Content uploaded by Micael S. Couceiro
Author content
All content in this area was uploaded by Micael S. Couceiro on Feb 04, 2020
Content may be subject to copyright.
Abstract—This article presents the full integration of compact
educational mobile robotic platforms built around an Arduino
controller board in the Robotic Operating System (ROS). To that
end, a driver interface in ROS was created to drastically decrease
the development time, providing hardware abstraction and intui-
tive operation mode, allowing researchers to focus essentially in
their main research motivation. Moreover, the full integration in
ROS provided by the driver enables the use of several tools for
data analysis, easiness of interaction between multiple robots,
sensors and teleoperation devices, thereby targeting engineering
education. To validate the approach, diverse experimental field
tests were conducted using different Arduino-based robotic plat-
forms.
I. INTRODUCTION
Mobile robotics is a technological field and a research area
which has witnessed incredible advances for the last decades. It
finds application in areas like automatic cleaning, agriculture,
support to medical services, hazard environments, space explo-
ration, military, intelligent transportation, social robotics, and
entertainment [1]. In robotics research, the need for practical
integration tools to implement valuable scientific contributions
is felt frequently. However, roboticists end up spending exces-
sive time with engineering solutions for their particular hard-
ware setup, often reinventing the wheel. For that purpose, sev-
eral different mobile robotic platforms have emerged with the
ability to support research work focusing on applications like
search and rescue, security applications, human interaction or
robotics soccer and, nowadays, almost every major engineering
institute has one or more laboratories focusing on mobile robot-
ics research.
Earlier, the focus of research was especially on large and
medium systems. However, with recent advances in sensor
miniaturization and the increasing computational power and
capability of microcontrollers in the past years, the emphasis
has been put on the development of smaller and lower cost
robots. Such low-cost platforms make affordable the experi-
mentation with a larger number of robots (e.g., in cooperative
robotics and swarm robotics) and are also ideal for educational
purposes. With such assumptions in mind, we have been doing
engineering and research work with two Arduino-based mobile
platforms [2]: the TraxBot [3] and the StingBot1. The choice
fell upon Arduino solutions, since it presents an easy-to-learn
programming language (derived from C++) that incorporates
various complex programming functions into simple commands
that are much easier for students to learn. Moreover, the sim-
plicity of the Arduino to create, modify and improve projects,
as well as its open-source and reduced cost makes it among the
This work was supported by the CHOPIN research project (PTDC/EEA-
CRO/119000/2010), by PhD scholarships SFRH/BD/64426/2009 and
SFRH/BD /73382/2010, and by ISR-Institute of Systems and Robotics (project
PEst-C/EEI/UI0048/2011), all of them funded by the Portuguese science
agency "Fundação para a Ciência e a Tecnologia" (FCT).
A. Araújo, D. Portugal, M. Couceiro and R. P. Rocha are with the Institute
of Systems and Robotics, University of Coimbra, Pólo II, 3030-290 Coimbra,
Portugal, email: {aaraujo, davidbsp, micaelcouceiro, rprocha}@isr.uc.pt.
1http://www.isr.uc.pt/~aaraujo/doc
most used microcontroller solutions in the educational context
[2].Following the trend of research, in this work the focus is on
educational, open-source platforms that enable researchers,
students and robot enthusiasts to quickly perform real world
experimentation, having access to the tools provided by the
Robotic Operating System (ROS) [4]. ROS is currently the
most trending and popular robotic framework in the world,
reaching critical mass and being the closest one to become the
standard that the robotics community urgently needed.
With the exponential growth of robotics, some difficulties
have been found in terms of writing software for robots. Differ-
ent types of robots can have wildly varying hardware, making
code reuse nontrivial. Opposing this tendency, ROS provides
libraries and tools to help software developers to create robot
applications. The major goals of ROS are hardware abstraction,
low-level device control, implementation of commonly-used
functionally, message-passing between processes and package
management. One of its gold marks is the amount of tools
available for the community like the Stage simulator [5], navi-
gation capabilities 2, visual SLAM [6] and 3D point cloud based
object recognition [7], among others. Regular updates enable
the users to obtain, build, write and run ROS code across multi-
ple computers.
In the next section, we review general purpose and educa-
tional mobile robots, focusing on those already integrated in
ROS and briefly describe our Arduino-based robot platforms.
