B 18 Final Evaluation Report

See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/290390411
ALL-TERRAIN MOBILE ROBOT FOR EXTRA-TERRESTRIAL APPLICATIONS
Thesis · May 2013

DOI: 10.13140/RG.2.1.2331.1449
CITATION READS
1 5,141
1 author:
Shivesh Kumar
Chalmers University of Technology
70 PUBLICATIONS 479 CITATIONS
SEE PROFILE
All content following this page was uploaded by Shivesh Kumar on 13 January 2016.
The user has requested enhancement of the downloaded file.

B18 Major Project 2012-13 / Mid-Term Evaluation Nov 2012
ALL-TERRAIN MOBILE ROBOT FOR

EXTRA-TERRESTRIAL APPLICATIONS
Final Report
of Major Project
Submitted in partial fulfillment of the requirements for the degree of
BACHELOR OF TECHNOLOGY
in
MECHANICAL ENGINEERING
by
MIHIR R BHAGAT SHIVESH KUMAR RAGHAVENDRA S

09M201 09M251 09M229
PANKAJ CHOUDHARY JAYANT ILME PRIYADARSHINI S

09M215 09M158 09M223
under the guidance of
Dr. K V GANGADHARAN
Professor, Dept. of Mech. Engg., NITK Surathkal
DEPARTMENT OF MECHANICAL ENGINEERING

NATIONAL INSTITUTE OF TECHNOLOGY KARNATAKA
SURATHKAL, MANGALORE
November, 2012
8
B18 Major Project 2012-13 / Final Evaluation May 2013
ACKNOWLEDGEMENT
This work is the result of efforts from various sources. We take this opportunity to express
our gratitude.
Firstly, we would like to thank the Department of Mechanical Engineering, National

Institute of Technology Karnataka for including this enriching opportunity as a part of the
degree requirements. The experience has been fruitful and our interests are developing.
We would like to thank our mentor and guide, Dr. K V Gangadharan, Professor in
Mechanical Engineering at NITK Surathkal. His zest and dynamism has propelled us to
work hard and long to achieve our goals. Truly, a great motivation.
Major thanks are in due to Mr. B. Sridhar, Director of Function Dynamics India Pvt. Ltd.
for issuing provisional licenses of RecurDyn for our project.
Special thanks to Robotics Club of NITK Surathkal for providing us with components for
making our prototype.
We would like to thank our classmates, friends and colleagues for their various ideas and
views during the entire length of the project till now.
Lastly, we would like to thank our families for their support and understanding.
2
ABSTRACT
With the recent success of several space missions, many public and private organizations
have invested heavily on developing technology for Extra-terrestrial exploration. It is just a
matter of few years before technology is able to meet our demands for frequent manned
and unmanned missions to bodies like Moon and Mars. This project aims at developing a
mobile robot (rover) with intended applications on Mars to help in exploration once
temporary base stations are set-up. This project involves designing, building and testing an
off-road, stand alone, off-the-grid, semi-autonomous mobile rover. Work for odd-semester
involved gathering background information regarding the mission requirements and
difficulties; learning on building wireless robots (and building a working model); doing
literature survey on existing rovers and their locomotion mechanisms; identifying the best
choice amongst it using methods of product development; and finally modeling and
simulating the mechanism using the RecurDyn software. Work in the even semester
included building the actual rover and testing its capabilities. In the process, we also
validated the simulation results from RecurDyn. Thus, this project was able to deliver on
two levels: (1) developing a prototype rover for extra-terrestrial applications and (2)
demonstrate a potential application of the RecurDyn software for use in space robotics.
3
CONTENT
I List of Tables . . . . . . . . . . . . . . . . . . . . . . . 5
II List of Figures . . . . . . . . . . . . . . . . . . . . . . 6
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . 8
2. Literature Survey. . . . . . . . . . . . . . . . . . . . . . 9
2.1 Set A.. . . . . . . . . . . . . . . . . . . . . . . . 13
2.1.1 Rocker Bogie (and its variants) . . . . . . . . . . . . . 13
2.1.2 4 wheel drive rover Zoë . . . . . . . . . . . . . . . 17
2.1.3 SpaceCat . . . . . . . . . . . . . . . . . . . . 18
2.1.4 Concept Screening Matrix . . . . . . . . . . . . . . . 20
2.2 Set B . . . . . . . . . . . . . . . . . . . . . . . 21
2.2.1 ATHLETE . . . . . . . . . . . . . . . . . . . . 21
2.2.2 STriDER . . . . . . . . . . . . . . . . . . . . 23
2.2.3 IMPASS . . . . . . . . . . . . . . . . . . . . 24
2.3 Set C . . . . . . . . . . . . . . . . . . . . . . . 26
2.3.1 Shrimp . . . . . . . . . . . . . . . . . . . . . 26
2.3.2 AMBLER
. . . . . . . . . . . . . . . . . . . . . . 27
2.3.3 GMD Snake . . . . . . . . . . . . . . . . . . . 28
2.4 Set D . . . . . . . . . . . . . . . . . . . . . . . 30
2.4.1 Chaos . . . . . . . . . . . . . . . . . . . . . 30
2.4.2 Oregon State University Rover 2011 . . . . . . . . . . . . . 31
2.5 Concept Scoring Matrix . . . . . . . . . . . . . . . . . 33
3. Realistic modeling and dynamic simulation of Shrimp rover in RecurDyn . . . 34
3.1 Realistic Modeling of Shrimp rover in RecurDyn . . . . . . . . . 35
3.2 Dynamic Simulation of Shrimp rover in RecurDyn. . . . . . . . . . 37
3.2.1 Ability to climb steps . . . . . . . . . . . . . . . . 37
3.2.2 Ability to climb an inclined surface . . . . . . . . . . . . 40
3.2.3 Ability to adapt passively with concave and convex surfaces . . . . 42
4. Shrimp rover manufacturing . . . . . . . . . . . . . . . . . . 43
4.1 Main body . . . . . . . . . . . . . . . . . . . . . . 43
4.2 Parallel bogie . . . . . . . . . . . . . . . . . . . . . 43
4.3 Front and rear.forks
. . . . . . . . . . . . . . . . . . . . 44
4.4 Electronics sub-system
. . . . . . . . . . . . . . . . . . . . . 44
5. Experimental Setup . . . . . . . . . . . . . . . . . . . . . 47
6. Results and discussions of the tests . . . . . . . . . . . . . . . . 49
7. Results of the project. . . . . . . . . . . . . . . . . . . . . 51
8. References . . . . . . . . . . . . . . . . . . . . . . . 52
48
LIST OF TABLES
Table 1. Concept Screening Matrix for Set A . . . . . . . . . . . . . 20

Table 2. Concept Screening Matrix for Set B . . . . . . . . . . . . . 25
Table 3. Concept Screening Matrix for Set.. C . . . . . . . . . . . . . 29
Table 4. Concept Screening Matrix for Set D . . . . . . . . . . . . . 32
Table 5. Concept Scoring Matrix . . . . . . . . . . . . . . . . . 33
Table 6. Joint definition of Shrimp rover . . . . . . . . . . . . . . . 36
Table 7. Maximum driving torque at each wheel . . . . . . . . . . . . 43
5
LIST OF FIGURES
Figure 1: Sojourner Rover using Rocker Bogie Mechanism . . . . . . . . . 13

Figure 2: Rocker Bogie Mechanism . . . . . . . . . . . . . . . . 13
Figure 3: Side View . . . . . . . . . . . . . . . . . . . . . 13
Figure 4: Top View . . . . . . . . . . . . . . . . . . . . . . 13
Figure 5: Terrain Adaptability . . . . . . . . . . . . . . . . . . 14
Figure 6: Deployed configuration of the suspension system of Sojourner . . . . 14
Figure 7: Changing Direction . . . . . . . . . . . . . . . . . . 15
Figure 8: Variants of Rocker Bogie mechanism . . . . . . . . . . . . . 16
Figure 9: Zoë Rover . . . . . . . . . . . . . . . . . . . . . 17
Figure 10: Zoë Chassis Configuration . . . . . . . . . . . . . . . . 17
Figure 11: Zoë Steering . . . . . . . . . . . . . . . . . . . . 17
Figure 12: SpaceCat rover . . . . . . . . . . . . . . . . . . . 18
. .
Figure 13: Locomotion Concept - Stepping on an object . . . . . . . . . 18
Figure 14: Locomotion Concept - Turning . . . . . . . . . . . . . . 18
Figure 15: Recovery after flipping over . . . . . . . . . . . . . . . 19
Figure 16: ATHLETE . . . . . . . . . . . . . . . . . . . . . 21
Figure 17: Design of each limb . . . . . . . . . . . . . . . . . . 21
Figure 18: Basic.Joint
. Design
. . . . . . . . . . . . . . . . . . . . . 22
Figure 19: STriDER . . . . . . . . . . . . . . . . . . . . . 23
Figure 20: IMPASS. . . . . . . . . . . . . . . . . . . . . . 24
Figure 21: IMPASS’ abilities . . . . . . . . . . . . . . . . . . 24
Figure 22: The Shrimp rover . . . . . . . . . . . . . . . . . . . 26
Figure 23: Steering . . . . . . . . . . . . . . . . . . . . . . 26
Figure 24: Coupling of wheels . . . . . . . . . . . . . . . . . . 26
Figure 25: Artists impression of AMBLER . . . . . . . . . . . . . 27
Figure 26: Limb of AMBLER . . . . . . . . . . . . . . . . . . 27
Figure 27: Snake rover . . . . . . . . . . . . . . . . . . . . 28
Figure 28: Schematic of the body . . . . . . . . . . . . . . . . . 28
Figure 29: Chaos . . . . . . . . . . . . . . . . . . . . . . 30
. . . .
Figure 30: Chaos’ variations . . . . . . . . . . . . . . . . . . 30
Figure 31: The OSU. rover
. . . . . . . . . . . . . . . . . . . . . 31
Figure 32: Chassis design. . . . . . . . . . . . . . . . . . . . 31
Figure 33: Steering arrangement . . . . . . . . . . . . . . . . . 31
Figure 34: Shrimp rover model imported to RecurDyn . . . . . . . . . . 35
Figure 35: Joint friction parameters . . . . . . . . . . . . . . . . 36
6
LIST OF FIGURES
Figure 36: Surface contact parameters . . . . . . . . . . . . . . . . 36

Figure 37: Shrimp rover climbing stairs
. . . . . . . . . . . . . . . . 37
Figure 38: Torque v/s Time graphs for all wheels while climbing step . . . . . . 39
Figure 39: Shrimp rover climbing inclined surface . . . . . . . . . . . . 40
Figure 40: Torque v/s Time graphs for all wheels while climbing inclination. . . . 42
Figure 41: Shrimp rover showing adaptability over concave surface . . . . . . 42
Figure 42: The rover during different stages of manufacturing . . . . . . . . 44
Figure 43: Schematic of control system of the rover . . . . . . . . . . . 45
Figure 44: The completed rover . . . . . . . . . . . . . . . . . . 46
Figure 45: Schematic for real time driving torque measurements . . . . . . . 47
Figure 46: Rover traversing terrain for experimental validation . . . . . . . . 48
Figure 47: Driving torque v/s time plots comparision between real & simulated data . 50
7
1. Introduction
With the recent success of several space missions, many public and private organizations
have invested heavily on developing technology for extra-terrestrial exploration. For
instance, the Americans, Russians and now also the Europeans have all had missions to
Mars, more or less successful. There are several risks and problems connected to such
missions, and this places high demands not just on the design but also the mission
equipment (Glette, 2004). Robotic rovers uniquely benefit planetary exploration - they
enable regional exploration with the precision of in-situ measurements, a combination
impossible from an orbiting spacecraft or fixed lander. (Tompkins, 2005). In the last two
decades, over 300 concepts have been generated and discussed (McTamaney, Douglas and
Harmon, 1989). Quite a few of these multi-functionality concepts have been designed and
developed for the deployment of such solo rovers. Characteristically, most designs now are
more or less identified, extensively analyzed and in a few cases even validated.
Sufficient rover designs exist for a solo mission of a rover to Mars, and hence, there has
been a shift in the trends now. Many organizations and universities look towards building a
robot ahead of its time: a rover for a time when humans have obtained / are in the process
of obtaining a permanent presence on Mars, and require robotic assistance to perform
remote tasks. These tasks can relate to astronaut safety, habitation-module maintenance,
and planetary exploration. Humans on Mars will want to limit their exposure to radiation,
hostile dust storms, and uncharted territory that may present physical risks; using a rover
provides a means to mitigate these risks. The project aims to create a wirelessly controlled
rover, able to traverse Mars-like terrain.
Our aim was to develop a prototype mobile robot with intended applications on Mars to
help in exploration once temporary base stations are set-up. Though similar in approach to
that of a solo-mission [ex: Sojourner (1996), Spirit & Opportunity (2003) and the more
recently, Curiosity (2012)], the difference in intended application brings about a different
set of challenges. This project involved designing and building an off-road, stand alone,
off-the-grid, semi-autonomous mobile rover whose life cycle runs into years, has superior
terrain handling capabilities, can be refueled / recharged and also undergo maintenance.
Through our project, we have explored possible designs for rover locomotion and
identified the best options available using product development methods. Shrimp, a passive
6-wheeled rover was determined to be the best rover. To understand its dynamics, we
simulated it using RecurDyn. A prototype rover was manufactured and its performance was
compared with the simulation results. The rover was equipped with adequate sensors and
auxiliary devices to primitively simulate its stand alone, off the grid, semi-autonomous
operation from a base station on the planet Mars.
Scope of the Project
1. To identify needs and associated problems of a Mars rover.

