INTERNATIONAL JOURNAL OF MATHEMATICAL MODELS AND METHODS IN APPLIED SCIENCES

Design of a Configurable All Terrain Mobile

Robot Platform
Gokhan Bayar, A. Bugra Koku, E. ilhan Konukseven

autonomous robots that can outsmart humans, wide range of

Abstract—With the increased funding from various agencies, applications for mobile platforms have been developed.
research conducted in the field of mobile robotics has significantly Despite the rich spectrum of applications, mobile platforms
increased during the last couple of decades. Due to wide range of available for research have been very few in number such as
applications mobile robots of different sizes and capabilities are
ATRV, Pioneer series robots [2, 5] and Packbot [7]. First of
required in the field. Despite the wide spectrum of applications to be
developed for mobile robots, available platforms on the market for all these robots are not economically feasible, second, they are
research purposes are very few in number, and limited in their designed to operate on a specific environment, hence
capability for multi-purpose use. Evidently purchasing new platforms developing diverse applications that require installing
for different applications is not a feasible solution for conducting additional hardware, and drive system manipulations cannot
research. Driven by the motive of having a configurable research be easily and effectively done on these robots. Purchasing a
platform, a mobile robot referred to as CoMoRAT (Configurable
new robot for every different application being developed is
Mobile Robot for All Terrain Applications) has been designed and
manufactured at METU. CoMoRAT can be driven by wheels, tracks not feasible both in terms of money and time.
or both. Besides its ability to ride effectively on various terrains, Different applications require the mobile robots to operate
robot body is designed in such a way that adding new hardware to on different environments [14]. Some robots are designed to
the platform requires minimal manufacturing and installation effort. operate indoors, some outdoors, and some have to be able to
This paper presents the design and construction details as well as the operate in both. Robots operating outdoors may stay on the
performance tests conducted on CoMoRAT. road or navigate off-road. For effective traversing diverse
terrain, it is important to have a driving system that can easily
Keywords—Mobile robot, design, construction, ground,
platform, base, terrain, track, wheel.
be configured to suit the needs of navigation task at hand.
Indoor environments are generally smooth and well
I. INTRODUCTION structured, whereas outdoors, especially off-road sections, are
rough and pose challenges for wheeled robots.
D EMAND for mobile robot applications has significantly
increased in the past couple of decades and mobile
robotics has become a rapidly developing field of
Various types of locomotion methods such as legs, wheels
and tracks are used on mobile robots with various
configurations. For instance, Plunstech walking robot, which
interdisciplinary research within robotics. This promising field has been designed for the forest applications, provides
has attracted the attention of academicy, industry, as well as automatic leg coordination while the operator selects a travel
several government agencies. Currently from security to direction [3]. Six-wheeled mobile robot, Sojourner, was
personal service, mobile robots are being used in a variety of designed for the mission of exploring Mars surface [4]. Both
tasks. Yet the use of such robots is expected to only increase of these robots have excellent rough terrain navigation
in the near future. capability, yet they are neither available to researchers nor
The demand for consumer robots for various applications versatile for diverse applications. Pioneer mobile robot is one
being on the rise [1], and the availability of various funds in of the mobile robot bases that are commonly used in research
addition to big stake competitions, both industry and academy [5].
have shown great interest in developing robotic hardware and Magellan mobile robot is another platform preferred by
applications. With the ultimate purpose of creating fully researchers. It is designed for indoor operation and commonly
used in indoor navigation applications. Being decorated with
Manuscript received November 02, 2009: This work was supported in part
by the project BAP-08-11-DPT-2002K120510.
various sensors, Magellan is not suitable for outdoor
Gokhan Bayar is with the Mechanical Engineering Department, Middle applications [6]. PackBot is a mobile robot platform used for
East Technical University, 06531 Balgat – Ankara TURKEY (phone: +90 the applications of explosive ordnance disposal, search-and-
(312)210 5273; fax: +90(312)210 2536; e-mail: [email protected]). rescue, surveillance, bomb squad, swat teams, and other vital
A. Bugra Koku is with the Mechanical Engineering Department, Middle
East Technical University, 06531 Balgat – Ankara TURKEY (e-mail: tasks [7].
[email protected]).
E. ilhan Konukseven is with the Mechanical Engineering Department,
Middle East Technical University, 06531 Balgat – Ankara TURKEY (e-mail:
[email protected]).