In section III, the main contributions of this work are revealed
and details on the development of the ROS driver and its fea-
tures are presented. In the subsequent section, preliminary
results with physical Arduino-based robots and a team of
mixed real and virtual cooperating agents are presented. Final-
ly, the article ends with conclusions and future work.
II. RELATED WORK
The following requirements, sorted by relevance, can be ex-
pected from robots to be used for educational purposes [8][9]:
xCost — Robots should be as cheap as possible to overcome
budget limitations and evaluate multi-robot applications
(e.g., swarm robotics);
xEnergy Autonomy — Robots should have a long battery life
since they may have to operate long enough during devel-
opment and experimentation;
xCommunication — Robots need to support wireless com-
munication to increase the range of applications (e.g., coop-
erative systems);
xSensory System — Robots should be equipped with some
form of sensing capability to allow interaction between
them and with their environment;
xProcessing — Robots need to be able to process information
about other robots and the environment (e.g., sensing data).
2http://www.ros.org/wiki/navigation
Integrating Arduino-based Educational Mobile Robots in ROS
André Araújo, David Portugal, Micael S. Couceiro and Rui P. Rocha
Proceedings of the 13th International Conference on Mobile Robots and Competitions
April 24, 2013. Lisbon, Portugal
© 2013 IEEE
Robotica 2013 was organized under the IEEE Robotics & Automation Society Technical Co-Sponsorship
978-989-97531-2-9
8
Fig. 1. Well-known educational and research mobile robotic platforms:
from left to right, iRobot Create, Turtlebot, Mindstorm NXT, e-puck, MarXbot, SRV-1 Blackfin and Pioneer 3-DX, respectively.
The following subsection reviews popular educational and
research platforms available in the market, after which we pre-
sent the Arduino-based educational platforms developed and
evaluate them according to the requirements presented above.
A. Educational Robotic Platforms
Several off-the-shelf mobile robots with various sensors and
diverse capabilities are illustrated in Fig. 1. We address their
mobility within different ground environments, capabilities,
size, sensing/perception, processing power, autonomous navi-
gation and integration in ROS.
The Roomba Create [10] from iRobot was designed for stu-
dents and researchers, being very popular in the robotics com-
munity due to its small size and low cost. It is a circular plat-
form, with extra space for larger sensors (e.g., 2D laser sensor
or Kinect). Many choose to utilize an external computer that
supports serial communication to control the Create robot, due
to troublesome limitations in storage space and processing
power. A ROS driver for the Roomba iCreate has already been
developed (irobot_create_2_1 package in the brown_drivers
stack), as well as the original vacuum cleaning Roomba (room-
ba_robot stack).
In fact, a popular off-the-shelf robot, developed at Willow
Garage, has been built upon an iRobot Create: the TurtleBot3.
This is a modular development platform incorporating an Xbox
Kinect and an ASUS eeePC 1215N netbook. TurtleBot pro-
vides 3D functionalities and ROS out of the box (through the
turtlebot stack), being fully open source and exploring all com-
bined capabilities of its components.
The Mindstorms NXT [11] from Lego is an educational, ac-
ademic robot kit, ideal for beginners. The robot is equipped
with drive motors with encoders and a good variety of cheap
sensors like an accelerometer, light, sound, ultrasound and
touch sensors. Support for interfacing and controlling this robot
with ROS is also available, through the nxt stack.
The e-puck [12] is an educational swarm platform for be-
ginners. It has tiny dimensions with only 80mm of diameter,
equipped with a vast set of sensors like microphones, infrared
sensors, 3D accelerometer and a VGA camera. Similarly, the
MarXbot [13] platform has 170 mm of diameter, being fully
equipped with infrared range sensors, 3D accelerometer, gyro-
scope, and an omnidirectional camera. It has a good processing
power with an ARM 11 processor at 533MHz. Both the e-puck
and the MarXbot are programmed in a user-friendly scripting
language, which uses ASEBA, an event-based low level control
architecture. In order to interface it with ROS, a ROS/ASEBA
Bridge has been released (ethzasl_aseba stack4).
Additionally, the SRV-1 Blackfin [14] from Surveyor is a
small-sized robot equipped with tracks with differential config-
uration. This robot has a good processing power with a
1000MIPS at 500MHz CPU, capable of running Linux Kernel
2.6. It is equipped with two IR rangers or optional ultrasonic
ranging and a 1.3MP camera. It also supports Wireless
802.11b/g communication and various I2C sensors. Unlike the
3http://www.willowgarage.com/turtlebot
4http://www.ros.org/wiki/ethzasl_aseba
previous platforms, SRV-1 Blackfin can be driven in rough
terrains due to its tracking system. At the time of writing, only
partial support for ROS is available through the ros-surveyor 5
stack, which offers a driver for the Surveyor Vision System in
ROS.