2. Study existing designs by Space Organizations, Universities and analyze them in
terms of utility, costs and feasibility.
3. Develop a design based on above analysis and simulate it.
4. Build a prototype rover with capabilities adjusted for earth.
5. Test the rover on key parameters; validate simulation results
8
2. Literature Survey
A mobile robot is an autonomous system capable of traversing a terrain with natural or
artificial obstacles. Its chassis is equipped with wheels, tracks or legs and possibly a
manipulator setup mounted on the chassis for handling of work pieces, tools or special
devices. Various preplanned operations are executed based on a pre-programmed
navigation strategy taking into account the current status of the environment.
The idea of sending a mobile robot to the surface of another planet is to allow earth bound
scientist’s access to specific areas of interest without enduring the harsh environments of
space. The robot carries instruments to various terrestrial formations for in-situ
experimentation. The goal of the rover is to move between areas of interest quickly and
safely. In order to better represent the planet of interest the rover must be able to travel tens
of kilometers.
Mobile robots can be classified by significant properties as:
1. Locomotion (Legged, wheeled, limbless, etc.)
2. Suspension (Rocker-bogie, independent, soft, etc.)
3. Steering (Skid, Ackerman, explicit)
4. Control Algorithm (Fully-Autonomous, semi-autonomous)
5. Body Flexibility (Uni-body, multi-body)
6. Usage Area (Rough Terrain, even surface, etc.)
7. Guidance and Navigation (Star field or Sun detection, GPS, sensor-based)
Design Challenges
1. Minimal weight and size
2. Reliable long term operation of all critical systems
3. Ability to survive and operate in abrasive and dusty environment
4. Safe utility in extreme radiation and thermal conditions
5. Efficient power utilization and transmission
6. Ability to maneuver in majority of terrains on Mars
7. Modular systems that can be utilized across other missions
Locomotion
Locomotion is a process which moves a rigid body. There is no doubt that a mobile robot’s
most important part is its locomotion system which determines the stability and capacity
while traversing on rough terrain. The difference of robotic locomotion is distinct from
traditional types in that it has to be more reliable without human interaction. While
constructing a robot, designer must have decided on the terrain requirements like stability
criteria, obstacle height, and surface friction. There is no only one exact solution while
comparing the mobility systems.
There are several types of locomotion mechanisms designed depending on nature of the
terrain. Locomotion systems can be divided into groups as; wheeled, tracked, legged
9
(walking robots), limbless (snake and serpentine robots) and hopping robots. Wheeled
rough terrain mobile robots are called as “Rover”.
In nature, insects are the fastest creatures, comparing to body/speed with their numerous
legs. There is no doubt that we are going to see legged robots more frequently in future,
with improved leg control algorithms and new lightweight materials. Limbless locomotion
is another terrain adaptive locomotion type for reptile creatures. Snakes can move very fast
on uneven terrain, additionally, they can easily climb on trees by their highly flexible body
structure.
Although animals and insects do not use wheels, wheeled locomotion has several
advantages for human-made machines. Rovers can carry more weight with high-speed
comparing to walking robots and snake robots. Another advantage of wheeled locomotion
is navigation. Wheeled robot’s position and orientation can be calculated more precisely
than tracked vehicles. Opposite to wheeled locomotion, legged locomotion needs complex
control algorithms for positioning.
Rovers designed for the exploration of other planets have had very complex mobility
systems using large numbers of wheels or legs and sometimes multiple bodies. Two
specific types of rovers have been to the surface of another planet: the Lunokhod rovers
using an eight wheel design and three Mars rovers using the six wheel rocker bogie
suspension. While the large number of wheels increases the stability over uneven terrain, it
also increases complexity in the design. Present day Mars rover suspension systems use six
wheels but require more than eight motors to drive them. Future rovers are also being
designed which use many wheels. New technology is being added to the rovers so that
when the drive train does fail the rover will remain mobile, though with reduced
capabilities.
Terrain / Environment
The various spacecraft that have landed on Mars provide sufficient evidence that its surface
is hard enough to support a small mobile vehicle. The images taken from the surface
indicate that it contains geological formations similar to places on Earth. Some have stated
that the areas resemble places in the deserts of Arizona and California where there is little
vegetation. The images from the Viking Landers in the 1970’s and the Pathfinder mission
in 1997 show rolling hills littered with rocks of various sizes. The twin MER rovers, which
are on opposite sides of the planet from each other, have landed in smooth dust covered
areas with an occasional impact crater. A long range rover may encounter these two
drastically different regions during its journey to the next science objective. The rover will
need high mobility features that allow it to pass through densely populated rock
outcroppings as well as efficiently make its way across vast dust covered plains (Cepolina,
1997), (Barlas, 2004).
A rover in natural terrain will encounter two types of obstacles; positive and negative.
Rocks that are above the ground plane are considered positive obstacles. Holes and craters
are examples of negative obstacles. Most rovers will stay clear of negative obstacles for
fear that it may get stuck or damaged from a fall more easily than hitting a positive
obstacle. There are two primary types of positive obstacles that a rover may come across;
bumps and steps. A bump is an obstacle that the rover can drive over a wheel at a time like
10
a rock shorter in length than the wheelbase of the rover. During the traversal of this type of
obstacle the remaining wheels maintain contact with the original ground plane. A step
obstacle will raise the entire vehicle to a new ground plane. As the rover traverses a step the
front wheels will remain on top of the obstacle once it has climbed it. The rear wheels will
then have to be pulled up. Of course the rover must be able to sense if the obstacle is
surmountable before attempting this.
Requirements
1. Suspension:
Wheeled locomotion’s main component is its suspension mechanism which connects
the wheels to the main body or platform. This connection can be in several ways like
springs, elastic rods or rigid mechanisms. Most of the heavy vehicles like trucks and
train wagons use leaf springs. For comfortable driving, cars use a complex spring,
damping and mechanism combination. Generally, exploration robots are driven on the
rough surface which consists of different sized stones and soft sand. For this reason,
car suspensions are not applicable for rovers. The requirements of a rover suspension
are:
1. As simple and lightweight as possible
2. Connections should be without spring to maintain equal traction force on
wheels
3. Distribute load equally to each wheel for most of the orientation possibilities to
prevent from slipping
2. Obstacle Climbing:
A rover’s obstacle limit generally compared with robot’s wheel size. In four wheel
drive off-road vehicles, limit is nearly half of their wheel diameter. It is possible to
pass over more than this height by pushing driving wheel to obstacle which can be
called as climbing. Step or stair climbing is the maximum limit of obstacles. The
contact point of wheel and obstacle is at the same height with wheel center for this
condition.
Field tests show that Mars mobile robots should be able to overcome at least 1.5
times height of its wheel diameter. This limitation narrows the mobile robot selection
alternatives and forces scientists to improve their current designs and study on new
rovers.
3. Control:
A passive system has very few control requirements than an active system. Legged
robots generally are active systems and each of their joints on each of their legs
requires to be controlled. A passive system on the other hand mirrors the contour of
the surface and thus requires minimal control.
Present technology and cost constraints, coupled with the erratic nature of martian
surface requires the system to be passive. However, if a robot could afford the
reliability, power and computational taxes levied by an active system, it obviously has
an upper hand.
4. Size:
The smaller the better. The lighter the better. Though the gravity of Mars is only 3/8th
11
of that on Earth and it makes things lighter, costs are still highly inhibitive to take
large and heavy machines from Earth. Also, the martian environment has soft sands
due to years of erosion and weathering. A heavy rover is more likely to have a wheel
sink into it.
5. Power:
Utilitarian power is scant on Mars. Solar power developed is dependent more on the
area of solar panels than on efforts to improve efficiency. Other sources are out of
reach and expensive, some even being hazardous. The rover, thus, requires to be
prudent with its activities.
6. Topple:
A toppled rover incapable of getting back is a wasted effort. The design should
account for instabilities and equip the rover to withstand a reasonable amount of pitch
and roll without toppling.
7. Complexity:
A complex design entails a complex manufacturing. It further involves even more
complex quality and validation checks. Besides, it may require regular servicing.
Hence, the intent remains to keep the design as simple as possible.
8. Payload:
The rover should be able to perform well while its payload is functional or not. Power
requirements may force the rover to stop moving while a particular function draws
majority of the power. However, higher the payload capability implies greater
functionality.
9. Speed:
Not the most important of factors, it can still bring about a major design change by
merely increasing speed requirements by 50%. It is essential to identify the speed
before hand and choose / design rover accordingly.
10. Cost:
Possibly, the single most important factor - a lot many ideas and designs get
abandoned due to lack of feasibility.
12
‘
B18At -80degCProject
Major and no load conditions,
2012-13 the geared
/ Final motors produce approximately 0.9RPM,
Evaluation
resulting in a vehicle which has a top speed of 0.4m/min at that temperature. At higher
May 2013
temperatures and similar no load conditions, greater speeds are possible. During steering,
the top speed is 7deg/sec.
The rover is 65cm in length, 48cm wide and 30cm tall in its deployed configuration
2.1(neglecting
Set
launch A the cruise-to-Mars phase of the MPF mission. In this stowed
the height of the UHF antenna). The rover is stowed on a lander petal for
and during
configuration, the rover height is reduced to 19cm. In this configuration, the rover has begn
tested and shown to withstand static loads of 66g, consistent (with margin) with the less
than 40g expected at impact upon landing on Mars. At deployment, the lander fires cable
2.1.1 Rocker bogie (and its variants)
cutting pyres, releasing tiedowns which restrain the rover to the stowed configuration.
Under command, the rover drives its wheels, locking the bogeys and deploying the antenna
so that the deployed configuration is achieved.
In mass
Inall, 4 vehicle
rovers
the deployed
on the have
arrangedlanded
configuration,
has been
the rover has ground on Mars
clearance
so that the center of mass
of 15cm.successfully
The distribution of
is nearly at the center of to explore its surface (Sojourner, 1996;
Spirit,1/3
the body (the of
Warm the
Opportunity,
consequence,
weight
Electronics Box (WEB))
2003;
the vehicle could withstand
of and the
at a rover.
height
a tilt of Curiosity,
at the base Together
of the WEB. As a the rear four wheels have enough traction
45deg in any direction2012.)
without over-They were developed at NASA and they all
turning, although fault protection limits prevent the vehicle from exceeding tilts of 35deg
have a six wheel rocker bogie suspension system invented by Donald Bickler.
during traverses.
Figure 1: Mars Pathfinder Microrover Flight Experiment (MFEX)
to keep the rover from slipping [10].
II------ UHF Antenna
Solar Panel II
APXS
\
A
b
II /7-
Carnerasl
Lasers
r-or-lb
65 cm ——
Figure 1: Sojourner Rover with Rocker Bogie. Courtesy: NASA Figure 2: Rocker Bogie Mechanism. Courtesy: NASA
(a) Pantograph (b) Rocker-Bogie
Each side of the suspension has two links, a main rocker and a forward bogie. A wheel and
steering mechanism
Figure 1.5:is Link
attached
styletomobility
one end systems
of the main rocker.
(images The term from
reproduced “rocker” comes
NASA)
from the rocking aspect of the larger links on each side of the suspension system. These
rockers are connected to each other and the vehicle chassis through a differential. Relative
to the chassis, when one rocker goes up, the other goes down. The chassis maintains the
average pitch angle of both rockers.
Each side of the suspension has two links, a main rocker and a forward bogie.
The opposite end of the rocker is connected to the forward bogie through a passive pivot
joint. A steering mechanism is attached to each end of the forward bogie with the pivot
mounted in-between.
A wheel The twomechanism
and steering sides of theissuspension
attached are
to connected to the
one end of a single
mainbody from aThe
rocker.
point on each main rocker. The length of the rockers and bogies and the position of each
joint are defined to distribute the weight of the body on the wheels with the lowest normal
opposite end is connected to the forward bogie through a passive pivot joint. A
force acting on the front pair.
steering mechanism is attached to each end of the forward bogie with the pivot
mounted in-between. The two sides of the suspension are connected to a single
body from a point on each main rocker. The length of the rockers and bogies and
the position ofSide

Figure 3: each
View joint are defined to distribute
Figure the
4: Top weight
View of the body on the
The body of rover is kept stable at the average angle between both sides of the suspension
wheels with the lowest normal force acting on the front pair. With more normal force
with a differential linkage. The linkage is connected to both main rockers and pinned at the
12
13
center on the back of the body. It assures that all six wheels have relatively constant loads
on them at all times which is a major advantage of an unsprung suspension system.
Types Of Joints Figure 3. Stowed MER Rover on Lander Basepetal
Robots using rocker bogie mechanism make use