Issue 4, Volume 3, 2009 366

 INTERNATIONAL JOURNAL OF MATHEMATICAL MODELS AND METHODS IN APPLIED SCIENCES

Table 1 Comparison chart about the features of the mobile robots (Adapted from [10])

Sojourney Magellan Pioneer Packbot MR-5 Talon CoMoRAT

Payload (kg) 4 9.1 22.7 25 80 150 15

Velocity (m/s) 2 2.5 1.6 1 1.5 2 1

Dimensions 225x400x 368 440x380x 525 600x480x 1025x680x 300x500x

(mm) 130 h:198 220 h:1060 465 440 80

Motors 2x12V 2x24V 2x12V 4x24V 2x48V 2x48V 2x24V

Embedded
Siemens Embedded
Comp. Platform Siemens - MPC555 MPC555 PC, Perip.
C166 PC
Mic.
Development VB, C++ C#
C++ C++ C++ C C
Tools Matlab Matlab
U-I-T U-I-T U-I-T
Sensors and O-I LS-V LS-C-V LS-C-V U-LS-C LS-G
I-U
Equipments G-L GPS-C GPS-Gr GPS-Gr V-Gr-Pt A-G-O
Gr-Pt Pt Pt
RS232 RS232 RS232 RS232
RS232 RS232
RS232 Wireless Wireless Wireless Wireless
Communication Wireless Wireless
RF RF RF RF RF
RF Ethernet
Ethernet Ethernet Ethernet Bluetooth
Application E E A,E,R A,E,R A,E,R A,E,R A,E,R
W W,T
Actuation W W W T T
W+T W+T
Permission for
Reconfiguration No No No Yes Yes Yes Yes
of Platform

Permission for
Actuation No No No No Yes No Yes
Configuration

The annotations are: E, A, R are education, application, and research, respectively. I, U, O, G are infrared sensor,
ultrasonic sensor, odometer, and gyro, respectively. A, L, T, LS are accelerometer, light captor, tactile sensor, and
laser scanner, respectively. V, Pt, C, GR are vision system, pan tilt camera, navigation compass, and grippers,
respectively

PackBot is a tough, light weight and quickly configurable tracks, MR-5 can be effectively operated in diverse
platform. It also allows integrating flippers, a camera and a environments and situations.
robotic arm on its frame. All the assembling and Another well-known mobile robot platform which is
disassembling stages take only a short period of time and commonly used by U.S. military is Talon. It has been
hence, it can be adapted to different situations very quickly. designed for the missions ranging from exploration of the
However, the price tag of this robot is quite a burden on a working environments to weapons delivery [9]. Even though
limited research budget. integrating some equipment to its base can be done very easily
MR-5 is a remotely controlled mobile robot base. It is and quickly, its track based driving system cannot be
ideally suited for explosive ordnance disposal, swat, harmful modified. Besides its legendry success in military
material search, surveillance and other hazardous environment applications, Talon is not suitable for research and it is even
and material tasks. It has been designed for operating with more expensive than other platforms.
wheel and/or tracks [8]. It has two motion modes: wheels As a result, commercially available robots mentioned so far
tracks on wheels. Being able to move on either wheels or fall short in terms of being the ideal mid-size research