Among the larger, more equipped and more powerful mo-
bile robots, a reference platform for research and education is
the Pioneer 3 DX from ActivMedia [15]. This is a robust differ-
ential drive platform with 8 sonars in a ring disposition, a high-
performance onboard microcontroller based on a 32-bit Renesas
SH2-7144 RISC microprocessor, offering great reliability and
easiness of use. Compared to the previously referred robots, this
robot has greater weight and less affordability. Two different
drivers are available to interface the Pioneer 3 DX with ROS:
ROSARIA6 and p2os7.
B. Arduino-Based Robotic Platforms
Even though most of referred platforms provide open source
software, they usually require a slow learning curve and the
hardware has limited expandability. Arduino solutions have
recently appeared in the market to work around such issues. For
this reason, our platforms were built around an Arduino control
board, which accesses the motor encoders and other infor-
mation from the power motor driver like temperature and bat-
tery state, being also able to send commands to the motors, read
sonar information and exchange messages natively through
Zigbee. Although this section briefly describes the platforms
assembled in our research laboratory, the proposed driver could
be applied to any other Arduino-based platform such as the
eSwarBot [8], the Bot’n Roll OMNI 8 and many others (e.g.,
[1]).
The Arduino-based platforms under consideration, namely
the TraxBot v1 and v2 and the Stingbot [3], are depicted in Fig.
2. All these platforms’ processing units consist of Arduino Uno
boards, which include a microcontroller ATmega 328p that
controls the platforms motion through the use of the Bot’n Roll
OMNI-3MD motor driver8.
Fig. 2. Arduino-based robotic platforms,
a) TraxBot v1; b) TraxBot v2; c) StingBot.
As for power source, two packs of 12V 2300mAh Ni-MH
batteries ensure good energy autonomy to the robots. For dis-
tance sensing, 3 Maxbotix Sonars MB1300 with a range of
approximately 6 meters were used. However, and as experi-
5https://github.com/rene0/ros-surveyor
6http://www.ros.org/wiki/ROSARIA
7http://www.ros.org/wiki/p2os
8http://botnroll.com/omni3md
a)
b)
c)
Proceedings of the 13th International Conference on Mobile Robots and Competitions
April 24, 2013. Lisbon, Portugal
© 2013 IEEE
Robotica 2013 was organized under the IEEE Robotics & Automation Society Technical Co-Sponsorship
978-989-97531-2-9
9
mental results depict, the sensing capabilities of the platforms
can be easily upgraded with other sensors, e.g., laser range
finders. Moreover, the platforms have the ability to also in-
clude a 10” netbook on top of an acrylic support, which extends
the processing power and provides more flexibility. In our case,
ASUS eeePC 1025C were used due to their reduced price and
size. The notebook provides communication via Wireless Wi-Fi
802.11 b/g/n to the robot and is dedicated to run ROS onboard,
providing the tools and means for enhanced control of the ro-
bot. Additionally, the platforms are also equipped with an Xbee
Shield from Maxstream, consisting on a ZigBee communication
module with an antenna attached on top of the Arduino Uno
board as an expansion module. This Xbee Series 2 module is
powered at 2mW having a range between 40m to 120m, for
indoor and outdoor operation, respectively.
C. Summary
Both Arduino-based platforms meet all the requirements pre-
viously pointed out, being ideal for multi-robot applications. In
terms of cost, our platforms have a similar price to the Mind-
storms NXT, being more affordable than the Turtlebot, e-puck,
MarXbot or the Pioneer. In terms of energy autonomy, both the
TraxBot and the Stingbot can operate continuously around 3
hours, which is a common operation time for compact plat-
forms. As for communication, unlike the iRobot Create and the
Pioneer, which do not offer multi-point communication out of
the box, our platforms support Zigbee communication, which is
extended with WiFi when using a netbook. Having distance
sensors and wheel encoders with high resolution, these plat-
forms have the flexibility to incorporate even more custom
sensors, as opposed to the SRV-1 Blackfin or the Mindstorms
NXT. Furthermore, its hybrid design enables not only to make
use of the 24 MIPS at 26MhZ Atmega 328 microcontroller, but
also the Intel Atom N2800 Dual Core at 1.86 GhZ processor of
the netbook, similarly to the Turtlebot and outperforming the
smaller platforms.