Folding and Unfolding the MER Mobility
of a suspension mechanism that consists of
Determination
several rigid elements connected of how the
through rocker-bogie suspension could be “broken” to enable it to stow within the
joints
allotted space was the first challenge faced in the design of the sub-system. As the name would suggest,
of a certain number of degrees
the two of freedom
primary (DoF)
components of this type of suspension are the rocker and bogie (see Figure 4). These
two structural elements are connected via a free rotating pivot dubbed the Bogie Pivot. The right and left
resulting in a structure that
setshas one system
of rocker-bogie DoF. are connected to each other via the vehicle’s differential, a passive,
assemblies
This enables them to move along uneven terrain
motion-reversal joint that constrains the two sides of the mobility system to equal and opposite motion.
Three unique break points were selected, the Rocker-Bridge Joint, a mid-span rocker folding joint; the
without losing contact with the Deployment
Rocker ground. Actuator (RDA) Joint, a motor driven deployment joint on the forward rocker arm, and
a telescoping prismatic joint on the bogie member. Thus, a total of six joints must be reliably locked and
Figure 5: Terrain adaptability
The suspension has 6 wheels
latchedwith
into place during deployment to provide the rover with a safe and stable platform for driving.
symmetric structure for both

sides. Each side has 3 wheels
which are connected to each other
with two links. The main linkage,
rocker has 2 joints. while first
joint connected to front wheel,
another joint assembled to bogie,
which is similar to train wagon Rocker DeplGyTIent
suspension member. (Stone, 1996) Actuator (RDA)
Wheel Strut Rocker ____-

/
The length of the rockers and W e e l Strut
bogies and the position of each

4.Figure
MER6:Suspension
joint are defined to distribute the Figure Sojourner. Deployed configuration of the suspension
Nomenclature (Deployedsystem of
Configuration)
Extra joints and kinks were needed to make it stow in
weight of the body on the wheels the space craft. Courtesy: NASA
with the lowest normal force 3

acting on the front pair. With more
normal force on the rear wheels there is more traction to push the front pair over an
obstacle. Unfortunately this works when the rover is moving in the forward direction only.
There is a possibility that the rover may drive into an area that it can not back out of. A
closer look at the MER rover will show that the suspension is on backwards so that the
rover can back out of anything it drives into. (Shirley, 1995)
Mobility
The primary advantage of the of the rocker-bogie mechanism is that it provides the rover
with a mobility system that has the kinematic range to permit the rover to safely traverse
obstacles higher then the size of the wheel diameter. It also allow for rover ‘arc turn’ and
‘turn-in-place’ maneuvers.
A rover is considered to have a high degree of mobility in natural terrain if it can surmount
obstacles that are large in comparison to the size of its wheels. A rover must have enough
traction from its rear wheels to push the front wheels against an obstacle with enough force
so that they can climb up it. Typically a four wheeled rover can not climb obstacles larger
than a wheel radius because the rear wheels do not have enough traction. Without traction
the wheels will slip and there will not be enough forward thrust to keep the front wheels in
14
contact with the obstacle. The rocker bogie suspension can surmount obstacles head on that
are larger than a wheel diameter because it uses an extra set of wheels to provide more
forward thrust. The extra wheels also reduces the normal force on each wheel by about 1/6
the weight of the rover. Less forward thrust is required because the front wheels only have
to lift 1/3 of the weight of the rover. Together the rear four wheels have enough traction to
keep the rover from slipping.
The rocker bogie suspension is capable of a high degree of mobility. It has a ground
clearance larger than a wheel diameter, unlike articulated body vehicles. The single rigid
body is more stable for sensor mounting and thermal control. The suspension mechanisms
and joints are above the wheels reducing the chances that the rover will get caught on an
obstacle. It can also perform multiple types of steering: Ackerman, Differential, Zero
Radius, and Crabbing.
Crabbing Zero Radius Ackerman Differential

(a) Crabbing (b)Figure7:
Zero radius
Changing Direction.(c) Ackerman
Courtesy: NASA (d) Diﬀerential/Skid
Figure 1.7: Changing direction

This mobility system requires that each wheel be driven by a separate motor and steering
mechanism, increasing the overall complexity. Rovers that use the rocker bogie suspension
This mobility system requires that each wheel be driven by a separate motor
can have 10 or 12 motors just for mobility all of which are exposed to the environment
including the drive train. Harmonic drives coupled to the motors are used to increase torque
rather than planetary
and steering or spur gear
mechanism, boxes because
increasing they complexity.
the overall save space andRovers
weight.that use the
Variants
rocker bogie suspension can have 10 or 12 motors just for mobility all of which are
There are a few different configurations of the rocker bogie but they all have six wheels
connected by four links (Roman, 2005). The series of “rocky” rovers was used to identify
exposed
what to the environment
configuration would workincluding the drive
best in Martian train. Harmonic
conditions. drives Rocky
The first Rocky, coupled3, to
and
Rock 8 use a gear differential between the two suspension sides. Sojourner, Rocky 4, and
Rocky 7 use are
the motors an external
used to linkage
increasedifferential to free
torque rather up planetary
than space insideor the
spurbody
gear[29].
boxesThe
suspension geometry of Rocky 7 is modified by moving the middle wheels forward and
eliminating the steering mechanisms on the front pair of wheels. This design reduces the
becauseofthey
number saveneeded
motors space and weight. from
for mobility During
tenoperation
to eight, athey have
motor high static
to drive friction
each wheel and
two for steering the rear pair. It was discovered that a rock can jam the tandem wheels
because
and canoflock
the short
up indistance between them.
cold temperatures which will overload the motors causing them
Rocky 8 also known as FIDO (Field Integrated Design and Operations) rover has a drive
motor in each wheel and has the ability to steer all six wheels independently. This gives
to fail prematurely. Sojourner had heating units on each motor to keep them within
FIDO the ability to perform a “crabbing” maneuver in which the rover can point all of the
wheels in the direction it would like to travel. Previous versions in the Rocky series can do
this
the as well but limits
operating the middle pairthat
in fear of wheels will scuff
the extreme coldacross
of thethe ground atmosphere
Martian because theymight
are not
steerable.
damage them [10].
15
Advantages
1. It can traverse obstacles more than the size of wheel diameter.
2. It is a passive mechanism.
3. All wheels are always on the ground. This provides equal traction.
4. It can perform 4 different types of steering: Ackerman, Differential, Zero Radius, and
Crabbing
Disadvantages
1. Its speed is very less (less than 0.1m/s).
2. It requires a differential mechanism for stability of the main body
3. It requires 10 motors - 6 for wheel and 4 for steering.
(a)Rocky
Rocky2 (b) Rocky
Rocky 33
(c) Rocky
Rocky 4 4 (d) Rocky
Rocky 7 7
(e) Rocky
Rocky 8 8
Figure 8: Variants of Rocker Bogie Mechanism. Courtesy: NASA
Figure 1.6: Rocky Rover series (images reproduced from NASA)

16
2.1.2 4 wheel drive rover - Zoë

Zoë, by the Robotics Institute, CMU, a solar powered bot meant to explore Atacama
Desert, Chile, developed as an alternative to 6 wheel drive, rocker-bogie rovers for extra-
terrestrial application. (Wagner et al, 2005), (Wettergreen et al, 2005)
Key Characteristics
1. It is a 4 wheel drive, with the wheels independently
driven by motors.
2. The chassis uses front and rear steering axles that are
articulated passively by driving wheel motors
differentially.
3. Higher maneuverability achieved by steering both
front and rear axis symmetrically. Figure 9: Zoë Rover. Courtesy: CMU
4. Uses an averaged chassis along the longitudinal axis,
thus distributing weight equally amongst all wheels, even when there is high roll.
5. During roll (case 3 in chassis configuration image), the chassis elements are made to
flex rather than having more number of attachments. This ensures fewer links and
This paper explores an alternative chassis design that is avoid or surmount obstacles and slopes without human
lowed
minimalist yet fully DOF
capable andofappropriate
such for attachments.
an intervention. Reliability is therefore a key metric.
autonomous, long-range Mars rover traveling at speeds Furthermore Zoë is built to survey large geological
6. Speeds
steering as high
axles that are articulated as by110cm/s
up to 100 cm/s. This chassis uses front and rear
passively driving achieved.
units, not to study specific rocks. Therefore it has
neither an instrument arm nor a need to make very tight
wheel motors differentially. An averaging mechanism turns.
7.
along The
the steering
chassis’ mechanism
roll axis distributes
weight equally amongst all four wheels. This
the vehicle’s requires traction, in absence of which the dual passive axle
The second motivation was minimizing Zoë’s power
configuration will not steer powered.
suspension system, passive like a rocker-bogie, also
improves obstacle climbing ability.
- as observed in most vehicles. This can lead to unreliable
consumption because the rover is entirely solar
This argues for a minimalist, low mass
We haveand built ainsufficient steering

this new conditions. It can be over come by two methods:
design with low part count and fewer degrees of
rover called Zoë to study freedom.
design. Zoë is a solar-powered rover designed to
1. Steering
explore the Mars-like landscapes ofactuators,
the Atacama like
Desert in Chile (see Fig. 1). Zoë traveled 50 km
those
Naturally theseused inconflict
motivations rocker-bogie.
with each other. The advantages are:
But the Zoë design does a good job at satisfying our
a. Steering even when
autonomously through the desert in a search for life
that was overseen by a team of remote scientists
traction is low.
goals of both reliably traversing rough desert terrain
while keeping power consumption low. Zoë uses four
b. Loss in momentum
operating as if the robot were on Mars [6].
driving due todifferentially,
varying causingvelocity
independently driven wheels. The vehicle steers by
its wheels its passive of Fig. 5: Analysis of the relationship between distance traveled and energ
motors while turning is
steering joints to pivot. Fig. 2 depicts the Zoë chassis
2. CHASSIS OVERVIEW reduced greatly. in several configurations. Zoë achieves higher6. KINEMATIC ANALYSIS 7. STEERING
maneuverability by steering both axles symmetrically,
2. Steering Control
Two competing factors motivated the design of the Zoë
chassis. First, Zoë was intended to traverse long by Feedback:
unlike our previous rover Hyperion [7] which had a
single steering axle. The body length identical for bothThere is a one-to-one mapping from the chassis’ Initial steering t
Thetakesystem isfielddesigned
distances over desert terrain at roughly human walking
speeds. Its travels would it beyond the of for
Hyperion anda symmetric
Zoë, but with axle(s) setsteering mechanism.
to 20 degrees,configuration This tends to shift
to the resulting path radius. The path
Hyperion travels an arc of about 5.5-m radius whileradius is determined by finding the point of intersection
towards incorrec
the right of Fig.
view of any panoramic image, so it must be prepared to
to incorrect chassis configuration. Also, the shiftofUnfortunately
Zoë travels an arc of about 2.7-m radius.
changes
lines extended based from on eachwhether an
steering axle.
many different steer axle angles can
obstacle climbing
axles is trailing or not. bring about a given arc path. To overcome th

