Multiple Festo Robotino Navigation using Gazebo-ROS Simulator

I Rokhim, P. Anggraeni and R. H. Alvinda

Automation Engineering Technology, Bandung Polytechnic for Manufacturing, Jln. Kanayakan 21, Bandung, Indonesia

Keywords: Autonomous Robot, Robotino, Robot Operating System (ROS), Gazebo.

Abstract: Two robotino are implemented on this study, the idea is to control two robotino as an Autonomous Robot to

achieve various tasks by using Robot Operating System (ROS) as their platform. Autonomous robots are

widely used in the unstructured environment and a robust system required to safely and efficiently achieve

the various tasks such as logistical task. By having a reliable system, many industries can be benefited by

using autonomous robot especially for open source system. The cost of building the robot itself has become

a burden, by using an open source operating system, maintenance cost can be reduced. The robots are using

motion planning to achieve their goal position using bug algorithm. This system is implemented on Gazebo

simulation workspace and can be developed further to the real robotino or other omnidirectional robot by

providing different namespace on ROS topics. The robots are able to be controlled by a ground station such

as Personal Computer or a Laptop through wireless communication. The result of this study shows

effectiveness of the algorithm that are being used by both robots in a certain environment.

1 INTRODUCTION

Autonomous robot type in this experiment is a land-

based mobile robot that able to traverse the

environment to perform various tasks such as

logistical task, exploration, etc (Ali and Ali, 2015);

(Anggraeni, Mrabet, Defoort, and Djemai, 2018);

(Borenstein and Koren, 1991). Autonomous robots

are widely used on large warehouses in industry to do

a repetitive delivery task that can help human job

easier by covering a large area of warehouse

(Draganjac, Miklić, Kovačić, Vasiljević, and Bogdan,

2016). In the recent years, the usage of autonomous

robots are becoming more often than past years and

many researchers have been studying the autonomous

robot (Klancar, Zdesar, Blazic, and Skrjanc, 2017) to

improve its autonomous behaviour (Sabattini,

Cardarelli, Digani, Secchi, and Fantuzzi, 2016),

safety, and reliability (Ali and Ali, 2015);

(Mylvaganam, Sassano and Astolfi, 2017).

The autonomous robot can be used to perform

various tasks such as material distribution or delivery

on a certain facility (Oltean, Dulău, and Puskas,

2010). The robot will travel the facility or

environment by using a predefined path to complete

their task without any direct control or direct

supervision from an operator on site (Yan, Jackson,

and Dunnett, 2017). As the warehouse grew larger,

more autonomous robots are needed to cover a large

area. This could be a problem because as the number

of robots added to an environment without a proper

system can lead to disastrous management of

materials and the company will have suffered a

financial loss (Borenstein and Koren, 1991).

One of autonomous robot back bones is obstacle

avoidance (Borenstein and Koren, 1991);

(Mylvaganam, Sassano and Astolfi, 2017). This

ability is one of the main reason such vehicle can be

operated on an environment without human

interruption, however for multiple robots the ability to

avoid obstacle becomes a challenge. This challenge is

going to be studied and experimented by using a

simple algorithm to aim the two robots finish their

tasks safely without any losses.

This study focuses on using bug algorithm for

robots to travel from a starting point to another. By

using a simple algorithm, the cost of computational

task can be reduced and this algorithm also can be

used elsewhere to do additional task for robots. The

environment area of 8x9 meters room will be filled

by some obstacles.

The remaining part of this paper will be organized

as follows. The section 2 will discuss kinematics

model of the robots and sensors that are attached on

the robots, and system design, section 3 deal with the

Robot Operating System (ROS) simulation using

Rokhim, I., Anggraeni, P. and Alvinda, R.

Multiple Festo Robotino Navigation using Gazebo-ROS Simulator.

DOI: 10.5220/0010967800003260

In Proceedings of the 4th International Conference on Applied Science and Technology on Engineering Science (iCAST-ES 2021), pages 1491-1495

ISBN: 978-989-758-615-6; ISSN: 2975-8246

 2023 by SCITEPRESS – Science and Technology Publications, Lda. Under CC license (CC BY-NC-ND 4.0)

1491

Gazebo and the algorithm of the motion planning.

The result of the experiment is shown on section 4 and

last section will discuss the conclusion of the

experiment.

2 SYSTEM DESIGN

This section presents the design of the robot, its

component and kinematics model of the robot.

2.1 Festo Robotino

Robotino Festo is one of the holonomics robots that

acts as autonomous robot. The robot has three degree

of freedom which can be controlled by the inputs of

translational velocity and angular velocity. The robots

are driven by DC motor which are equipped with

encoder. Robotino has a maximum velocity of

10km/h.