Issue 4, Volume 3, 2009 367

 INTERNATIONAL JOURNAL OF MATHEMATICAL MODELS AND METHODS IN APPLIED SCIENCES

platform. They are either not suitable for different terrain effectively where tracked robots have relatively better
conditions, or they are not configurable enough to meet the performance when the robot moves off the road, traverse over
needs of diverse research requirements. Evidently, purchasing steep hills. In conclusion choosing one of these three methods
a new robot for different research projects is not economically to mobilize a robot introduces a trade-off that has to be faced
feasible in most cases. Hence, CoMoRAT is designed to be a by a researcher.
modular, easy to reconfigure, on and off road capable mobile Indeed the real trade-off that robotic researchers frequently
platform. CoMoRAT is compared to some well-known robots face is when a robot has to be purchased. It is evident that a
on the market in Table 1. Payload that robot can carry, researcher has to deal with different applications in time, yet
velocity, dimensions, type of the motors that drive the robots, most of the commercially available robots on the market do
computational platform, development tools, sensors and not provide a generic solution to suit the needs of a diversified
equipments that can be attached, types of communication, robotics researcher. Such a researcher has to deal with robots
application areas, and actuation types are compared in this moving on different environments (indoor, structure and
table. unstructured terrain etc.). Based on the applications various
During research, developers often need to add and remove types of hardware have to be installed on the development
hardware when deploying new applications to a platform. platform. As a result, a robotic researcher often feels the need
Hence, physically adding new hardware (i.e. sensors, for a mobile platform; that can effectively move on different
actuators, computational units) to the robotic platform should terrain conditions and to which addition of new hardware is
be practical. Diverse applications require the robot to move on practical. Since purchasing a new robot for every different
various types of terrain. To increase the efficiency of the robot application is neither economically nor time-wise practical.
on different terrain, it is important if the robot can be Hence, we have felt the need to develop a single, yet easily re-
configured to drive on wheels, tracks or both. CoMoRAT is configurable mobile platform at the Department of
designed with the expectations that traction system can be Mechanical Engineering at METU which can be used wide
easily changed (to be wheeled, tracked and wheeled + tracked) range of applications with minimal effort.
to suit a wide range of terrain, and allow additional hardware At the beginning of this study, the design requirements of
to be installed with minimal manufacturing and effort. In CoMoRAT are determined as follows:
terms of size, it is aimed to create a robot that can be - Selection of one of the three locomotion method is going
backpacked by the user (end user, or the grad-student) if to restrict the effectiveness of the designed robot.
needed. This study does not attempt to individually address Locomotion is to be selected in such a way that a wide
any specific robotic application, indeed it is aimed to shape up range of applications should be deployable to the
a robotic module that can be used in a wide range of developed platform.
applications on different terrain with proper modifications. It - Diverse applications require different hardware to be
is tested with tracks, wheels, and with both tracks and wheels. present on a robot. Hence, the robot body should enable
The results are can potentially provide some guidelines and addition and relocation of hardware.
directions for other researches designing mobile robots. - Addition of new hardware to the robot should require
In the remaining of the paper, the design requirements are minimal manufacturing and installation should not be
presented in Section II followed by the brief introduction of cumbersome.
the design process. Simulations and experiments in order to - If possible a standard mechanical interface should be
evaluate the performance of CoMoRAT are given in Section determined and any hardware following this interface
IV followed by the conclusions. should be easily mounted on the mobile platform.
- Any reconfiguration work on the robot should be
II. DESIGN REQUIREMENTS AND DECISIONS practical, consume minimal time and not require
A researcher has to deal with problems, the solution of personal talent.
which requires robots moving on various terrain conditions. - The robot should easily be adopted by different students
Commonly used locomotion methods for mobile robots can be with different programming backgrounds in order to
classified into there as legged, wheeled and tracked. Hybrid increase the usability of the platform.
locomotion methods that fit a robot into more than one of
these classes are also available in the literature. Legged With these requirements in mind, CoMoRAT design
locomotion provides perfect terrain adaptation hence process is initiated. Due to its mechanical and control
increased mobility on diverse terrain conditions for mobile complexity, relatively slow motion, low reliability and
robots. Yet, this kind of locomotion is mechanically complex, possibility of high maintenance requirements legged
which in turn brings the disadvantage of slower motion and locomotion is opted out. Between wheeled and tracked
decreased system reliability. Wheeled and tracked locomotion locomotion, neither one is preferred to the other basically due
offers faster and more reliable solutions, yet they lack the to the fact that, wheeled robots perform better on structured
adaptation capability of legged locomotion. Wheeled robots roads whereas tracked robots outperform wheeled ones when
can move and maneuver rapidly on structured roads the terrain is unstructured and especially when the surface is
very rough and steep hills is to be traversed. However, it is