loop feedback c
While steering actuators are necessary for ensuring that the
Ideally the axles
steering gets
angle of steered,
the front the
axle is symmetric Although propor
to that of the rear, having the same magnitude but the angles does redu
feedback control loop is necessary to ensure that the axles gets
opposite sign. stay in position.
In any other configuration, more range state error remai
of motion is needed to achieve the same path radius. A and the rear axle
8. Unlike 6 wheel drive, the steepest grade that Zoë can drive is limited
visual explanation can bebyfoundthe traction
in Fig. 6. On the of
side of the figure, the steer angles are symmetric and
left turns, the rear ax
against the electro
the wheels rather than by its drive torque capacity. When approaching
the robot is driving along anvertical steps,
arc with radius R . On
robot
the right side, the steer angles are not symmetric: the
2 We observed that
wheels are not sufficient to produce enough tractionfront to israise the while
under-steered rover. Hence,
the rear the
is over-steered. the magnitude o
Compared to the symmetric configuration, the commanded robo
rover has to approach the step diagonally and thus allow onlyconfiguration
asymmetric one wheel requiresto be raised
a wider range of the feedback loo
axle rotation is necessary to drive a given path. when Zoë is com
at a time, while three wheels produce the sufficient traction to move forward. degree steering ax
steer axle settles a
settles at about -7
is not due to an
frictions within s
leading or trailing
We measured thi
number of tests.
driving command
Fig. 2: Detailed design of Zoë chassis (top) and three chassis configurations: steering, driving straight, surmounting
Figure 10: Zoë Chassis configuration:
rough terrain (bottom, left to1. Steering 2. Neutral 3.
right).
Figure 11: 1. Symmetric steering 2. steering axle ang
Incorrect steering.
steering Courtesy:
angles (left)CMU constant for all t
Surmounting a rough terrain. Courtesy: CMU Fig. 6: Symmetric provide more
maneuverability for a limited range of steering axle drive the robot fo
motion than do other configurations (right). and straight. In al
are given in [9] b
17
2.1.3 SpaceCat
SpaceCat, an innovative micro-rover designed by the European Space Agency, which can
climb stairs and optimally balance rover in all situations by adjusting the centre of mass. It
has a unique stepping wheel design.but has capabilities comparable to 6WD rocker-bogie
without the the corresponding expense or complexity. Adjustable centre of mass capability
makes the rover an interesting subject to study. (Siewart et al, 1998), (Lauria et al, 1998)
the center of gravity of the rover is moved outside the contact surface formed by the four
Key characteristics
wheels. Thus the rover gets out of balance and ‘falls’ with its upper wheel onto the object.
Scientific Instruments
1. The rover is 30cm long, 20cm wide and 20cm high. It
CUI APX Camera MOS
has capabilities of overcoming obstacles as high or
wide (in case of a hole) as 10cm.
slave module H-Bridges
2. The robot has bothsensors
distance walking and• rolling capabilities -
low level tasks
• motor control 2 DC-motors
it allows for efficient
inclinometers
rolling over planeswithand stepping
• interface
over obstacles. robot sensors digital encoders
3. It has 2 independently
distance sensors
driven set of 3 wheels aredistance Figure 12: SpaceCat rover. Courtesy: ESA
sensors
joined by a frame. The wheels in a set are placed in
slave module master module slave module
triangular form, without
• low level tasks touching • data each
collectionother. This • low level tasks
forms the wheel frame.
• motor control
• interface with
• navigation • motor control
sliprings • path planing sliprings • interface with
4. The payload robot frame,
sensors
which can• interface
be independently
with
external world
designed,
robot sensors
consists of scientific
payload and itsH-Bridges
support. H-Bridges
tethered link
5. Each wheel is4 DC-motors
driven by an independent• motor. power Also, each wheel frame is capable of
4 DC-motors
• data
rotating independently
digital encoders
around the payload frame by a motor.
digital encoders
external world
6. The adjustable centre of mass is achieved (lander) by the relative motion between the wheel
and payload frame.
FIGURE 2. Electronics scheme of the micro-rover
CoG
CoG
1 2
Co
CoG G side view
Co
G
A B
3 CoG 4 CoG
CoG top view CoG
5 CoG 6 CoG
CoR
CoR
FIGURE 3. Left: Locomotion concept: Stepping on an object
FigureFor
13:simplicity
Locomotion concept.
the Figure
Center of Gravity (CoG) 14: Locomotion
is assumed concept:
to be equal to Turning.
that of the
Stepping
payloadon an object. (CoR = Centre of Rotation)
Courtesy:
Right: ESA
Locomotion concept: Turning A: turning
(CoR: on 3 wheels
Center B: turning on 4 wheels;
of Rotation)
A: turning on three wheels main load on front wheels.
Courtesy: ESA
B: turning on four wheels with main load on front wheels
4 Siegwart, Lauria, Mäusli, Van Winnendael
18
FIGURE 4. Real movements of the robot while stepping over an object
CoG CoG
CoG
Co
G
1 2 3 4
Co Co
G
top view G
Figure 15: Recovery after flipping over.

FIGURE 5. Recovery after flipping1. Flipped
over: rover. CG is well inside the central axis.
2. The CG is shifted by rotating the frame
1: initial position with CoG
3. Rover within
tries to central
balance itself axis
because of shift in CG
4. Rover back on its wheels in upright position.
2: movement of the CoG to its most outside position -> rover flips back on wheels
Courtesy: ESA
3: situation just after flipping back on wheels
4: situation after moving the CoG back to its initial position
4. CONTROL OF THE MICRO-ROVER
Control is assumed to be distributed over the rover, the lander and ground (Earth). Various ar-
chitectures and concepts, some of them more hierarchical [14], [15] and some of them more
behavior based [16], [17], have been proposed for space robots. Our approach fits within the
unified control architecture for planetary rovers which has been proposed in [18].
The degree of autonomy of a robot depends directly of the complexity of its local control. To-
tally autonomous mobile systems are very hard to build because of energy and navigation prob-
lems. In our case, a semi-autonomous robot is more adapted to fulfill the mission’s
requirements.
A crucial aspect of the whole system is the low transmission rate from Mars to Earth. A teleop-
erator should act only for path planning and in case of hazards. Regularly, all decisions are
made by the lander and/or the rovers. Local intelligence must therefore exceed the reflexive be-
haviors and provide problem solving for more complex situations. An essential skill of the rov-
5 Siegwart, Lauria, Mäusli, Van Winnendael
19
2.1.4 Concept Screening Matrix
Table 1: Concept Screening matrix for Set A
Parameter Rocker Bogie Zoë SpaceCat

Suspension + 0 -
Obstacle Climbing + 0 0
Control 0 0 0
Size 0 0 0
Power - 0 -
Topple + 0 +
Complexity - 0 -
Payload + 0 +
Speed 0 0 +
Cost - 0 -
Total 1 0 -1
Rank 1 2 3
20
2.2 Set B
2.2.1 ATHLETE
The ATHLETE vehicle system is a new form of two cooperative robotic vehicles that can
act individually or physically connect together through a structural pallet to transport and
manipulate cargo (Wilcox, 2008), (Wilcox, 2009).
The basis of the ATHLETE (All Terrain Hex Limed Extra Terrestrial Explorer) robot is the
wheel-on-limb vehicle concept. This hybrid mobility platform enables the vehicle to
traverse at high speeds across benign terrain, as well as enabling walking, by locking the
wheels and using them as feet, on extreme terrain. This vehicle architecture also allows for
manipulation since the vehicle is
stable on three or more wheels. Non-
adjacent limbs can be lifted and used
to interact with the environment. A
tool mechanism at the end of the
limb, attached to the wheel hub,
allows for interchangeable tools, such
as a gripper or an auger, to be used
for manipulation.
This unique vehicle design allows for
significant weight savings over a
traditional planetary roving vehicle
that must have large wheels to allow
Figure 16: ATHLETE. Courtesy: NASA
for low ground pressure as well as
high torque wheel actuators since
traditional vehicles cannot walk in extremely soft or steep terrain.
It has 6 identical limbs – each having 7 DOF. Each joint has a unique gear ratio and torque
capability, but basic joint design is the same. Every joint has a brushless DC motor with a
power-off holding brake and an incremental encoder. Low power motor is used in Hip
Yaw, Knee Roll, Ankle Pitch, and Ankle Roll Joints. High power motor is used in the Hip
Pitch, Thigh Pitch, and Knee Pitch joints.
Joint Sensing and Control
Overall Coordination of joints is achieved using a main CPU installed over the rover body.
Motor controllers are mounted on the motors locally. Joint positions are updated during
motion and are fed from the main CPU to the local controllers. The motor controllers then
update the position set-point and smoothly transition to the new motion profile. Each joint
has both an incremental encoder on the motor input as well as a 12 bit absolute encoder on
the joint output. The motion of each joint motor is controlled via the incremental encoder,
and the limb kinematics is determined from the high precision output encoder. Using these
two position sensing devices, the mechanical windup of the joint can be determined. This
measurement combined with a characterization of the torsional stiffness of the joint gives
the torque experienced at each joint. Using the joint torques and the vehicle’s kinematic
pose, the robot can determine the ground contact force of each limb and can autonomously
make adjustments to correctly distribute the vehicle load over all the limbs.
21
Figure 17: Design of each limb. Courtesy: NASA
Figure 18: Basic Joint Design. Courtesy: NASA
Advantages
1. Wheel – leg hybrid design makes the rover very robust and versatile for any terrain.
2. CG can be changed easily.
3. Modular design
4. Compliant and highly adaptive for any terrain.
5. No spring like suspension required.
6. End effector can also act as a tool holder and can be used for operations like drilling,
pick & place etc.
Disadvantages
1. High costs involved in even building a very basic prototype. (up to Rs. 5,00,000)
2. Achieving a realistic computer simulation for this rover design itself can take more
than 4-5 months.
3. Requires a time line greater than available for Major Project
4. Very complex design and dynamics
5. Active control system
22
2.2.2 STriDER
STriDER (Self-excited Tripedal Dynamic Experimental Robot) is a novel three-legged
walking machine that exploits the concept of actuated passive dynamic locomotion to
dynamically walk with high energy efficiency and minimal control. Unlike other passive
dynamic walking machines, this unique tripedal locomotion robot is inherently stable with
its tripod stance and can change directions while walking.
During a step, two legs act as stance legs while the other acts as a swing leg. The legs are
oriented to push the center of gravity outside of the stance legs to initiate a step. As the
body of the robot falls forward, the swing leg naturally swings in between the two stance
legs and catches the fall. The body also rotates 180 degrees, preventing the legs from
tangling up. Once all three legs are in contact with the ground, the robot regains its
stability and the posture of the robot is then reset in preparation for the next step. Gaits for
changing directions are implemented in a rather interesting way: by changing the sequence
of choice of the swing leg, the tripedal gait can move the robot in 60° interval directions
for each step (Ren et al, 2008), (Heaston et al, 2007).
Advantages
1. It has a simple kinematic structure which is inherently stable (like a camera tripod).
2. It is simple to control as the motion is a simple falling in a predetermined direction
and catching its fall.
3. Very energy efficient, exploiting the actuated passive dynamic locomotion concept
utilizing its built in dynamics.
4. Lightweight, thus, enabling it to be launched to difficult to access areas.
5. It is tall making it ideal for deploying and positioning sensors at high position.
Surveillance, for example.
Figure 19: STriDER. Courtesy: IEEE
23
2.2.3 IMPASS
IMPASS (Intelligent Mobility Platform with Active Spoke System) is a high mobility
locomotion platform for unmanned systems in unstructured environments. Utilizing
rimless wheels with individually actuated spokes, it
can follow the contour of uneven surfaces like
tracks and step over large obstacles like legged
vehicles while retaining the simplicity of wheels.
Since it lacks the complexity of legs and has a large
effective (wheel) diameter, this highly adaptive
system can move over extreme terrain with ease
while maintaining respectable travel speeds, and
thus has great potential for search-and-rescue
Figure 20: IMPASS. Courtesy: ASME
missions, scientific exploration, and anti-terror
response applications. It has a total of 9 DOF
(Laney and Hong, 2006), (Laney and Hong, 2005).
The kinematics of the robot is analyzed and the motion of the robot is simulated using two
actuated spoke wheels on flat terrain using a one-, two-, and three-point contact per wheel
scheme. It is shown that the one-point contact mode has two degrees of freedom and that
the motion output can be arbitrarily selected. This mode would allow for moving while
maintaining a constant height for the center of mass, which we have demonstrated by
simulation. Turning for this mode is shown to occur discretely by changing the heading
angle for every step by taking steps of different lengths with the right and left wheels. The
two-point contact mode is shown to have one degree of freedom, and that by choosing a
step length, the path of the center of the axle in the sagittal plane is determined as a
function of the wheel angle. This mode of locomotion allows for statically stable walking
with only two wheels, and could be used for carrying heavy payloads. The three-point
contact scheme is shown to have zero degrees of freedom, but would allow for additional
stability during stationary tasks by letting the robot assume a wide stance.
Climbing stairs upto 4 times its nominal height
Staying level and / or maintaining pitch and roll on all inclines
Terrain adaption and multiple points of contact for stability

Figure 21: IMPASS’ abilities. Courtesy: ASME
24
Table 2: Concept Screening matrix for Set B
Parameter ATHLETE IMPASS STriDER

Suspension + 0 -
Obstacle Climbing + 0 -
Control - 0 0
Size + 0 -
Power + 0 +
Topple + 0 -
Complexity - 0 +
Payload + 0 -
Speed + 0 -
Cost - 0 +
Total 4 0 -3
Rank 1 2 3
25
2.3 Set C
2.3.1 Shrimp
Shrimp Rover has six wheels that operate separately. It is similar to rocker bogies
mechanism in its capabilities. It has 4 wheels on a pair of parallel bogies. It has one front
and one rear wheel. Special design, flexible elbows, a spring fitted in the front elbow that
work as a pushing force, makes it possible for robot to adjust rough areas and obstacles
such that all six wheels touch the ground simultaneously. (Seigwart et al, 2002)
The rover’s front fork has three roles:
1. The spring makes it possible for wheels to touch the ground all the time.
2. When the robot encounters an obstacle, the horizontal force acting on the front wheel
creates a torque around the instantaneous rotating center of front wheel. The four bar
mechanism design in the front wheel shows that the instant center is set under the
horizontal line, and therefore causes the wheel to move up accordingly.
3. When the front wheel is going up, spring is compressed and energy will be stored in
the front wheel. Although, other wheels are not in a good condition during climbing
and they don't touch the ground completely, but this stored energy helps them move
up easier.
(a) (b)
Figure 22: (a) The Shrimp rover’s adaptability (b) Front fork. Courtesy: EPFL
In this robot, each of the six wheels have separate drivers. Also, the front and back wheels
have angle adjusting and controlling system. So for steering, there is appropriate speed
difference in the side wheels and angle of front and back wheels is adjusted. This steering
strategy increase the accuracy of robot maneuvers, and the robot can also turn in its place
with minimum slip.
Wheels are coupled so that the force distribution is the best as
possible. Spring and dimensions of the robot are designed in a
way that when it is standing on a planar surface, forces acted on
all six wheels are the same.
Figure 23: Steering. Figure 24: Coupling of wheels. Courtesy: EPFL

