1

CHAPTER 1
INTRODUCTION

1.1 EVOLUTION OF QUADRUPED ROBOT

Quadruped robots, inspired by four-legged animals, have undergone

various Signiant changes since the beginning. The first model began in the 1980s
with the earliest model focusing on basic locomotion and basic activities. The
latest innovative technology development meant to support human efforts at
almost every walk of life is the use of quadruped robots in Reconnaissance and
Surveillance, Maintenance and Inspection, Assistance and Support, Logistics
and Delivery and much more. Thus, considerable focus is created on quadruped
robots with an increasing demand for high efficiency and effectiveness in
activities like reconnaissance and search-and-rescue operations.

With the world's population estimated to have more than 7.8 billion people, and
the increase rate in urbanization is rising basically, the required answers were
under exploration and surveying. Exploration and surveillance using traditional
methodologies by their conventionality, which always suffered from
inefficiencies and inadequacies, had arrived as a paradigm shift using the
quadruped robot technology to offer unprecedented flexibility and adaptability.
Similarly, the rapid urbanization has made traditional delivery methods, through
the congested urban areas and remote regions, very hard. These robots however
present a revolutionary approach, offering flexibility and adaptability to navigate
through the tough environments. Their ability to handle the stairs, narrow
pathways and uneven surfaces, making them ideal for last mile delivery and
logistics in both urban and rural settings

As the advancement in the materials, actuators and control systems have since
helped in propelling the development. In the recent years, integration of Artificial
intelligence has enhanced their adaptability and functionality by helping them
 2

have the cognitive thinking abilities to have terrain appropriate gait algorithms
and proper battery power segregation. Modern quadruped robots can navigate
complex terrains, perform dynamic manoeuvres, and execute tasks previously
deemed impossible for robotic systems. Recent development in AI has enhanced
the quadruped robot in various ways:

i. Made the decision making and path planning tasks much advanced and
quicker
ii. Assists in object detection and semantic mapping
iii. Made the robot quicker to adaptive to varying terrains and improving its
performance by enabling the its learning capabilities
iv. Enables Human- Robot interaction via voice or gesture control
v. It enhances the Energy efficiency by optimizing the power consumption
by predicting the power needs at the current run conditions
vi. The most reliable of all is the SWARM intelligence, where many
quadruped robots can work collaboratively in swarm-like behaviour for
large applications like industrial setting or search and rescue operations.

1.2 CLASSIFICATION OF QUADRUPED ROBOT

The classification of quadruped robots is basically based on their

manipulators and mobility mechanisms. Some design gives more importances to
the speed and agility aspect of the robot where the frame is lightweight and joint
actuators are quick and precise. The other focuses on strength and stability
making it fit for tasks requiring heavy lifting. The choice of these characters is
based on the intended application
 3

Figure 1.1 :Classification of Quadruped Robot

1.3 APPLICATION OF QUADRUPED ROBOT

Quadruped robots have found applications across various sectors, each

addressing specific challenges.

1.3.1 Reconnaissance and Surveillance

These robots play a crucial role in these applications, especially in

military, security and disaster management scenarios. They can be equipped with
cameras, LiDAR and infrared sensors and AI based analysis models that can
autonomously patrol and detect anomalies. This data can be transferred real-time
to the operator. The quadruped robot’s ability to traverse rugged terrains, climb
stairs and stability making them ideal for these applications. The Figure 1.2 They
can also be deployed in stealth military operations to operate covertly behind
enemy lines, avoiding any risking of human life.

Figure 1.2: ADIN ROBOTICS Quadruped Robot for surveillance

 4

1.3.2 Maintenance and Inspection

In manufacturing and construction, quadruped robots are utilized to

transport materials across complex, instructed environments, acting as
automated pack mules. They can be used to transport parts around factories and
construction sites, thereby enhancing safety. Its ability to climb stairs and
manoeuvre over uneven terrains and avoid unnecessary human fatigue. They are
also employed to carry out routine inspections in hazardous areas and reduce the
risk of loss of human life who would be employed with the same task. This is
achieved by equipping thermal sensors, cameras or detectors to provide real-time
data to operators and have repetitive checks. They can also be used to deliver
tools and materials between assembly lines.