Issue 4, Volume 3, 2009 368

 INTERNATIONAL JOURNAL OF MATHEMATICAL MODELS AND METHODS IN APPLIED SCIENCES

also possible that during operation a robot will not stay in one software. At this point, Maxon motors and EPOS controllers
type of terrain at all times. In such cases, bringing the best of provide an easy yet affordable and reliable solution.
two worlds together will increase the effectiveness of the CoMoRAT’s Maxon motors can easily be controlled via
robot, hence, it is decided that the robot locomotion should RS232 which enables the user command the base through any
enable it to be configured as wheeled, tracked or wheeled + device that implements a not-so-complicated serial protocol.
tracked at the same time.
Necessity to add different hardware to a robot is frequently
faced by robotics researchers. Especially, research on limited III. DESIGN PROCEDURE
budget demands such multi-purpose use. To highlight this fact In order to select CoMoRAT's actuators, a simple model of
consider two grad students working on different projects. the platform as illustrated in Figure 2 is derived. Note that the
They may end up using the same robot base but different model has been developed for an ideal flat surface condition
hardware has to be installed on them so that they can conduct in order to have an idea about the general requirements of the
their own research. In many such cases purchasing a new motion. In the mathematical model the forces coming from
robot for each and every student in a research lab is not ground has been taken into account. The applied wheels'
possible; hence, these robots should enable hardware re- traction forces, which are needed to give the motion to the
configurability as much as possible. robot base, have been modeled as given in Equations (1-6).
Many of the commercially available robots do not provide a The abbreviations used in Figure 2 are specified as: m denotes
practical solution to this hardware re-configurability need. the mass of the platform. g denotes the gravitational
Robot bodies should be drilled when an installed hardware is acceleration. hmg denotes height of the center of mass. FNr and
removed; most of the installation holes remain on the robot FNf denote the normal forces acting on rear and front wheel,
body. This is not only cosmetically a flaw; it also is a treat to respectively. FTr and FTf denote the total tractive effort for rear
the sealed inner content of the robot. As shown in Figure 1, and front wheel, respectively. γf and γr equal 1 for the four
the robot body frame is constructed from extruded aluminum wheel drive. θ is the slope angle of the terrain. The details of
with groves on both side of the frame. This frame structure the model are given in [11].
enables us to design simple adaptor plates for individual
hardware components and locate the hardware either inside or
outside the robot with the same adapter. It is also very
practical to move the hardware once it is installed on a groove
to open up some space for additional hardware. Another
advantage of the grooved frame structure is that, by moving
the hardware (including the batteries) within and around the
robot, the center of gravity of the robot can be adjusted as
needed.

Fig. 2 Motion model of the mobile robot platform

∑ Fx = μ w γ r FNr + μ w γ f FNf − mg sin θ (1)

∑ Fy = FNr + FNf − mg cos θ (2)

∑ M = FNr L w − mg(L cg cos θ + h sin θ) (3)

Fig. 1 Frame of the robot base

mg(L cg cos θ + h sin θ)
FNr = (4)
Finally, although configurable locomotion and hardware are Lw
necessary conditions for a multi-purpose mobile research
platform, hardware integration to the overall control system is mg((L w − L cg ) cos θ − h sin θ)
an issue that has to be handled via software. Hence, the base FNf = (5)
Lw
should not restrict the user to a specific operating system.
Considering the fact that, bare bone CoMoRAT contains only
the driving system, the basic need is portable motor control

Issue 4, Volume 3, 2009 369

 INTERNATIONAL JOURNAL OF MATHEMATICAL MODELS AND METHODS IN APPLIED SCIENCES

the profiles of the frames should be reliable under the load

Fapp = FNr μ + FNf μ + mg sin θ (6) corresponding to a weight of 70 kg man. Frame of the mobile
robot has been manufactured from 6061 series of Aluminum
profiles. Two sliding slots have been manufactured inside the
CAD models of the wheeled, tracked and wheeled + tracked
profile. As shown in Figure 1, the beams, which are fastened
versions of CoMoRAT are given in Figures 3-4-5,
to the sliding slots, can be easily moved. By this way, internal
respectively. The mathematical model of the mobile robot
space of the platform can be rearranged and used effectively
platform has been simulated in Matlab. In the design process,
whenever it is required. Linear sliding slots are designed at the
SolidWorks is used. The dimensions of the parts have been
outside of the base profile. In order to connect the motor
optimized using the CosmosWorks finite element analysis
mounting to the frame and to adjust the tightness of the
(FEA) software.
pallets, these slots can be used. Moreover, using these slots,
the driving type can be rapidly configured within a few
minutes.