Courtesy: EPFL
26
2.3.2 AMBLER
Ambler, an acronym for Autonomous Mobile Exploratio
n, was developed by Carnegie Mellon University and
the Jet Propulsion Laboratory. Ambler’s configuration
consists of six legs stacked co-axially at shoulder joints.
Each leg is mounted at a different elevation from the
body and can be independently rotated around the full
body. Further, each leg consists of 2 revolute joints that
movie in a horizontal plane to move the leg and a
prismatic joint at the end of the leg affects the variable Figure 25: Artists impression of AMBLER.
Courtesy: CMU, JPL
telescopic motion to extend over and retract the bot.
Thus, it has 18 degrees of freedom. (Krotkov, 1995).
The AMBLER body consists of place to mount power generation devices, computing
devices and equipments for sampling. Amber's legged configuration overcomes three
significant liabilities of precedent walkers:
1. Complexity of coordination control
2. Resultant energy losses
3. Redundancy for continued function after loss of some motions
The Amber's actuator groups are orthogonal; the Ambler can thus level without propelling,
can propel without leveling. It exhibits no power coupling between the two. This
configuration enables a tractable control model and eliminates the energy loss of actuator
conflict.
In addition, the Ambler enables energy-efficient overlapping gaits unprecedented by
animals and other robot walkers. The Ambler incorporates true functional redundancy. It
can lose up to two legs and still walk. Other critical issues include perception and
locomotion of rugged terrain, self-assessment, safeguarding, gait planning, control, and
ultimate self-reliance.
Design employed by ambler can overcome very rough terrains, high slopes, but the
difficulties associated with it are lot more.
Advantages:
1.Can overcome very rough, high terrain.
2.Functional even after losing 1 - 2 limbs.
3.Easier to control than other legged rovers.
Disadvantages:
1.Its speed is very slow.
2.Design is very complex and includes lot of active joints.
3.Controlling and configuring it would be difficult.
Figure 26: Limb of AMBLER. 4.Mounting the sensors and actuators is a tedious task.
Courtesy: CMU, JPL
27
2.3.3 GMD Snake

The elegance and efficiency of a real snake’s
movements can hardly be reproduced by a robot
but it’s still a task to learn from nature. Watching a
snake’s path (on normal ground) we may observe
that the whole body is following the head in a
trace like a railway train. The main propulsion
here comes from a source not to be observed at
first glance - hundreds of tiny scales are moving
ahead and back on the bottom side of the snake.
To imitate such a motion, the mechanism is made
Figure 27: Snake rover. Courtesy: ICRA
to follow two main principles for the parts (or
sections) of a mechanical snake:
1. Each section should be actively moveable in direction of its longitudinal axis
2. Each joint should be bent according to the movement of its predecessor with a certain
delay
The first principle is realized by adding wheels around the snake’s body driven by a DC
motor in each section. The second principle is an application for distributed computing:
Each section has its own processor and a communication unit to talk to its neighbor
sections. Hence, if it is possible to estimate the speed of the snake’s actual forward
movement, every section can calculate the delay after which its own joint position must be
identical to the predecessors’ former position and has to send these data to its successor.
Each section is built like an aluminum cylinder with some holes in its surface for sensors.
Within each cylinder there are three 5W-DC motors to control the position of one universal
joint. To realize a technically equivalent movement we implemented a ring of wheels
around each section of our snake. Each ring of wheels is driven (via special joints) by one
additional DC motor per section such that every section can control its own forward force.
GMD snake encompasses numerous disadvantages. An uncontrolled torsion effect occurs
when the snake rover lifts some of its parts. The forward forces are generated by hundreds
of active scales under the snake’s body. These forces allow a normal land snake to slide
within its curved track. This robot does not provide sufficient space to mount rovers and it
is very difficult to overcome extra terrain surface. By the active joints we can obviously
adjust the curvature of the snake’s body. This is only limited by the maximal angle (here 45
degree) and by the discretization of the path induced by the straight rigid parts between the
joints. (Klassen and Paap, 1999)
Figure 28: Schematic of the body. Courtesy: ICRA
28
Table 3: Concept Screening matrix for Set C
Parameter Shrimp AMBLER GMD Snake

Suspension 0 - 0
Obstacle Climbing 0 + -
Control 0 - -
Size 0 0 +
Power 0 - -
Topple 0 + 0
Complexity 0 - -
Payload 0 - -
Speed 0 0 +
Cost 0 - -
Total 0 -4 -4
Rank 1 2 3
29
2.4 Set D
!"#$%&'()$*)('+,-$$.+*()-$(/$0(,(1(-2(/$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$3&-+4"56""6"7$
2.4.1 Chaos
Chaos, is a small unmanned ground
vehicle (SUGV) designed as a platform
for search, reconnaissance and
surveillance. The Chaos mobile robot
!"#$%&'()$*)('+,-$$.+*()-$(/$0(,(1(-2(/$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
from Autonomous Solutions is an
autonomous tracked robotic platform Locomotion of C haos:
designed for high mobility in areas
W alking A rticulated T racks:
with challenging terrain. It can carry a
wide range of sensors and payloads. Track beam movements can be coordinated to achieve different vehicle g
undulating, crawling, waddling, flailing, and other behaviors
The high mobility is a result of four
tracked arms that provide both high
Figure 29: Chaos
speed and strength. A real-time Performance
Mobility: Q uad independent drive/track arm hybrid
distributed control system controls xx Speed: Up to 8 K P H
each tracked arm independently, allowingxx agility and versatility in steep, uneven, and loose
V ertical Step: Up to 45 cm
G ap C rossing: Up to 60 cm
terrain. x Slope: O ver 45 degrees
x Side-Slope: O ver 45 degrees
x Runtime: 2 hours
The Chaos robot's four tracked arms use two motors - for!"#$%&'()$*)('+,-$$.+*()-$(/$0(,(1(-2(/$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$3&-+4"56""6"7$
rotating the arm and for track
Dimension
locomotion. The arm's rotation allows the robot to change itsBeam
posture,
PassiveLocomotion
climb over obstacles,
of C haos:
Q uad T racks :
elevate its body, etc. and the tracks provide2. tracks
the raised:
locomotion
49 cm (19.3 in) and steering. The Whistle's
1. H eight: tracks fl at: 23 cm (9.1 in)
3. W idth: 66.5The passively
cm (26.2 in) pivoting tracks conform to uneven terrain, maximizing surface co
high power output provides high arm torque 4. Land
ength: rapid changes
distributing
tracks extended±131 load. of in)arm
cmcrawling,
undulating, (51.6 position to allow
Track beam movements can be coordinated to achieve different vehicle gaits such as
waddling, flailing, and other behaviors
5. tracks retracted±78 cm (30.7 in)
high agility under heavy loads. While moving across terrain the Whistle's high power
!"#$%&'()$*)('+,-$$.+*()-$(/$0(,(1(-2(/$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$3&-+4"56""6"7$
6. T urning Radius: Zero within 105 cm (41.3 in)
Passive
7. W eight w/battery: 59.9 motion
kg (132 is one
lbs, achieved
pack) bykgmechanically
,69.4 (153 lbs, two packs) disengaging the track beam m
allows the robot to move at speed both on flat terrain sending
and steep commands inclines.
Locomotion of C haos:
to the motors and allowing them to be back driven by terrain
$
Variations Track beam movements can be coordinated to achieve different vehicle gaits such as
undulating, crawling, waddling, flailing, and other behaviors
1. Walking Articulated Tracks:

Track beam movements can be coordinated to achieve Passive Beam Q uad T racks :
different vehicle gaits such as undulating, crawling, waddling,

The passively pivoting tracks conform to uneven terrain, maximizing surface contact area and evenly
distributing load.
flailing, and other behaviors. Passive motion is achieved by mechanically disengaging the track beam motors or by not
sending commands to the motors and allowing them to be back driven by terrain forces.
2. Passive Beam Quad Tracks: Passive Beam Q uad T racks :

The passively pivoting tracks conform to uneven Shape Shifting
terrain, T racks : Shape shifting tracks can change from a track to a w
The passively pivoting tracks conform to uneven terrain, maximizing surface contact area and evenly
track mode, the track can be driven around the tensioning arms while the tension
maximizing surface contact area and evenly distributing
conform to the load.
distributing load.
terrain. In wheeled mode, the vehicle is capable of high speed on
Passive motion is achieved by mechanically disengaging the track beam motors or by not
Passive motion is achieved by mechanically disengaging the sending commands to the motors and allowing them to be back driven by terrain forces.
track beam motors or by not sending commands to theShape motors

Shifting T racks : Shape shifting tracks can change from a track to a wheel configuration In
track mode, the track can be driven around the tensioning arms while the tensioning arms pivot to
and allowing them to be back driven by terrain forces. conform to the terrain. In wheeled mode, the vehicle is capable of high speed on hard, flat terrain.
3. Shape Shifting Tracks:
Shape shifting tracks can change from a track to a wheel
Shape Shifting T racks : Shape shifting tracks can change from a track to a wheel configuration In
configuration. In track mode, the track can be driven around track mode, the track can be driven around the tensioning arms while the tensioning arms pivot to
Figure
conform to the terrain. 30:
In wheeled The
mode, 3 variations
the vehicle is capable of high speed on hard, flat terrain.
the tensioning arms while the tensioning arms pivot to
conform to the terrain. In wheeled mode, the vehicle is capable
of high speed on hard, flat terrain.
At each corner of the vehicle, rotary power must be available for two functions
simultaneously - drive function (high speed, low torque) and walk function (vice-versa).
30
244+'5&/A# 0%&# -+*&(+5# +(&# 4&55# 0%+.# 0%(&&# 8&&0# 8(2*# & 0%&# 3(2B./=# D.# 0%)5#
>)5)74&# 02# 0%&# /()>&(A# +4421).3# 82(# *2(&# '(&-)5&# .+>)3+0)2.# 2>&(# /)88)-B40#
-+.#+452#7&#(&*2>&/#8(2*#0%&#0()'2/#1%&.#0(+.5'2(0).3#0%&#(2>&(=#$%(&&#
1&(&#B5&/#0%(2B3%2B0#0%&#-2*'&0)0)2.#EF)3B(&#:!
2.4.2 Oregon State University Rover 2011 )=#
Overall rover configuration is a six wheeled
triple rocker system with two front rockers
and a rear rocker. The purpose of the chassis
serves as a mounting point for the rockers,
electrical box, and the robotic arm. It uses
linear actuators to turn the four corner
wheels, allowing the rover to rotate around a
central point
#$"%&'()"*+"'",-./&0&"12#"34/526&",78"91/)"'))&%#:$"+/;)"/()/<&"*(&"*+";1&".:');/0"6/%"1':=&)">1/01"/)";1&("
The chassis is designed to keep all six