Additionally, when developing our educational robots other
requirements were taken into account: all hardware is either
made of aluminium or stainless steel, being extremely robust;
their dimensions are adequate for both indoor and outdoor ex-
periments; and they have the ability to run ROS.
III. ROS DRIVER FOR ARDUINO-BASED ROBOTS
The key contributions of this work are the development and
description of a driver that enables fast prototyping through the
interface and control of custom educational platforms with
ROS, which can be generalized to different Arduino-based
platforms.
ROS provides tools to interface with the Arduino family of
boards through the rosserial stack 9.However, it was verified
that rosserial is not suitable for this work, due to the high over-
head imposed by its data acquisition and commands, which
result in an excessive workload to the Arduino microcontroller
Atmel 328p SRAM. In fact, the microcontroller presents lim-
ited SRAM memory and for standard ROS topics (float32 mes-
sages + message headers), stress tests have shown that only a
maximum of 15 ROS topics can be used in parallel and the
message buơer is limited to 70 standard messages.
The most important feature in rosserial is to add libraries to
the Arduino source code, in order to emulate ROS language
directly in Arduino code. This results in high overhead in com-
munication between PC / ROS and the Arduino, due to the
9http://www.ros.org/wiki/rosserial
structures used, for example, when publishing messages from
the Arduino side. For this reason, a custom driver was created,
being able to adopt a faster and more transparent communica-
tion between any Arduino board and ROS. We propose a solu-
tion based on the serial_communication stack 10, where the
messages sent from the Arduino only consist of arrays of char-
acters, which are parsed to integer variables on the PC / ROS
side, hugely decreasing the communication load.
A. Driver Description
The mrl_robots11 driver herein presented was developed for
integration and control of the platform using ROS Fuerte ver-
sion running on Ubuntu 11.10 “Oneiric Ocelot”. The seri-
al_communication stack 10, was used to establish point-to-point
serial communication between the Arduino and the PC / ROS
side, without the overhead of rosserial. This enables robust and
fast communication in more complex applications, such as
teleoperation, crossing of sensory information, the integration
of the navigation stack, among others. It also has the versatility
of creating protocols to exchange data between the Arduino
and the PC/ROS side, which enables the creation of a custom-
ized and transparent serial communication.
The Arduino firmware code was developed taking into ac-
count all components and their features, which are required for
the robots’ operation. In Fig. 3 the architecture of the ROS
Driver is illustrated. The power motor driver OMNI-3MD
provides libraries to control the motors (i.e., velocity or posi-
tion control), read encoders and temperature, as well as setting
the parameters for the initial configurations of the PID control-
ler, among others. The motor driver is connected to the Ar-
duino Uno through I2C communications. C/C++ language was
used as the programming language for the ATmega328p mi-
crocontroller. Algorithm 1 illustrates the resident Ro-
bot/Arduino Firmware code.
ROS Driver
Connection node
TraxBot Source Code
Cereal_port node establish I2C
communication via serial
protocol and Driver source
code interprets the frame sent
by the firmware program to
perform the desired task.
Driver Source Code
USB cable
connection
Firmware
Implemented in C with specific
protocol frame
Motor Driver
Power driver to control motors,
Encoders and temperature among others.
I2C
Fig. 3. ROS driver architecture diagram.
The protocol developed to interface ROS with the Arduino
board consists on sending a frame with the configuration
shown in Fig. 4. The character ‘@’ is used at the beginning of
every frame, and commas ‘,’ separate the different parameters.
Character ‘e’ identifies the end of the frame. Regarding the
content of the protocol, the first parameter corresponds to the
action command; like move motors, and others (Algorithm 1).
Following the action command, commas separate the argu-
ments of the designated commands which have been defined as
signed integers.
10 http://www.ros.org/wiki/cereal_port
11 http://www.ros.org/wiki/mrl_robots
Proceedings of the 13th International Conference on Mobile Robots and Competitions
April 24, 2013. Lisbon, Portugal
© 2013 IEEE
Robotica 2013 was organized under the IEEE Robotics & Automation Society Technical Co-Sponsorship
978-989-97531-2-9
10
Fig. 4. Frame protocol to receive/send data from/to the Arduino Uno
Let us suppose, for instance, that we want the platform to
move with a linear velocity of 0.5 m/s and an angular velocity
of -0.8 rad/s, the frame would be, “@11,500,-800e” represent-
ing “@command,(lin_vel×103),(ang_vel×103)e”.