Fig. 3 CAD model of wheeled model of CoMoRAT

Fig. 6 Exploded view of the robot’s configurable driving systems

Wheels and pulley wheels have been connected to the

frame by two shafts. Dimensions of the shafts, shaft hole of
the wheels and pulley wheels have been designed in a way
that they get along well with each other. In other words, inner
and outer shafts can be used for wheels or pulley wheels. The
stages of the assembly are given in Figure 6 where the two
shafts are joined with a set screw and the shafts are connected
with wheels and pulley wheels by keys.

Fig. 4 CAD model of tracked model of CoMoRAT

Fig. 7 Manufactured reconfigurable mobile platform

Figure 7 shows the manufactured wheeled + tracked

version of the mobile robot. Tracked and wheeled versions are
given in Figure 8 and 9, respectively. The physical parameters
of the mobile robot shown in Figure 2 and 10 are Drw = Dfw =
Fig. 5 CAD model of wheeled + tracked model of CoMoRAT 200 mm, Lw = 347 mm, Lcg = 173.5 mm, h = hmg = 100 mm,
μw = 0.5, γf = γr = 1. DC motors which drive the mobile
In the strength analysis of the components, it is aimed that platform have an operating voltage range of 12 to 24 V,

Issue 4, Volume 3, 2009 370

 INTERNATIONAL JOURNAL OF MATHEMATICAL MODELS AND METHODS IN APPLIED SCIENCES

maximum speed of 72 rpm (0.35 A) at no load and speed of simulation results have been figured out in Table 2 in order to
36 rpm with 54.6 kg.cm torque at maximum efficiency. The make a comparison. Tipping over angles (up and side motion)
motors give 40 Watt output power at maximum efficiency. of the configurations has been also obtained during those
They draw 6 A current and produce 240 kg.cm torque at stall. experiments. Wheeled configuration has been failed for
The details of the model and the design parameters are given climbing more than 190 of slope angle. The mobile platform
in [11]. with wheeled + tracked configuration has climbed through 280
of slope angle. This result gives a nearby result that has been
obtained by the only tracked version. Hence a brief conclusion
can be made for this experiment that wheeled configuration
can be more powerful if it is combined with tracks. Wheeled +
tracked version of the mobile platform has designated an
advantage that it is more successful for climbing a slope than
wheeled one. As a result of this test it, can be concluded that
tracked or tracked + wheeled mobile robots can be surely used
in an environment having the features of mixed terrain and
slope, rather than using wheeled robot.

Table 2. Experimental and simulation results for observing

capability of slope climbing in degrees (T denotes tracked and
W denotes wheeled mobile platform)
Fig. 8 Tracked mobile robot
Tip Over Tip Over
Exp. Sim. Angle Angle
(Up) (Side)
T 31 34 60 70
W 19 20 60 70
T+W 28 31 62 72

Mobile platform with its three configurations has been

tested for velocity analysis. In order to show the performance
of the configurations, no load speed of the mobile platform
has been set to 1 m/s. Then, they have been tested on an
Fig. 9 Wheeled mobile robot indoor flat surface and on an outdoor surface composed of
small size gravel and worn concrete. Configurations have
As mentioned in this section, the mobile robot platform has been tested for a short path. While the mobile platform is
been designed for flat, rough and uneven terrain applications. traversing the path, the average velocity is obtained. The
Three driving systems have been merged into one platform. experimental results given in Table 3 show that the velocity of
By this way mobile robot can be used for different the tracked version of the mobile platform can be increased by
applications with proper configuration. adding wheels.