%';&<" ;*" ;1&"*;1&6" 1':+" *+" ;1&" 6/%" >/;1" ;1&" ;/6&" /("#&;>&&(8" 91&" >1*:&" '))&%#:$" /)" 1&:<" ;*5&;1&6" #$"
%&'()"*+"'"12#"0'.?">/;1";>*")06&>)";1';";16&'<"/(;*";1&"12#8"@1&(";/51;&(&<?";1&)&")06&>)".2::";1&"
wheels on the ground over any terrain
;>*"6/%"1':=&)";*5&;1&6">1/:&")&026/(5";1&"12#"/()/<&";1&">1&&:8"
without the use of heavy springs or shock
absorbers. The center frame has three pivot
points on it, two on either side of the front
and one on the back. Each of these pivot Figure 31: The OSU rover. Courtesy: OSU #
points has a hinged set of two wheels
&-'-$%"/#'0,&'(4)4'5/"6&%3"07'826&%'9,:;;&/#&*'<!%2=';&.0'02'%"#,0>'3"-&?=2$/0&-'.2%':'
attached to it, called a "bogie". This insures that when driving over uneven terrain, the
%' -$%"/#' 0,&' 3:=1;&' %&0$%/' 0:3C' <0,&' 3:=&' 123"0"2/' A:3' :;32' $3&-' .2%' 0,&' &D$"1=&/0'
& the wheel pair, keeping
center pivot of any given bogie maintains the average deflection of
6&%6"&A'2.'0,&'3$%%2$/-"/#':%&:'.2%'210"=:;'3&:%@,"/#'-$%"/#'0,&':30%2/:$0'%&3@$&'0:3CE'
!"#$%&'()'*+&&,'-&."#/'0/-'-%"1&'2343%'"/4&%506&'374"2"8&-'53%'9&"#+4'0/-':$"6;'4"%&'6+0/#"/#)'<0,,33/'4"%&.'%&-$6&'
all#%3$/-'7%&..$%&='&/0>,"/#'4+&'%31&%'43'/01"#04&',33.&'.3",'9"4+3$4'-"##"/#'"/)'?%&0-'90.'0--&-'43'"/6&0.&'4%064"3/)'
wheels on the ground, providing an overall smoother ride. The " real advantage to this
&13'0,&'A,&&;3'"/'6"&A'A,";&'-%"6"/#'26&%'%2$#,'0&%%:"/'-$%"/#'0,&'3"0&'3$%6&7'0:3C*'
design is that it eliminates the need & for any sort of heavy
A("*6<&6";*"5'/("%*6&";6'0;/*("*(":**)&")26+'0&)?";1&";/6&)">&6&"2.56'<&<">/;1"'".*:$26&;1'(&";6&'<8"91&"
suspension, allowing for a simpler and more light-weight design (less things to break).
);6/.)";1';"%'B&"2.";1&";6&'<">&6&"02;"+6*%");*0B")1&&;)"*+".*:$26&;1'(&">/;1"'");'(<'6<")1&&;-%&;':"
spring or shock absorber
&
#6&'B".6&))?"'(<"#*(<&<";*";1&".*:$26&;1'(&";/6&"2)/(5"'"0$'(*'06$:';&"'<1&)/=&"3C6'D$"E:2&78"
!"#$$%$& &
This concept is not unique in itself, it can be seen on off-highway articulated trucks
In this rover, the rear bogie is similar in function to
91&"01'))/)"34/526&"F7">')"<&)/5(&<";*"B&&."'::"
the front of the truck. In effect, the chassis is
)/G" >1&&:)" *(" ;1&" 56*2(<" *=&6" '($" ;&66'/("
;!;#J+(5#K2>&(#1&(&#/&5)3.&/#02#'(2>)/&#+#%)3%&(#4&>&4#28#(&4)+7)4)0,#+./# >/;1*2;" ;1&" 2)&" *+" 1&'=$" ).6/(5)" *6" )1*0B"
flipped91&"180
'#)*6#&6)8" degrees.
0&(;&6" +6'%&" 1')"This;16&&"is done to insure the front
./=*;"
of the rover would experience a smoother ride than
B5#J+(5#K2>&(=#$2#+-%)&>&#0%)5A#0%&#&4&-0()-+4#0&+*#5&0#5&>&(+4#32+45=# .*/(;)"*("/;?";>*"*("&/;1&6")/<&"*+";1&"+6*(;"'(<"
*(&"*(";1&"#'0B8"H'01"*+";1&)&"./=*;".*/(;)"1')"
the back, as this is where the arm is mounted. This
'"1/(5&<")&;"*+";>*">1&&:)"';;'01&<";*"/;?"0'::&<"
orientation
'" I#*5/&I8" also
91/)" /()26&)" eliminates
;1';" the raised "tail" from the
>1&(" <6/=/(5" *=&6"
#&>&(,#&4&-0()-+4#*2/B4&#1+5#(&MB)(&/=#D8#'(2'&(#/2-B*&.0+0)2.#'(+-0)-&5# camera's
2(&=&(" ;&66'/(?" forward field*+"of
;1&" 0&(;&6" ./=*;" '($"view.
5/=&("
#*5/&" %'/(;'/()" ;1&" '=&6'5&" <&+:&0;/*(" *+" ;1&"
.3#+./#).0&3(+0).3#5,50&*5#12B4/#%+>&#0+C&.#*B-%#42.3&(=# In .'/6?"
" >1&&:" order to allow
B&&./(5" the
'::" >1&&:)" *(" rover to turn in place without
;1&" 56*2(<?"
!"#$%&'@)'?+&'6+0..".'-&."#/='5&04$%"/#'0'>3#"&A4B7&'.$.7&/."3/'43'
Figure 32: Chassis design. Courtesy: OSU
70.."1&,B'63/53%2'43'6327,&C'4&%%0"/='-"%&64'2343%'-%"1&'43'&06+' skidding (desirable in very complex terrain), a set
.6*=/</(5" '(" *=&6'::" )%**;1&6" 6/<&8" 91&" 6&':"
9+&&,='0/-',"/&0%'064$043%'.4&&%"/#'63/4%3,)' '<='(;'5&";*";1/)"<&)/5("/)";1';"/;"&:/%/(';&)";1&"
of +*6"
actuators
'($" )*6;" *+"controls
1&'=$" ).6/(5"the
" front and rear wheels,
/# 7&#*2/B4+(#52#0%+0#)0#-2B4/#7&# MB)-C4,#+./#5+8&4,#).50+44&/#2(#(&*2>&/#
'" )/%.:&6" '(<" %*6&"
(&&<"
!"#$%&'()'*+,--"-'.&-"#/'"/-0"%,1"2/',/.'.&-"#/30+,-&'.&42/-1%,1"2/'25',6631&%%,"/'7,0,8"6"19)'
"
rotating them about a vertical axis via a swivel inside the bogie so that all wheel axes
'#)*6#&6" )2).&()/*(?" '::*>/(5" +*6" :/51;->&/51;" <&)/5(" 3:&))" ;1/(5)" !
;*"
*6" )1*0B"
#6&'B78" 91/)"
intercept the same point at the center of the rover. (

0*(0&.;"/)"(*;"2(/J2&"/("/;)&:+?"/;"0'("#&")&&("*("*++-1/51>'$"'6;/02:';&<";620B)"34/526&"! )8"
&#'(&>)2B5#(2>&(#8&+0B(&/#'22(4,#/&5)3.&/#%+(/1+(&#0%+0#1+5#/)88)-B40#02#
!"##$%&'()&*+$,-"%+&(
#$" %&'(&" )%" *++%,"A(";1&"KLML"N'6)"O*=&6"01'))/)"<&)/5(?";1&"6&'6"#*5/&"/)")/%/:'6"/("+2(0;/*(";*";1&"+6*(;"*+";1&";620B8"A("
)-(" &%.(&" )%" )/&$" 0$" 1+*2("
&++&0;?" ;1&" 01'))/)" /)" +:/..&<" MPL" <&56&&)8" 91/)" >')" <*(&" ;*" /()26&" ;1&" +6*(;" *+" ;1&" 6*=&6" >*2:<"
2/B4+(#5,50&*#+>2)/5#0%)5#0,'&#28#-2*'4)-+0)2.#EF)3B(&#::G=#
,0)-%/)" 340''0$5" 6'(30&*7+(" 0$" .(&8" 2%91+(:"
&G.&6/&(0&" '")%**;1&6"6/<&" ;1'(";1&"#'0B?" ')";1/)"/)">1&6&";1&"'6%" >/::"#&"%*2(;&<8"91/)"*6/&(;';/*("
)(&&*0$;<" *" 3()" %=" *2)/*)%&3" 2%$)&%+3" )-(" =&%$)"
':)*"&:/%/(';&)";1&"6'/)&<"I;'/:I"+6*%";1&"0'%&6'Q)"+*6>'6<"+/&:<"*+"=/&>8"R".6&=/&>"*+";1&"01'))/)Q"'#/:/;$"
*$'"&(*&",-((+3<"&%)*)0$5")-(9"*7%/)"*".(&)02*+"
;*" 0*(+*6%" ;*" ;&66'/(" >')" 5&(&6';&<" <26/(5" ;1&" <&)/5(" .1')&" ;*" .6*=/<&" /()/51;" /(;*" ;1&" '::-;&66'/("
-+4# 5,50&*5# 1&(&# (&MB)(&/=# $22# *+.,# *)50+C&5# 1&(&# *+/&# ).# '(&>)2B5#
*:03" .0*" *" 3,0.(+" 0$30'(" )-(" 7%50(" 3%" )-*)" *++" )8"
.&6+*6%'(0&"*+";1&"01'))/)"34/526&"!
,-((+" *:(3" 0$)(&2(1)" )-(" 3*9(" 1%0$)" *)" )-(" !"
+>2)/&/#)8#0%&(&#%+/#7&&.#*2(&#-244+72(+0)2.#1)0%).#0%&#&4&-0()-+4#0&+*=#
"
2($)(&"%=")-("&%.(&"6>05/&("!;?"
@-(" *$5+(" )-(" ,-((+3" $(('" )%" )/&$<" )-(" =/++8"

(:)($'('" *$'" &()&*2)('" *2)/*)%&" +($5)-3<" *$'"
2-*3303"5(%9()&8",(&("2%970$('"0$)%"*"383)(9" !"#
(A/*)0%$3?" B30$5" CD@EDF" )%" 3%+.(" )-(" 383)(9<" "
!"#$%&'=)'>+&'7+,--"-'-1&&%"/#'42.&)';"/&,%',71$,12%-'&?1&/.',/.'
)-(" 3%+/)0%$3" )%" )-(3(" (A/*)0%$3" '(=0$('" )-(" %&1%,71'Figure 33:<+&&6-@'
12' 1$%/' Steering arrangement.
&/,86"/#' Courtesy:
72/1%266&.' OSU
%21,1"2/' ,82$1' ,'
+($5)-"%=")-("3,0.(+"*&93<"*$'",-(&(")-("+0$(*&" 7&/1%,6',?"-)'
*2)/*)%&"9%/$)3")%")-("7%50(?"@-("*'.*$)*5("-(&("03")-*)")-("=/++"&*$5("%="9%)0%$"%="(*2-"*2)/*)%&",*3"
" 31
/3('<"*++%,0$5"=%&"39*++(&"*2)/*)%&3<"&('/20$5")-("%.(&*++",(05-)?"hƚŝůŝǌŝŶŐƚŚĞĂĐƚƵĂƚŽƌ͛ƐĨƵůůƌĂŶŐĞ"*+3%"
1&(.($)3"'*9*5(")%")-("&%.(&"=&%9"/$'(&G"%&"%.(&G3)((&0$5")-(",-((+3"0$")-("(.($)"%="*"2%$)&%+"=*0+/&(?"
./"0-"+$1(
Table 4: Concept Screening matrix for Set D
Parameter Chaos Oregon Rover

Suspension 0 -
Obstacle Climbing 0 -
Control 0 +
Size 0 -
Power 0 +
Topple 0 0
Complexity 0 +
Payload 0 -
Speed 0 -
Cost 0 +
Total 0 -1
Rank 1 2
32
2.5 Concept Scoring Matrix
Table 5: Concept Scoring matrix
Rocker Bogie ATHLETE Shrimp Chaos

Parameter Weight
Score Weighted Score Weighted Score Weighted Score Weighted
(1-5) Score (1-5) Score (1-5) Score (1-5) Score
Suspension 0.10 5 0.50 4 0.40 3 0.30 4 0.40
Obstacle 0.10 3 0.30 4 0.40 4 0.40 4 0.40
Control 0.15 5 0.75 1 0.15 5 0.75 1 0.15
Size 0.10 4 0.40 0 0.00 4 0.40 2 0.20
Power 0.10 3 0.30 1 0.10 3 0.30 1 0.10
Topple 0.05 3 0.15 4 0.20 3 0.15 3 0.15
Complexity 0.15 3 0.45 0 0.00 3 0.45 1 0.15
Payload 0.05 3 0.15 5 0.25 4 0.20 2 0.10
Speed 0.05 0 0.00 3 0.15 2 0.20 5 0.25
Cost 0.15 5 0.75 0 0.00 5 0.75 2 0.30
Total 1.00 34 3.75 22 1.65 36 3.90 25 2.2
Rank 2 4 1 3
33
3. Realistic modeling and dynamic simulation of the Shrimp rover in