In the ROS side, a computation process (robot_node) has
been programmed, which starts the serial connection using the
cereal_port library of serial_communication stack and receives
streams of information from the robot. Whenever a data frame
has been received, a callback is triggered, publishing the corre-
sponding message into appropriate ROS topics, thus providing
updated information to the rest of ROS ecosystem. Algorithm
2 shows how the driver works. In Fig. 5, it is shown how a
ROS user application node (e.g., a mapping algorithm) can
interact with robot_node by sending velocity commands to the
base and receiving information like sonar range, odometry,
transforms, etc. One of many ROS tools, rxgraph, has been
used to allow real time monitoring of the available nodes, as
well as topics exchanged by each node. Note also the interac-
tion with other existing nodes in ROS like the wiimote_node12,
used for teleoperating the robot through a Nintendo’s Wii
remote controller (WiiMote), and the hokuyo_node13 to add an
Hokuyo laser range sensor to the robot. ROS provides many
different built-in sensor message types which are appropriately
assigned to the topics of each component of the driver.
The ability to stream data from the Arduino board is an in-
teresting feature of the driver because it does not require a
synchronous communication involving requests and responses
between ROS and the Arduino board. Hence, it frees the serial
communication channel since it only needs a starting request
and can be stopped at any time. Furthermore the mrl_robots
driver has the ability to enable and disable debugging options
to track eventual errors.
B. Driver Features and Potential
The driver presented in the last subsection offers several fea-
tures, many of which are inherited by the direct integration
with the ROS middleware. The driver enables the interface
with ROS tools for data process and analysis of the platforms,
like 3D visualization (rviz), logging real-time robot experi-
ments and playing them offline with (rosbag/rxbag), plotting
data (rxplot) and visualizing the entire ROS network structure
(rxgraph).
Beyond the easiness of using the available tools, ROS also
provides effortlessly integration of new sensors without need-
ing hardware expertise, as it will be seen in section IV. This
opens a new range of possibilities since several well-known
stacks from the ROS community comprise algorithms for ro-
botics development such as the navigation2 and
slam_gmapping14 stacks.
12 http://www.ros.org/wiki/wiimote_node
13 http://www.ros.org/wiki/hokuyo_node
14 http://www.ros.org/wiki/slam_gmapping
Fig. 5. Rxgraph topics and nodes provided by the mrl_robots driver.
Algorithm 1. Robot/Arduino Resident Firmware
1:
2:
3:
4:
5:
#Omni3MD library // main motor driver command functions
#EEPROM library // storage robot particular specifications: robot ID,…
#Robot library
// range sonars acquisition, calibration, PID gains
#
RobotSerialComm library // protocol serial communication
#Standard libraries
6:
7:
Setup Functions(); // PID motor gains, using ports, encoders scale, set I2C
connection,…
8:
Streaming Functions():
9:
sendEncodersReads()
10:
11:
Read encoder 1 and 2 pulses;
Serial reply encoder data;
12:
sendEncodersSonarsReads()
13:
15:
16:
Read encoder 1 and 2 pulses;
Read sonars 1, 2 and 3 ranges;
Serial reply encoder and sonar data;
17:
sendRobotInfo()
18:
19:
20:
21:
22:
23:
Read from EEPROM robot ID;
Read internal board temperature;
Read Omni
-3MD driver firmware version;
Read TraxBot battery voltage;
Read firmware version;
Serial reply info data;
24:
Main loop():
25:
Switch (action):
26:
27:
28:
29:
30:
31:
32:
33:
34:
35:
36:
37:
38:
39:
40:
41:
42:
43:
44:
45:
sendEncodersReads;
sendEncodersSonarsReads;
sendRobotInfo;
Omni
-3MD auto-calibration motors for controller purposes;
Set PID gains;
Receive Xbee message;
Send Xbee message;
Xbee node discovery;
Set prescaler from encoders;
Set
desire encoders values;
Robot info;
Single encoders reading;
Single sonars reading;
Linear motors move with PID controller;
Linear motors move;
Stop motors;
Reset encoders;
Debug action;
// (Des)Activate debug option
Start streaming data;
// Activate desire streaming data
Stop streaming data;
Algorithm 2. PC/ROS Driver.