The Operating System of the Robotino consists of

two layers, the first layer is Linux and the second

layer is Real Time Linux. Linux layer provides a

standard use of Robotino through direct control on the

machine while the Real Time Linux allow usage of

wireless communication control of the robot.

2.2 Kinematics of Robotino

Inverse kinematics on global coordinate are obtain by

calculating the translational velocity.

The final sentence of a caption must end with a

period.

Figure 1: Robotino on its own coordinate.

𝑣=



𝑥



+𝑦



)

And angular velocity 𝜑 The velocity from each

wheel, in this case wheel 1 are defined by this

equation below:

𝑣



=𝑣



+𝑣



)

The wheel velocity is part of the translational velocity

of the robot that are defined by this equation below:

𝑣



=𝑥 sin



𝜑



+𝑦  cos



𝜑



)

nd an

ular velocit

defined as follow:

𝑣



=𝑅𝜑

)

Distribute equation (3) and (4) to equation (2), then

we get the complete equation of the velocity of wheel

𝑣



=𝑥 sin



𝜑



+𝑦 cos



𝜑



+𝑅𝜑

(5)

By using same method, we can calculate the velocity

of wheel 2 and 3 by adding the respective angle for

each wheel ϕ and other configuration of each wheel

respective to the global angle on (x,y) (0,0) of the

robots that are located in the center of the intersection

of robot wheel coordinate.

2.3 Robot Operating System

As stated in the previous sub section, Robotino Festo

were operated by embedded Linux Operation System

in a minicomputer. The linux version in the Robotino

Festo is Ubuntu Xenial. This Ubuntu Xenial has a

middleware called Robot Operating System (ROS)

that allows the robot to be controlled via Nodes that

are available in ROS packages. This Nodes can

publish or subscribe messages to topics. The topic

controls the actuators and sensors on the robot

directly by sending messages from Nodes. ROS

programming using both C++ and python.

Figure 2: General System for multi robotino control.

Robotino on the simulator are controlled by

sending a message from a control program to topics

of the actuator. The topics are in form of the

translational and angular velocity with its own

namespace for each robotino. The positions of the

robotino are known through odometry topics that are

iCAST-ES 2021 - International Conference on Applied Science and Technology on Engineering Science

1492

being published by the robots and subscribe to a

control program.

The detection of the obstacle is known by

subscribing sensors data to control program. Then the

sensors data that are obtained can be used to generate

the map of environment that has been travelled by

each robotino in form of 2D map. In t h i s study, the

environment based on the Robotics Laboratory in

Bandung Polytechnic for Manufacturing. The

generated map will display the path that robot has

been travelled and the obstacle that are detected

during the time of the robot travelled.

The entire system is focused to safely and

efficiently travelled the environment while

generating the information of the environment that

are travelled by robots. This can be further used for

both robots to travel back safely after knowing the

obstacle position is located and using more efficient

path planning and efficient cost trajectory algorithm.

3 GAZEBO ROS SIMULATION

This section discusses the simulation of the multi

robotino system on Gazebo simulator and present the

detailed information about the motion planning

algorithm of the system simulation.

Figure 3: Map Figures in Gazebo.

4 SYSTEM MODEL ON GAZEBO

ROS

Robot Operating System allows to simulate a system

that has been planned on the previous section. There

are two type of simulation, Gazebo and Rviz. Each

simulation has different purpose, Gazebo simulator is

aimed as the visual representation and as the main

workspace to integrate the 3d models of the robotino

and environment to the ROS messages or services. Its

allows to simulate robot on an environment without

the real robot. While Rviz are mainly functioned to

display the data of the messages or nodes that are used

on the Gazebo simulator.

Figure 4: Gazebo Workspace.

The simulation requirement is the description of

the Robotino and environment. This simulation

includes the plugin for each component (sensors and

actuators) to work accordingly as the real component.

The description of the Robotino in ROS is as Unified

Robot Description Format (URDF)form that has the

function to describe the configuration of the robot

such as dimensions and plugins that are needed to

control the robot on Gazebo. The URDF file

formatted as XML file. Meanwhile the environment

of the simulation can be made directly on Gazebo

using building editor and model editor.

Figure 5: Multi Robotino on one environment.

The Rviz workspace that represent 2d map and

topics can be used for navigation planning. After all

environments are explored by the robots, 2d map can

be saved by using a map saver message.

Multiple Festo Robotino Navigation using Gazebo-ROS Simulator

1493

Figure 6: Rviz workspace.

4.1 Motion Planning Algorithm

Motion planning is the key to achieve the travelling

task of robotino on the environment without having

any collision and losses. It is required two

components, the path and the behaviour of Robotino

to avoid any obstacle while moving to the goal

position. The path will be selected as straight line

between the goal and the initial position.