RecurDyn
In this section, we present the needs for a computer simulation; the reasons for choosing
RecurDyn; and the realistic modeling and dynamic simulation of the Shrimp rover.
An extra-terrestrial rover, like Shrimp, is a spatial multi-body system with many degrees of
freedom. It is essentially a multi-variable, multi-parameter coupled non-linear system. This
makes the kinematic and dynamic analyses complex and difficult to solve. Also, it is
difficult for beginners to learn and visualize how these manipulators would perform under
certain dynamic conditions.
Computer Simulation have been recognized as an important research tool since the
beginning of the 20th century as reported by Zlajpah (2010). It was an academic research
tool in the beginning but now it has evolved as a powerful tool supporting the design,
planning, analysis, and decisions in different areas of research and development. It has
been used by various researchers, developers and manufacturers in a variety of fields. The
use of computer simulation in the study of robotics is not an exception. For complex
systems as robots, the simulation tools can certainly enhance the design, development, and
even the operation of the robotic systems. Augmenting the simulation with visualization
tools and interfaces, one can simulate the operation of the robotic systems in a very
realistic way.
With the advent of various dynamic simulation packages (e.g., ADAMS, RecurDyn,
Autodesk Inventor, CATIA etc.), one can easily perform dynamic simulation of complex
mechanical systems like an industrial robot assuming an oversimplified control scheme, as
demonstrated by Doan (2008) and Alshamasin (2009). Similarly, with various control
simulation packages (e.g., MATLAB/Simulink, LabVIEW etc.), one can simulate the task
of a controller but without a realistic model of the plant, as shown by Rastogi (2011), Sun
(1995) and Schlotter (2003). For example, while performing the control simulation of a
robot using MATLAB, one generally needs to prepare the mathematical model of the robot
first which requires the derivation of complex kinematics and dynamics equations. This
makes the process very time consuming. MATLAB/SimMechanics can however aid this
process.
RecurDyn, as a multi-body dynamics analysis software, can overcome some of the
shortcomings in other traditional dynamics analysis software such as excessive
simplification, low solving efficiency and bad solving stability. The solver of RecurDyn
consists of equations of motion theory with recursive formulation, which makes RecurDyn
to be a fast and high precision method. RecurDyn also supports almost all types of file
created by different CAD software. Furthermore, it can be interfaced with MATLAB/
Simulink, which facilitates co-simulation of mechanical and control system [Wang (2011)
and Oh (2007)]. Hence using these two software, dynamic co-simulation of robots can
easily be performed. For performing dynamic co-simulation of any robot, RecurDyn and
Simulink can be connected such that RecurDyn provides the actual plant model to
Simulink to perform its control system simulation. This strategy saves a lot of time as one
need not derive the complex kinematics and dynamics equations of the robot to prepare its
mathematical model in Matlab/Simulink. However, with the recent introduction of CoLink
toolkit in RecurDyn, the whole task of dynamic simulation can be performed
34
independently by RecurDyn. Thus, the use of this toolkit eliminates the use of Matlab/
Simulink from this process.
A potential application of this software for realistic modeling and dynamic co-simulation
of the Shrimp rover is presented. Time dependent variations of several parameters such as
joint torques, velocity, acceleration are easily obtained from the software. Also, the actual
robot motion is visualized.
3.1 Realistic Modeling of Shrimp rover in RecurDyn

The solid modeling of the rover was done in CATIA V5. The CAD model of the full rover
was imported into RecurDyn/Professional environment using STEP file format as shown in
Fig. 35. The STEP file does not contain any information about the kinematic constraints
present in the rover assembly. Hence, mechanical joints in the form of kinematic
constraints were defined in RecurDyn/Professional environment. The details of all the
joints present in the rover are presented in Tab. 1.
All the wheels were constrained to move with a constant angular velocity of ! rad/s (30
rpm). Sliding joint friction (µstatic=0.5 and µdynamic=0.3) was introduced at all the revolute
joints for more realistic modeling.
Standard “Ground” body in RecurDyn was modified for the road/track definition. The road
outline was made using the outline tool and the outline surf tool was used to make the road
surface. The road was rendered as a surface on which the rover is expected to move.
Structured terrains involving slopes, steps, concave & convex surfaces were modeled to
evaluate the rover’s performance. To make the contact of rover wheels with the road,
surface to surface contacts were defined between the road surface and wheel surface. These
surface contacts constrain the wheels to make a contact with road whenever possible. To
make the tire to road contact more realistic, certain contact parameters such as spring
coefficient = 1000 N.mm, damping coefficient = 1 and dynamic friction coefficient = 0.9
were set.
Figure 34: Shrimp rover model imported to RecurDyn
35
Table 6: Joint definition of Shrimp rover
Type of Joint No. of Joints Type of Actuation
Between body and wheels 6 Active
In front fork 4+1 4 Passive, 1 Active

Revolute
joints In rear fork 1 Active
In right parallel bogie 6 Passive
In left parallel bogie 6 Passive
Fixed Between various parts of 31 No actuation

Joints the body
Figure 35: Joint Friction parameters
Figure 36: Surface contact properties
The use of spring is suggested in the front fork design to enable the rover to adapt to the
terrain more effectively. The “spring force” feature was used to create a spring that was
placed in an appropriate position. The spring properties such as stiffness, damping
coefficient, size, diameter, etc. were defined.
36
3.2 Dynamic Simulation of SHRIMP rover in RecurDyn

Kinematic and Dynamic analyses for the rover was performed using the multi-body
dynamics solver provided in RecurDyn. The solver uses fully recursive formulations which
are highly efficient, stable and fast. The accuracy of the simulation could be varied easily
by changing the number of time steps. Gravity (g=9806.5 mm/s^2) was applied in the -Y
direction. IMGALPHA integrator was used for numerical integration of differential
equations.
The rover capabilities are tested on 3 types of terrains:
1. Ability to climb steps
2. Ability to climb a inclined surface
3. Ability to adapt passively with concave/convex terrains
3.2.1 Ability to climb steps

The ability of the rover to climb the steps is measured by the maximum size of step it can
climb for a given wheel diameter. As Shrimp rover can climb steps of 2 times the size of
the wheels, the rover has the ability to climb step with size varying from 0 to 200 mm,
provided wheel diameter being 100 mm. A track with a 200 mm high step is created in
RecurDyn and the rover’s ability to climb this step is checked by performing its dynamic
simulation. The animation of the rover climbing the step is shown by the sequence of
pictures as shown in Figure 37.
Using RecurDyn’s plot tool, we obtained the driving torque v/s time graphs. These,
obviously, had some inherent noises affecting the actual data. Hence, a Fast Fourier
Transform (FFT) was applied on all individual driving torque v/s time plots. The
corresponding cut-off frequencies were also identified. The results were then filtered using
a Butterworth filter of order 3 and the cut-off frequency as that was just identified. Fig. 38
(a) & 4 (b) show the driving torque v/s time variation of the wheels. From such driving
torque v/s time plots, maximum torque requirements at each of the active joints can be
predicted. From such driving torque v/s time plots, maximum torque required at each joint
can be easily predicted.
Figure 37: Shrimp rover climbing a step
37
(a) Front tyre, Torque v/s Time
(b) Front right tyre, Torque v/s Time
(c) Rear right tyre, Torque v/s Time
38
(d) Rear tyre, Torque v/s Time
(e) Rear left tyre, Torque v/s Time
(f) Front left tyre, Torque v/s Time
Figure 38: Torque v/s Time graphs for all wheels while climbing step of 20 cm height
39
3.2.2 Ability to climb an inclined surface

The ability to climb an inclined surface is measured by the maximum inclination angle it
can climb. Various terrains are designed with different angles of inclination and rover’s
ability to climb them is tested with this software. It is found that the rover is able to climb
vertical inclinations of upto 40 degrees. The animation of the rover climbing the inclined
surface is shown by the sequence of pictures as shown in Figure 39.
Using the plot tool in RecurDyn, the time dependent driving torque variation at all the
active wheel-bogie joints were plotted as shown in Figure 40.
Figure 39: Shrimp rover climbing inclined surface
(a) Front tyre, Torque v/s Time
(b) Front right tyre, Torque v/s Time
40
(c) Rear right tyre, Torque v/s Time
(d) Rear tyre, Torque v/s Time
(e) Rear left tyre, Torque v/s Time
41
(f) Front left tyre, Torque v/s Time
Figure 40: Torque v/s Time graphs for all wheels while an incline of 40 degrees
3.2.3 Ability to adapt passively with concave and convex surfaces

It is very important for the rover to adapt passively to concave and convex surfaces
because actual terrains are not straight paths. For testing this ability, the rover was made to
fall from a certain height on a concave and convex surfaces and the behavior of the bogie
while it gets in contact with the terrain was observed. It was found that the rover has an
excellent ability to adapt with such terrains. Figure 41 shows the animation of the rover
passively adapting to a concave terrain.
Figure 41: Shrimp rover showing adaptability over concave surface.
After analyzing the torque requirement curves in all these cases, the maximum driving
torque required at each wheel was found out.
42
4. Shrimp rover manufacturing

Knowledge of torque requirements, as obtained from section 3, helped us in selecting the
motors for the rover. Tab. 7 shows the maximum torque requirements of each wheel as
obtained from the simulations. We notice that the maximum torque required is 32.1 Kg-cm.
Using a factor of safety (FOS) of 1.25, the torque requirement becomes 40.1 Kg.cm. The
closest available off-the-shelf motor had a torque rating of 45.0 Kg.cm, thus, it made the
most ideal choice.
The availability of components and facilities affected the design process greatly. Our final
rover was manufactured with using as many off-the-shelf products as possible.
Table 7: Torque requirements at each wheel.
Type of Joint Torque (kg-cm)
Front 32.1
Right bogie front 7.5
Right bogie read 7.4
Left bogie front 7.0
Left bogie rear 6.8
Rear 16.3
4.1 Main body
We began the construction with a reasonably accurate (up to 0.5 mm accuracy) cut-out of
2mm thick aluminium plates, which form the top cover and the base plate of the rover. The
shape of these twin plates is a square with a triangle appended on one of its sides. Thus, it’s
a pentagon, with the vertex of the triangle forming the rear of the rover. Fixed between the
two plates, on either side, is an aluminium block, which has a pair of ball-bearings set into
it. This block joins the two plates, serves as a column for load bearing and provides
mounting points for the parallel bogies. The bearings on these blocks prove to be a revolute
joint between the bogie and the main body. There are 2 other such blocks for mounting the
front fork, while a plain, bearing-less block for supporting the rear fork.
4.2 Parallel bogie
It consists of a set of links, which form a couple of two wheels, mounted on a support that
can freely rotate around a central pivot. We used C-section links to build the frame of the
bogies. The C-section allows for the frame to be sufficiently light without compromising
on its strength and rigidity. We used two different cross-section sizes for the C-section
links such that amongst the two, the smaller one could be perfectly inserted inside the
bigger one. The frame was so formed that no adjacent links were of the same cross-section,
thus, permitting us to create a freely rotating revolute joint by merely using a rivet. It is
advised to exercise caution during the manufacture of the two bogies, because they need to
be greatly identical. Any mismatch between the twins will give rise to non-uniform travel
of rover.
43
(a) Alumnium plates with 2 side blocks fixed (b) Parallel bogie mounted upon main body;
.....(c) Rover (incomplete) setup place on an inclined surface.........................(d) The basic mechanical structure of the rover....................
Figure 42: The rover during different stages of manufacturing
4.3 Front and rear forks
The front fork consists of a 4 bar mechanism robustly mounted on 2 aluminium blocks. In
all, 4 bearings are used to move the fork. The front fork has a servo mounted on it, to assist
in the steering process. It does so by rotating the wheel about an axis passing through the
wheel centre perpendicular to the ground.
The rear fork is a fixed link, at the end of which a wheel is mounted. It too, has a steering
system to rotate the wheel.
4.4. Electronics sub-system
The electronics sub-system of the rover is distributed over the base station and on-board.
On-board control system includes two Atmega16 based development boards interfaced in
Master-Slave configuration (SPI Interface) [6], which controls various sensors and
actuators installed on the rover for its operation. The sensor data and control signals are
wirelessly transmitted to the base station using a pair of Xbee trans-receiver modules. The
base station involves a PC running on Intel Core 2 Duo processor, which provides a
MATLAB based GUI to the end-user to interact with the rover. A wireless camera has also
been installed on board which uses a separate RF channel and sends the live video feed to
44
the GUI. Fig. 43 shows the overall schematic of the control system of the rover. Whereas,
Fig. 44 shows the completed rover.
Figure 43: Schematic of control system of the rover
45
Figure 44: The completed rover
46
5. Experimental Setup
The rover was tested to experimentally validate the results of RecurDyn. The objective was
to compare RecurDyn results against experimental results with respect to the torque
requirements of each individual motor. The rover was made to traverse on two types of
terrains, one including an inclined surface and the other having a step. As observed in
section 2, using RecurDyn, the driving torque requirements of each wheel were directly
available as time graphs, for both these terrains. However, for obtaining experimental
results, we measured the armature current variation at each motor while the rover fared
through these obstacles. These readings were later multiplied by Kt (the motor torque
constant) to obtain the corresponding torque values. A separate experiment was conducted
to measure the value of Kt
! = Kt " I (1)
where, ! is the torque, N.m

I is the current, A.
Fig. 10 shows the basic experimental setup and better explains the method of obtaining
data for real-time driving torque measurement.
Figure 45: Schematic for real time driving torque measurements
1.1. Step test
This test consisted of putting a step in the path of the rover. The step size was arbitrarily
chosen as 8 cm high. The setup comprised of a base, a step and a raised platform, all of
which were made of plywood. This ensured that the coefficient of friction between the
wheel and surface remained constant. The rover had no difficulty in climbing the step and
we were successfully able to measure the current variation at all motors independently as
the rover traversed the path. Care was taken to ensure that the experimental setup matched
the setup used for simulations in RecurDyn. Fig. 46(a) shows the rover during step test.
1.2. Inclined slope test
An inclined surface, big enough to accommodate the rover completely on it, was used as
the next obstacle. The inclination was arbitrarily chosen as 30 degrees. This setup also
consisted of a base and a raised platform. We performed the experiment in a similar
manner to obtain real-time driving torque values. The experimental setup was made to
match the setup used for simulation in RecurDyn. Fig. 46(b) shows the rover during slope
test.
47
(a) Rover during step climb test
(a) Rover during inclination test
Figure 46:Rover traversing terrain for experimental validation
48
6. Results and discussions of the tests