1:
2:
#ROS_msgs library // ROS type messages
#Cereal_port library // protocol serial communication
3:
Robot data callback():
4:
UpdateOdometry()
5:
6:
7:
8:
Read encoder pulses;
Pulses convert to cartesian pose (ݔǡݕǡ ߠ
);
Publish in ROS topic updated pose;
Publish tf: odom → base_link
9:
DriveRobot()
10:
11:
Subscribe ROS topic differential velocity commands;
Send to robot angular and linear speeds;
12:
RangeUltrasonicSonars()
13:
Publish ROS topic range ultrasonic sonars;
14:
XbeeMsgs()
15:
16:
Publish ROS topic with Xbee message received;
Subscribe ROS topic with Xbee message to send from user node
17:
UpdateRobotInfo()
18:
Publish ROS topic robot information;
19:
Main loop():
20:
Establish a serial connection;
21:
Receive data streaming from robot (activate callbacks);
Proceedings of the 13th International Conference on Mobile Robots and Competitions
April 24, 2013. Lisbon, Portugal
© 2013 IEEE
Robotica 2013 was organized under the IEEE Robotics & Automation Society Technical Co-Sponsorship
978-989-97531-2-9
11
As a result, the overall time spent in robotics research is
greatly reduced due to code reuse and therefore the driver
represents a valuable tool that enables fast prototyping and
opens a gateway into the world of ROS.
Another interesting feature of the driver is the simplicity for
enabling heterogeneous multi-robot coordination and coopera-
tion. Running the same hardware abstraction layer in the team
of robots, ROS takes care of the communication messaging
system using a publish/subscribe method, which enables all
kinds of interaction between members of the same team, as
seen in Fig. 6, where an example of a ROS network is depic-
ted.
ROS
Gazebo
ROS Gazebo Node
Driver Node Stingbot
...
Robot N
Driver Node Traxbot v2
Robot 2
Driver Node Traxbot v1
Robot 1
Computer
Multi
Virtual robots in Stage
Nodes
Main
Algorithm
Node
Add specific
task
Node Hokoyo Laser Node
Kinect Node
Wii Remote Controller
Node
...
Navigation
Stack
Fig. 6. Network topology example with multiple robots, sensors,
tele-operation devices and applications.
ROS also has the potential to integrate mixed real and virtu-
al robot teams. With the use of the driver herein presented,
together with the Stage15 multi-robot simulator, the same code
can be used to drive either real robots or virtual agents running
on Stage. Therefore, the developed ROS driver allows the
integration of virtual robots with different sizes and driving
characteristics, as seen later on. In addition, the communica-
tion between real and virtual agents is completely transparent
since they are both operating in the same ROS network. This
major feature is ideal to perform multi-robot tasks, allowing
the use of a large population of robots, when no extra physical
robots are available, being cheaper and promoting safer test
scenarios by making interactions between physical and virtual
world objects.
Fig. 7. Evaluation of the ROS driver in Traxbot v1 with different sensors. a)
Ultrassound Range Sensors integration; b) Hokuyo URG-04LX Laser Range
Finder integration.
IV. RESULTS AND DISCUSSION
In order to experimentally evaluate the ROS driver, some
tests were conducted using physical Arduino-based robots and
15 http://www.ros.org/wiki/stage
stage15, which provides essential options like the information
about the ground truth pose and odometry of virtual robots.
We present experimental tests that validate the aforemen-
tioned claims and we also show cooperative behaviors with
real multi-robot systems, as well as mixed real and virtual
robotic teams 16. Therefore, the experiments will allow to eval-
uate the driver flexibility to different sensors, the driver porta-
bility to different robotic platforms and the driver expandabil-
ity and integration with the existent ROS tools.
The first experiment aims to demonstrate the driver
flexibility to integration of different sensors. The TraxBot v1
platform was equipped with a laser range finder (LFR) and its
performance was compared against the native ultrasonic range
sonars on a simple mapping task. The Hokuyo URG-04LX is a
LRF classified as an Amplitude Modulated Continuous Wave
(AMCW) sensor. In order to test the sonars performance, an L-
shaped scenario of 2 m by 1.6 m was set up, with a 1 m width
(Fig. 7). To perform this test, two lateral sonars placed at ±45
degrees were used.
In this test, the robot movement relies solely on odometry.