Figure 7: Calculation of the Hypothenuse as the path

between point A and B.

The algorithm that are being used in this study are

Bug Algorithm. This algorithm has a simple thinking

process for each Robotino to be computed on their

system and two behaviours while executed, move to

goal behaviour and avoid obstacle. Move to goal

behaviour will calculate the initial position of each

robotino to their goal position and begin to move the

robot on a straight line between those position, if the

robot detect an obstacle, they begin to change the path

to avoid obstacle by following the contour of the

obstacle. In the same time, the algorithm calculates

the new position and try to maintain the straight line

path that were calculated on the beginning of the

algorithm. This process is repeated until the robots

reach the goal.

Figure 8: Motion Planning Algorithm.

While traversing the environment, each robotino

has to leave a trace of the obstacle that were

encountered and their traversable path. This allows

the robot to return to their original position by

choosing a more effective path and trajectory.

Figure 9: Visualization of the algorithm.

iCAST-ES 2021 - International Conference on Applied Science and Technology on Engineering Science

1494

5 EXPERIMENTAL

In the experiment scenario, each robot will travel to

the goal position on the other side of the room from

their initial position. Both robot will encounter

obstacle and trying to maintain the path that were

defined in the beginning of the algorithm, a straight

line between the initial position and the goal position.

On Rviz, each robot will leave a 2d map along the

way to their goal as shown in Fig. 10.

Figure 10: 2D Map generated from the encounter of the

robot with obstacle.

As shown on the Fig. 11 that the original path (the

straight line) as the reference of the path for the robot.

Thus the robot will try to avoid the obstacle (the block

on the middle of the room) while trying to return the

original path.

Figure 11: Visualization of the algorithm.

6 CONCLUSIONS

The robots are able to travel safely to the goal position

while leaving a trace on the Rviz to visualize the

obstacles that have been encountered. However, the

Gazebo simulation has successfully spawn both

robots on the same environment with each namespace

attached to their topics. Bug Algorithm has its own

limitation, for instance while the robot trapped on U

shape obstacle. However, this can be overcome by

changing the obstacle avoidance behaviour to

circumnavigate the obstacle, therefore the robots will

select the traversable side point of the obstacle to

reach the goal position.

REFERENCES

Ali, T. Y., & Ali, M. M. (2015, November). Robotino

obstacles avoidance capability using infrared sensors.

In 2015 IEEE Jordan Conference on Applied Electrical

Engineering and Computing Technologies

(AEECT) (pp. 1-6). IEEE.

Anggraeni, P., Mrabet, M., Defoort, M., & Djemai, M.

(2018, October). Development of a wireless

communication platform for multiple-mobile robots

using ROS. In 2018 6th International Conference on

Control Engineering & Information Technology

(CEIT) (pp. 1-6). IEEE.

Borenstein, J., & Koren, Y. (1991). The vector field

histogram-fast obstacle avoidance for mobile

robots. IEEE transactions on robotics and

automation, 7(3), 278-288.

Draganjac, I., Miklić, D., Kovačić, Z., Vasiljević, G., &

Bogdan, S. (2016). Decentralized control of multi-AGV

systems in autonomous warehousing

applications. IEEE Transactions on Automation

Science and Engineering, 13(4), 1433-1447.

Klancar, G., Zdesar, A., Blazic, S., & Skrjanc, I.

(2017). Wheeled mobile robotics: from fundamentals

towards autonomous systems. Butterworth-Heinemann.

Mylvaganam, T., Sassano, M., & Astolfi, A. (2017). A

differential game approach to multi-agent collision

avoidance. IEEE Transactions on Automatic

Control, 62(8), 4229-4235.

Oltean, S. E., Dulău, M., & Puskas, R. (2010, May).

Position control of Robotino mobile robot using fuzzy

logic. In 2010 IEEE International Conference on

Automation, Quality and Testing, Robotics

(AQTR) (Vol. 1, pp. 1-6). IEEE.

Sabattini, L., Cardarelli, E., Digani, V., Secchi, C., &

Fantuzzi, C. (2016). Multi-AGV Systems in Shared

Industrial Environments: Advanced Sensing and

Control Techniques for Enhanced Safety and Improved

Efficiency. In Autonomous Industrial Vehicles: From

the Laboratory to the Factory Floor. ASTM

International.

Yan, R., Jackson, L. M., & Dunnett, S. J. (2017).

Automated guided vehicle mission reliability modelling

using a combined fault tree and Petri net approach. The

International Journal of Advanced Manufacturing

Technology, 92(5), 1825-1837.

Multiple Festo Robotino Navigation using Gazebo-ROS Simulator

1495