The driving torque data obtained from the experiments was imported into MATLAB. The
noise was filtered away using FFT and Butterworth filter. RecurDyn simulation results
were also exported in text file format and were subsequently imported into MATLAB to
perform a comparative study on the driving torque v/s time plots obtained from
experiments and RecurDyn. A common sampling frequency of 100 Hz was used in both
real and virtual environments. The simulation time in inclined slope test was 22 seconds
and 16 seconds in step test. Fig. 47 (a) shows the driving torque v/s time plot of back
wheel in the slope experiment and Fig. 47 (b) shows the driving torque v/s time variation
of front wheel in the step experiment.
The mean absolute percentage error (MAPE) is calculated with the help of Eq. 2.
!
! !!"# ! !!"#
!"#$ !
! !!"#
! (2)
MAPE for rear wheel (in slope test) was found to be 42.87%. MAPE for front wheel (in
step test) was found to be 35.62%. Similar comparisons can be performed for each motor
in both the cases easily. To avoid redundancy, in this paper, we present the comparisons
only for the above two cases. It was found that MAPE value typically varies between 35 to
45 % for different cases.
The primary reason for having such large variations in results is that the electrical DC
drive motor is not included in the RecurDyn simulation. The DC motor adds various non-
linearities to the system such as BEMF voltage, friction at motor bearings, etc. which
haven’t been accounted for.
(a) Slope Test; Rear wheel; 30 degree inclincation
49
(a) Step test; Front wheel; 8cm height
Figure 47: Driving torque v/s time plots comparison between real and simulated data
Other reasons contributing to the error could be the incorrect modelling of joint friction at
each revolute joint; incorrect estimate of contact friction between the road and wheel
surface; mismatch between modelled and actual mass-inertia properties of different parts in
the assembly. Also, nuts, bolts, clamps, bearings and other such parts were not included in
the RecurDyn model. This simplification might have contributed to the error as well. If
these sources of error can be suitably accounted for, a better comparison can be drawn.
However, the software was found to be particularly useful for selection of actuators
required by the rover for particular payload capabilities. It shows potential of being able to
greatly assist in the design process.
50
7. Results of the project

The results of this project can be summarized as follows:
Conceptual Results:
1. Identified the problems and issues faced by a high-performance rover for an intended
application in extreme environments
2. Identified the most effective means of rover locomotion by use of product development
methods
3. Used a new multi-body dynamics software for simulation purposes. In the process,
presented a case-study: potential application of the software for space robotics.
4. Made use of Concurrent Engineering: Design and manufacturing was a iterative process
that was greatly done simultaneously.
Tangible Results:
1. Built the rover
2. Tested its capabilities
3. Validated simulation results
4. Developed a research paper; It will be presented at the 12th Symposium on Advanced
Space Technologies in Robotics and Automation, Netherlands (May 15-17, 2013)
Further, successful completion of such projects might motivate more students / colleges to
take up research or projects in this field. Currently, for numerous reasons, only a select few
individuals who hold high posts in elite organizations, carry out research work in space
robotics in India. Looking at the big picture, this is quite contrasting to the situation in
other developed countries where the space industry is (partly) open for privatization.
Indian space research industry could benefit more with an open community with a
common objective; not a closed elite community with a classified objective.
51
5. References
Alshamasin, M.S., Ionescu, F. and Al-Kasasbeh, R.T., (2009), Kinematic Modeling and Simulation of a
SCARA Robot by Using Solid Dynamics and Verification by MATLAB/Simulink, European Journal of
Scientific Research, Vol. 37 No. 3, pp. 388-405
Barlas, F., (2004), Design of a Mars Rover Suspension Mechanism, Masters Thesis, Department of
Mechanical Engineering, Izmir Institute of Technology
Cepolina, F., (1997), Design and simulation of an all terrain mobile robot, Undergraduate Thesis, Department
of Mechanical Engineering, University of Leeds
Doan, T.T., Kim, I.S., Kim, H.H., Jeong, W.J. and Kang, B.Y., (2008), Developed Simulation Model-
Kinematics for Robotic Arc Welding. Asian International Journal of Science and Technology in Production
and Manufacturing, Vol. 1, No. 2, pp. 69-76
Hardarson, F., (1997), Locomotion for difficult terrain, Technical Report, Dept. of Machine Design, Royal
Institute of Technology, Sweden
Heaston, J. R., Hong, D. W., Morazzani, I., Ren, P. and Goldman, G., (2007), STriDER: Self-Excited Tripedal
Dynamic Experimental Robot, 2007 IEEE International Conference on Robotics and Automation, Roma,
Italy
Klaassen, B. and Paap, K.L., (1999), GMD-SNAKE2: A snake-like robot driven by wheels and a method for
motion control, Proceedings of International Conference on Robotics and Automation
Krotkov, E., Simmons, R. and Whittaker, W.L., (1995), Ambler: Performance of a Six-Legged Planetary
Rover, Acta Astonautica, Vol. 35, No. 1, pp. 75-81.
Laney, D. and Hong, D.W., (2005), Kinematic Analysis of a Novel Rimless Wheel with Independently
Actuated Spokes, 29th ASME Mechanisms and Robotics Conference, Long Beach, California
Laney, D. and Hong, D.W., (2006), Three-Dimensional Kinematic Analysis of the Actuated Spoke Wheel
Robot, 30th ASME Mechanisms and Robotics Conference, Philadelphia, Pennsylvania
Lauria, M., Conti, F., Mäusli, P.-A., Van Winnendael, M., Bertrand, R. and Siegwart, R., (1998), Design and
Implementation of an Innovative Micro-Rover, ASTRA, ESTEC, The Netherlands
Lindemann, R. A. and Voorhees, C. J., (2005), Mars Exploration Rover mobility assembly design, test and
performance, http://hdlhandlenet/2014/37604
Oh, K. Hwang, J.P., Kim, E. and Lee H., (2007), Path Planning of a Robot Manipulator using Retrieval RRT
Strategy, World Academy of Science, Engineering and Technology 25, pp. 171-174.
Rastogi, A. and Guruprasad K.R., (2011), Kinematic Analysis of 5DOF manipulator arm for mine detection,
7th International Conference on Trends in Industrial Measurements and Automation
Ren, P., Hong, D.W. and Morazzani, I., (2008), Forward and Inverse Displacement Analysis of A Novel
Three-Legged Mobile Robot Based on the Kinematics of In-parallel Manipulators, ASME Journal of
Mechanisms and Robotics
Roman, M.J., (2005), Design and Analysis of a Four Wheeled Planetary Rover, Masters Thesis, University of
Oklahoma
Schlotter, M., (2003), Multibody System Simulation with SimMechanics, Technical Report, Darmstadt
University of Technology
Shirley, D. L., (1995), Mars Pathfinder Microrover Flight Experiment - a paradigm for very low-cost
spacecraft, Acta Astronaut. (UK), Vol. 35, Suppl. issue, pp. 355 - 365
Siegwart, R., Lauria, M., Mäusli, P.-A. and Van Winnendael, M., (1998), Design and Implementation of an
Innovative Micro-Rover, Proceedings of Robotics, 3rd Conference and Exposition on Robotics in
Challenging Environments, Albuquerque, New Mexico
52
Siegwart, R., Lamon, P., Estier, T., Lauria., M., Piguet R., (2002), Innovative design for wheeled locomotion
in rough terrain, Robotics and Autonomous Systems, Vol. 40, Number 2, pp. 151-162
Stone, H.W., (1996), Mars Pathfinder Microrover: A Small, Low-Cost, Low-Power Spacecraft, http://
hdl.handle.net/2014/25424livepage.apple.om
Sun, S. S. and Meng, Q.H.M., (1995), Dynamic Simulation of Robot Manipulators Using Graphical
Programming Packages, Proceedings of IEEE Pacific Rim Conference on Communications, Computers, and
Signal Processing, pp. 521-524.
Tompkins, P., (2005), Mission-directed path planning for planetary rover exploration, Masters Thesis, The
Robotics Institute, Carnegie Mellon University
Wagner, M., Heys, S., Wettergreen, D., Teza, J., Apostolopoulos, D., Kantor, G. and Whittaker, W., (2005),
Design and Control of a Passively Steered, Dual Axle Vehicle, i-SAIRAS
Wang, Y.-S., (2011) Dynamics co-simulation of a type of spot welding robot by RecurDyn and Simulink,
International Conference on Consumer Electronics, Communications and Networks
Wettergreen, D., Cabrol, N., Baskaran, V., Calderon, F., Heys, S., Jonak, D., Lüders, A., Pane, D., Smith, M.,
Teza, J., Tompkins, P., Villa, D., Williams, C., Wagner, M., Waggoner, A., Weinstein, S. and Whittaker, W.,
(2005), Second Experiments in the Robotic Investigation of Life in the Atacama Desert of Chile, i-SAIRAS
Wilcox, B.H., (2009), ATHLETE: A Cargo and Habitat Transporter for the Moon, Proceeding of the 2009
IEEE Aerospace Conference, Big Sky, Montana,
Wilcox, B.H., (2008), ATHLETE: An Option for Mobile Lunar Landers, Proceedings of IEEE Aerospace
Conference, Big Sky, Montana
Zakrajsek, J.J., McKissock, D.B., Woytach, J.M., Zakrajsek, J.F., Oswald, F.B., McEntire, K.J., Hill, G.M.,
Abel, P., Eichenberg, D.J. and Goodnight, T.W., (2005), Exploration Rover Concepts and Development
Challenges, http://ntrs.nasa.gov/search.jsp?R=20050175879
Zlajpah, L., (2010), Robot Simulation for Control Design, Robot Manipulators Trends and Development
53
View publication stats

B 18 Final Evaluation Report

Uploaded by

Copyright:

Available Formats

B 18 Final Evaluation Report

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

B 18 Final Evaluation Report

Uploaded by

Copyright:

Available Formats

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

ALL-TERRAIN MOBILE ROBOT FOR EXTRA-TERRESTRIAL APPLICATIONS

Thesis · May 2013

The user has requested enhancement of the downloaded file.

ALL-TERRAIN MOBILE ROBOT FOR

MIHIR R BHAGAT SHIVESH KUMAR RAGHAVENDRA S

PANKAJ CHOUDHARY JAYANT ILME PRIYADARSHINI S

under the guidance of

DEPARTMENT OF MECHANICAL ENGINEERING

Firstly, we would like to thank the Department of Mechanical Engineering, National

Table 1. Concept Screening Matrix for Set A . . . . . . . . . . . . . 20

Figure 1: Sojourner Rover using Rocker Bogie Mechanism . . . . . . . . . 13

Figure 36: Surface contact parameters . . . . . . . . . . . . . . . . 36

Scope of the Project

1. To identify needs and associated problems of a Mars rover.

the position ofSide

Types Of Joints Figure 3. Stowed MER Rover on Lander Basepetal

Robots using rocker bogie mechanism make use

symmetric structure for both

Wheel Strut Rocker ____-

bogies and the position of each

with the lowest normal force 3

Crabbing Zero Radius Ackerman Differential

Figure 1.7: Changing direction

Figure 1.6: Rocky Rover series (images reproduced from NASA)

2.1.2 4 wheel drive rover - Zoë

We haveand built ainsufficient steering

axles is trailing or not. bring about a given arc path. To overcome th

CoG top view CoG

4 Siegwart, Lauria, Mäusli, Van Winnendael

Figure 15: Recovery after flipping over.

5 Siegwart, Lauria, Mäusli, Van Winnendael

2.1.4 Concept Screening Matrix

Table 1: Concept Screening matrix for Set A

Parameter Rocker Bogie Zoë SpaceCat

Figure 17: Design of each limb. Courtesy: NASA

Figure 18: Basic Joint Design. Courtesy: NASA

Figure 19: STriDER. Courtesy: IEEE

Climbing stairs upto 4 times its nominal height

Staying level and / or maintaining pitch and roll on all inclines

Terrain adaption and multiple points of contact for stability

2.2.4 Concept Screening Matrix

Table 2: Concept Screening matrix for Set B

Parameter ATHLETE IMPASS STriDER

Figure 23: Steering. Figure 24: Coupling of wheels. Courtesy: EPFL

2.3.3 GMD Snake

Figure 28: Schematic of the body. Courtesy: ICRA

2.3.4 Concept Screening Matrix

Table 3: Concept Screening matrix for Set C

Parameter Shrimp AMBLER GMD Snake

W alking A rticulated T racks:

1. Walking Articulated Tracks:

different vehicle gaits such as undulating, crawling, waddling,

2. Passive Beam Quad Tracks: Passive Beam Q uad T racks :

track beam motors or by not sending commands to theShape motors

The chassis is designed to keep all six

intercept the same point at the center of the rover. (

@-(" *$5+(" )-(" ,-((+3" $(('" )%" )/&$<" )-(" =/++8"

2.4.3 Concept Screening Matrix