In Fig. 7a it can be seen in the first rectilinear motion, that the
sonars readings are stable (red dots) and coincident with the
ground truth scenario limits (blue line). Green dots represent
the midpoint of sonars acoustic beam while turning. Some
issues arise during the 90 degrees rotation, since the sonar
beam cone has an opening of approximately 36 degrees, thus
presenting a much poorer resolution than the LRF, as
illustrated in Fig. 7a. In the case of Fig. 7b, the Hokuyo LRF
was used to perform the same test. The overture of the laser
was set to 180 degrees with a rate of 512 samples per reading.
It is possible to observe some discrepancy in some readings
especially at the end of the movement due the odometry
position error accumulated during motion.
In the second experiment, the main goal is to demonstrate
the portability of the driver to different robots and sensors,
which enables testing in our Arduino-based robots the existent
algorithms in ROS. Hence, a mapping task with the
incorporation and interaction of drivers for different sensors
like LRF and a joystick controller was performed. Along these
lines, this time the Traxbot v2 platform was equipped with an
Hokuyo LRF and teleoperated with a WiiMote for a mapping
task using Hector Mapping [16], which is available in the
hector_slam stack17.
Fig. 8. Map generated by the Traxbot v2 with hector mapping.
16 A video of the experiments is available at:
http://www.isr.uc.pt/~aaraujo/videos/Robotica2013
17 http://www.ros.org/wiki/hector_slam
a)
b)
Position x [mm]
Position x [mm]
Position y [mm]
Position y [mm]
Proceedings of the 13th International Conference on Mobile Robots and Competitions
April 24, 2013. Lisbon, Portugal
© 2013 IEEE
Robotica 2013 was organized under the IEEE Robotics & Automation Society Technical Co-Sponsorship
978-989-97531-2-9
12
The teleoperation node runs on the eeePC netbook
connected over the ROS network. The node subscribes to a
topic which has the information of the Wiimote state and
assigns functions for each pressed button, publishing then
velocity commands, which are interpreted by the ROS driver,
resulting then on motor commands. Additionally, the
hector_mapping node subscribes to the scans provided by the
hokuyo_node and publishes the estimate of the robot's pose
within the map, while generating the map. Fig. 8 presents the
resulting map in rviz, which was obtained with
hector_mapping on our experimental lab arena.
In the third and final experiment, we show not only the
possibility to have coordinated behaviors with two physical
robots, but also the possibility to include a third simulated
robot running in stage which communicates with the other two
forming a mixed team of real and virtual robots, as described
in section III.B. In addition, we also integrate navigation
capabilities in our robots, by running the navigation stack2
with a known map, in this case, the map of our experimental
lab arena (Fig.9).
The robots were commanded to navigate cyclically between
a set of waypoints in the arena, as seen in the video of the
experiment16. To further demonstrate their coordinating abili-
ties, a common waypoint for all three robots was defined, and
robots had to exchange messages through a shared ROS topic
to avoid going to the common point at the same time. They
would wait to go to the point at the center of the arena and
priority was given to the robot who expressed firstly its inten-
tion to move to that point. All three robots were able to coordi-
nate themselves in the environment without colliding to each
other, due to the integration of the navigation stack2.Fig.10
presents a snapshot of rviz, illustrating the three robots moving
in the arena.
Fig. 9. Experimental arena with a Traxbot v2 and a Stingbot cooperating
with a virtual robot, running on stage.
Fig. 10. The three robots coordinating their behaviors by exchanging ROS
messages (rviz).
V. CONCLUSIONS AND FUTURE WORK
In this paper a solution for integrating Arduino-based robot-
ic platforms in ROS, through the development of a ROS driver,
was presented. It was shown the great advantages of integrat-
ing the platform with ROS middleware, enabling the usage of a
wide range of tools and reducing the development time
through code reuse. The robots, alongside with Arduino and
ROS open-source development tools, present themselves as
ideal platforms for educational robotics. Beyond providing
access to all ROS tools, the driver also simplifies the robotic
development by: i) supporting hardware abstraction to easily
control the platform; ii) allowing for the extension and integra-
tion of all kinds of sensors; and iii) enabling multi-robot coop-
eration and coordination through the operation in a ROS net-
work, both for real teams of homogeneous and heterogeneous
robots, as well as hybrid teams of real and virtual agents, run-
ning the same code. Results from the experiments that were
conducted demonstrate all these features and the insignificant
overhead imposed by the driver was discussed.
