any servers
or tunnels or any such complications. The execution would be sightly smoother, but the output generated is
the same and the use of csv files helps us to see the values for ourselves and understand the system
extensively. In automation of vehicles or small robots in local areas, this proposed system can be used
accordingly, as it is easy to understand and easy to modify. The main advantage of the system is that even a
newbie can understand the working and they can modify according to their needs. For instance, a cleaning bot
in one’s room. The bot needs the map data of the room and the data about the obstacles to clean the room
precisely. This can be easily achieved by testing the bot through simulation after the execution of the program
in MATLAB. Furthermore, many automatic vehicles can be created for special purposes in warehouses or
such. There are many applications for the proposed system as the simulation part can detect many problems
before making a live prototype of the project.

VI. REFERENCES
[1] Richardo acatillo et.al “Implementation and assembly of a robotic module for mecabot-3 reconfigurable
system” International Journal of Applied Engineering Research 2016 vol-11 no.21 pp 10681- 10684
[2] Jong-Hun Park and Uk-Youl Huh “Path Planning for autonomous mobile robot based on safe space”
Journal Electrical Engineering Technology.2016; vol .11 pp1921-718
[3] Bornstein, J. and Y. Koren, Real-time obstacle avoidance for fast mobile robots. Systems, Man and
Cybernetics, IEEE Transactions on,1989. 19(5): p.p 1179-1187.
[4] P. E. Hart, N. J. Nilsson, and B. Raphael, “A formal basis for the heuristic determination of minimum cost
paths in graphs”, IEEE Transactions of Systems Science and Cybernetics. vol.4, no.2, pp.100-107, 1968
[5] German A. Vargas et.al “Simulation of e-puck path planning of Webot” International Journal of Applied
Engineering Research 2016 vol-11 no.19 pp 9772- 9775
[6] Yudi pratama & Wansu lim “Development application of indoor logistics transportation using amr”
International Journal of Applied Engineering research 2016 vol-11 no.20 pp 10183-10189
[7] LaValle, S. M., 2006. Planning Algorithms, Cambridge University Press, New York, USA. František
Duchoň et al. / Procedia Engineering 96 (2014) 59 – 69 69
[8] Van Den Berg, J., Abbeel, P., Goldberg, K., 2011. LQG-MP: Optimized path planning for robots with
motion uncertainty and imperfect state information. The International Journal of Robotics Research, 30(7), p.
895-913.
[9] Liu, S., Sun, D., Zhu, C, 2011. Coordinated motion planning for multiple mobile robots along designed

Volume XIV, Issue VIII, AUGUST 2021 Page No: 303

JAC : A Journal Of Composition Theory ISSN : 0731-6755

paths with formation requirement. Mechatronics, IEEE/ASME Transactions on, 16(6), p. 1021-1031.
[10] Hajduk, M., Sukop, M., 2009. Multiagents system with dynamic box change for Microsot. Progress in
Robotics: Communications in Computer and Information Science 44, p. 287-292.

Volume XIV, Issue VIII, AUGUST 2021 Page No: 304