IV. EXPERIMENTS CONDUCTED WITH COMORAT Table 3. Velocity (m/s) of different mobile platform
configurations
CoMoRAT gives the opportunity for various applications
performed in both indoor and outdoor environments. In order Tracked +
Wheeled Tracked
to show the usability of the platform, some experiments have Wheeled
been conducted and the results are presented in this section. Indoor 0.8 0.6 0.7
These experiments are slope climbing, obstacle crossing over Outdoor 0.7 0.5 0.6
and power consumption. Moreover, tip over angles for the up
and side motions have been obtained during the slope
B. Crossing Over an Obstacle
experiments. All the experiment outcomes have been
compared with the simulation results. In order to analyze obstacle crossing over capabilities of
CoMoRAT, a mathematical model is derived and a scenario is
A. Slope Climbing considered. In this scenario, robot makes an effort to cross
Mobile platform with its three configurations has been over an obstacle that is placed just in front of it (Figure 10).
tested in the mixed terrain that is composed of small size Images from the experiments are also given in Figure 11.
gravel and worn concrete. The surface on which experiments
have been conducted has also inclination. Experiment and

Issue 4, Volume 3, 2009 371

 INTERNATIONAL JOURNAL OF MATHEMATICAL MODELS AND METHODS IN APPLIED SCIENCES

2mgL w − L cg
FNf = (8)
(L w + (D fw / 2) sin β)(cos β + μγ f sin β)

CoMoRAT's different configurations have been tested with

various obstacles having different heights. The experiment
results are given in Table 4 (+ and - signs indicate successful
and unsuccessful results about crossing over an obstacle
respectively). Note that radiuses of the wheels and pulley
wheels of the robot platform are 10 cm and 9 cm, respectively.
Tracked configuration of the robot platform has been
successful for the five different obstacles that heights are
ranging from 6.4 cm to 13 cm. Tracked + wheeled
configuration has managed to cross over all the obstacles
except the last one having 13 cm height. Tracked and tracked
+ wheeled configurations of the platform give acceptable
Fig. 10 Motion model of the mobile robot platform for crossing
over an obstacle
results. However, wheeled configuration is successful for only
two obstacles having heights of 6.4 cm and 8 cm. From these
experiments, tracked and tracked + wheeled versions have
shown their dominancy about crossing over an obstacle.

Table 4. Experiment results for crossing over obstacles

Height of
6.4 8 9 11 13
Obstacle (cm)
Tracked + + + + +
Wheeled + + - - -

Tracked +
+ + + + -
Wheeled

Fig. 11 Crossing over a step

For the obstacle crossing over experiments, CoMoRAT has
been tested with the obstacles in two cases. In the first case,
Normal forces and total tractive efforts required for one wheel of the platform has blocked with the obstacle. In
crossing over an obstacle obtained from the contact points the second case two wheels have blocked with the obstacle.
between wheel(s) and obstacle are specified in Figure 10. Drw Experiment and simulation results are given in Table 5.
and Dfw denote the rear and front wheel diameter,
respectively. μ denotes the coefficient of friction between the Table 5. Experimental and simulation results for crossing over
wheel/track and obstacle. h denotes the height of center of obstacle
mass about plane. hstep denotes the height of the obstacle. ГRw
and ГFw indicate the required torques for rear and front wheel,
respectively. β indicates the contact angle between the One Wheel Contact Two Wheel Contact
wheel/track and the obstacle. Normal and tractive forces are Analytical Exp. Analytical Exp.
obtained for one wheel and two wheel blocked with an Results Results Results Results
obstacle. Normal forces for the rear and front wheels in case (cm) (cm) (cm) (cm)
of two wheel contact are given in Equation (7-8).The details
of the mathematical model can be found in [11]. Note that the T 9 8 7 6.4
shape of the obstacles used in these experiments is W 18 13 11 9
rectangular. T+ W 14 11 9.5 7

2mg (L cg + R sin β )
FNr = (7)
L w + (D fw / 2 )sin β
C. Power Consumption
Three different configurations of CoMoRAT have been
tested in indoor and outdoor environments for measuring the