REFERENCES
[1] Brooks, Rodney A., “New Approaches to Robotics”, Science, vol. 253,
pp. 1227- 1232, 13 September, 1991.
[2] John-David Warren, Josh Adams and Harald Molle “Arduino Robotics”,
Springer Science and Business Media, 2011.
[3] A. Araújo, D. Portugal, M. Couceiro, C. Figueiredo and R. Rocha,
"TraxBot: Assembling and Programming of a Mobile Robotic Platform".
In Proc. of the 4th International Conference on Agents and Artificial In-
telligence (ICAART 2012), Vilamoura, Portugal, Feb 6-8, 2012.
[4] M. Quigley, B. Gerkey, K. Conley, J. Faust, T. Foote, J. Leibs, E. Berger,
R. Wheeler, and A. Y. Ng, "ROS: an open-source Robot Operating Sys-
tem," in Proc. Open-Source Software workshop of the International Con-
ference on Robotics and Automation, Kobe, Japan, May, 2009.
[5] B. Gerkey, R. Vaughan and A. Howard, "The Player/Stage Project: Tools
for Multi-Robot and Distributed Sensor Systems", In Proc. of the Intl.
Conf. on Advanced Robotics, pp. 317-323, Coimbra, Portugal, 2003.
[6] G. Grisetti, C. Stachniss and W. Burgard, "Improved Techniques for Grid
Mapping with Rao-Blackwellized Particle Filters", In IEEE Transactions
on Robotics, 2006.
[7] R. Rusu and S. Cousins, "3D is here: Point Cloud Library (PCL)", In
Proc. of International Conference on Robotics and Automation (ICRA
2011), Shanghai, China, May 2011.
[8] M. S. Couceiro, C. M. Figueiredo, J. M. Luz, N.M. F. Ferreira & R.P.
Rocha. "A Low-Cost Educational Platform for Swarm Robotics", Int.
Journal of Robots, Education and Art, Vol. 2(1), Feb., pp. 1-15, 2012.
[9] Park, I. W. and Kim, J. O., “Philosophy and Strategy of Minimalism-
based User Created Robots (UCRs) for Educational Robotics - Educa-
tion, Technology and Business Viewpoint”, International Journal of Ro-
bots, Education and Art, vol. 1, no. 1, 2011.
[10] M. Kuipers, “Localization with the iRobot Create”. In Proceedings of the
47th Annual Southeast Regional Conference ACM (ACM-SE 47), Clem-
son, South Carolina, USA, March 19-21, 2009.
[11] B. Bagnall, “Maximum LEGO NXT: Building Robots with Java Brains”.
Variant Press, 2007.
[12] Mondada, F. et al., “The e-puck, a Robot Designed for Education in
Engineering”. In Proc. of the 9th Conf. on Autonomous Robot Systems
and Competitions, 1(1), pp. 59-65 2009.
[13] Bonani, et al, “The MarXbot, “a Miniature Mobile Robot Opening new
Perspectives for the Collective-Robotic Research” In Int. Conf. on Intel-
ligent Robots and Systems, Oct. 18-22, 2010, Taipei, Taiwan, 2010.
[14] Cummins J., Azhar, M.Q. and Sklar, E.: Using Surveyor SRV-1 Robots
to Motivate CS1 Students. In: Proceedings of the AAAI 2008 Artificial
Intelligence Education Colloquium, 2008.
[15] S. Zaman, W. Slany and G. Steinbauer, "ROS-based Mapping, Localiza-
tion and Autonomous Navigation using a Pioneer 3-DX Robot and their
Relevant Issues", In Proc. of the IEEE Saudi International Electronics,
Communications and Photonics Conference, Riad, Saudi-Arabia, 2011.
[16] S. Kohlbrecher, J. Meyer, O. von Stryk and U. Klingauf, "A Flexible and
Scalable SLAM System with Full 3D Motion Estimation", In Proc. of the
IEEE International Symposium on Safety, Security, and Rescue Robotics
(SSRR’2011), 50-55, Kyoto, Japan, Nov. 1-5, 2011.
Shared
Point
Proceedings of the 13th International Conference on Mobile Robots and Competitions
April 24, 2013. Lisbon, Portugal
© 2013 IEEE
Robotica 2013 was organized under the IEEE Robotics & Automation Society Technical Co-Sponsorship
978-989-97531-2-9
13