Issue 4, Volume 3, 2009 372

 INTERNATIONAL JOURNAL OF MATHEMATICAL MODELS AND METHODS IN APPLIED SCIENCES

current drawn for a certain traversed distance. During tests the [6] iRobot Magellan Pro Compact Mobile Robot User’s Guide, 2000.
[7] iRobot PackBot Tactical Mobile Robot Platform User’s Guide, 2007.
voltage of the motors is set to 24V and three different [8] M. Gifford, C., “Review of Selected Mobile Robot and Robotic
configurations have been tested for 10 m long straight path. Manipulator Technologies”, Center of Remote Sensing of Ice Sheets,
The indoor test area is smooth flat surface. On the other hand, Technical Report CReSIS TR 101, University of Kansas, 2006.
[9] TALON Mobile Robot, Foster-Miller, http://www.fostermiller.com,
surface of the outdoor test area is composed of small size
Cited 20 October 2009.
gravel and worn concrete. The results are listed in Table 6. [10] The International Unmanned Vehicle Systems Information Source,
http://www.uvs-info.com, Cited 20 October 2009.
Table 6. Power consumption experiments [11] Bayar, G., “Configurable Robot Base Design for Mixed Terrain
Applications”, MS Thesis, METU, 2005.
[12] Y. Maddahi, “Design and Laboratory Tests of Wheeled Mobile Robots”,
Current 7th WSEAS Int. Conf. on Mathematical Methods and Computational
drawn Time Techniques In Electrical Eng. (MMACTEE 2005). Bulgaria - Sofia. Oct.
2005.
during traveled (s)
[13] A. Segovia, M. Garduno, A. Gonzalez, “4 Steel-Robot : A Climbing
motion (A) Mobile Robot for Gas Containers Inspection”, 3rd WSEAS/IASME Int.
Indoor 1.1-1.3 16.6 Conf. On Dynamical Systems and Control (Control’07), Arcachon,
Tracked France, October 13-15, 2007.
Outdoor 1.4-2.2 20
[14] F. Yanqiong, S. Libo, “Design and Analysis of Modular Mobile Robot
Indoor 0.9-1.1 12.5 with Magnetic Wheels”, WSEAS Trans. On Applied and Theoretical
Wheeled
Outdoor 1.1-1.5 14.5 Mechanics, Issue 12, Vol. 3, December 2008.
Tracked Indoor 1.3-1.4 14.7 [15] R. Siegwart, P. Lamon, T. Estier, M. Lauria, R. Piguet, “Innovative
+ Design for Wheeled Locomotion in Rough Terrain”, Journal of Robotics
and Autonomous Systems, 40:151-162, 2002.
Wheeled Outdoor 1.5-2.5 17
[16] R. Siegwart, “Robox at Expo.02: A large Scale Installation of Personal
Robots”, Journal of Robotics and Autonomous Systems, 42:203-222,
Power consumptions of three configurations can be 2003.
[17] L. Song, Y. Fei, T. Lu, “Study on Effects of Nonholonomic Constraints
evaluated using Table 6. Results indicate that wheeled on Dynamics of a New Developed Quadruped Leg-Wheeled Passive
configuration draws the lowest current and it has the highest Mobile Robot”, WSEAS Trans. On Systems, Vol. 8, 2009.
speed.

V. CONCLUSION
While developing diverse applications for mobile robots, a
researcher inevitably feels the need for a configurable robot
that can efficiently operate both indoors and outdoors, onto
which various hardware can easily be installed and removed.
CoMoRAT is developed with this perspective, to be a small
size mobile robot suitable for academic research running on a
limited budget. The traction system can be configured to use
wheels, tracks or both. Body of the robot provides a closure
for batteries, motor drivers and main computer. Additional
hardware can easily be installed both inside and outside of the
robot. Main frame is constructed from Aluminum profile
which has groves to facilitate addition of hardware easily.
Field tests conducted on the robot are in agreement with the
simulations run prior to manufacturing the robot. As a result, a
configurable robot platform is designed, manufactured and
tested and CoMoRAT proved to meet the expectations as a
versatile mobile robot platform to be used in academic
research.

REFERENCES
[1] UNECE/IFR issues its 2005 World Robotics
survey,http://www.unece.org/press/pr2005/05statp03e.pdf, Cited 20
October 2009.
[2] UMass Lowell Robotics Lab, http://robotics.cs.uml.edu/robots, Cited 20
October 2009.
[3] R. Siegwart, I.R. Nourbakhsh: “Autonomous Mobile Robots”,
Cambridge MIT Press, Londan 2004.
[4] Nasa Space Telerobotics Program, http://www.nasa.gov, Cited 20
October 2009.
[5] Boersch, I., Heinsohn, J., Loose, H., 49. Internationales
Wissenschafliches Kollognium Teshnische Universitat Iımenau, 2004.

Issue 4, Volume 3, 2009 373