Figure 1.3: Quadruped Robot Used for Inspection and

Surveying of Power Plants

1.3.3 Assistance and Support

Quadruped Robots are also being used as assistive companions, similar to

that of the support animals, in healthcare and personal support. They can assist
the elderly by giving them physical support or keeping essential goods handy. In
disaster response, they can help the rescue teams by carrying the kits and
 5

communication equipment to the last mile. Their ability to work in such harsh
terrains helps and reduces the risk of first responders.

1.3.4 Logistics and Delivery

The industry is rapidly adopting automated supply chain management and

that has led to integration of these robots into the logistics field. These robots are
capable of transporting goods across locations within the plant. They also aid the
last-mile delivery solutions. Their ability to traverse crowded places and rough
terrains are effective for home delivery.

The anatomy of the quadruped robot helps in navigating through rubbles, debris
or collapsed structures in earthquakes or disaster-stricken areas to locate
survivors. The robots can access areas that are inaccessible by humans and can
operate in extreme conditions and work tirelessly. Robots can be used to
transport medical aids, water and foods during extreme scenarios.

The advancements in modern engineering, and the availability of advanced

sensor systems, autonomous navigation algorithms, and robust locomotion
mechanisms make quadruped robots epitomized by modern man. These are agile
in the transition from rugged mountainous terrain to very densely urbanized
environments that allow easy traversal for operations in exploration and
surveillance activities. Also, intelligent monitoring systems can capture and
analyse real-time data so that action and the best decisions may be taken by
operators. Artificial intelligence and machine learning are guiding principles in
quadruped robots that guide the adaptation of behaviour based on continuous
refinement of strategies considering environmental cues and mission objects.
 6

Figure 1.4: Puddu D1 is commercial delivery quadruped robot

Therefore, integration of quadruped robots into exploration and surveillance

represents the latest advance in technological innovation-capable of
unprecedented abilities across a broad spectrum of operational contexts. This
eventually brings about an advance in robotics which will redefine the
boundaries of exploration and surveillance, fulfilling the expansion of human
quest in knowledge and safety.

1.4 ON BOARD COMPUTERS

On-board computers (OBCs) are referred to as the central processor of

various robot systems. One standard OBC configuration is as depicted by Figure
1.4. They are nothing but the units that take input from various sensors related
to the robot's external and internal situations, and upon processing signals,
control actuators that are for movement and various other robot processes. The
OBC has input and output interfaces as pins and slots, with a processing unit,
insulated from external disturbances for optimal functioning. Under isolation,
the system functions reliably under stated environmental conditions, in which
both computation effectiveness and real-time processing are critical Khalastchi
& Kalech et al (2018).
 7

Figure 1.5: Raspberry PI on board computers

1.4.1 Integration of OBC with Quadruped Robot

The combination of on-board computers is extremely important for

facilitating autonomous movement in quadruped robots. The integration of such
a system is shown in Figure 1.5. The control system of quadruped robots can
either be implemented via microcontrollers or higher-level embedded computing
platforms. This project uses an OBC instead of a simple microcontroller because
of its greater processing capability, integration with AI environments, and
support for real-time communication. Unlike microcontrollers that need heavy
programming in low-level languages like C or C++, OBCs have support for high-
level robot frameworks like ROS (Robot Operating System), making
development and programming easier in addition, OBCs provide not only
accurate locomotion control but also on-line sensor fusion, terrain adaptation,
and obstacle avoidance. Their long-term scalability and compatibility with
machine learning models make them the best option for controlling the
quadruped robot in the current project work Rashtian et al. (2020); Tan et al.
(2021).
 8

1.5 NEED FOR THE STUDY

This study is timely because there is a possibility of wide-scale deployment

to overcome problems of exploration and surveillance. Ambitious though this
project is, a system that would advance reconnaissance and rescue operations
merits an exploration into its possible application to augment security and safety.
This report focuses on basic factors required for scaling of quadruped robot
systems for exploration and surveillance: reliability, as these have to play a very
crucial role in gathering diverse and hazardous environments.

Currently, quadruped robot applications are limited to particular missions, but

recent advancements open the door for a more general use: from surveillance in
urban conditions to disaster response. Using the capabilities of quadruped robots,
situational awareness will be enhanced, and resources will be optimally allocated
so that people's lives and assets are protected. Transformative long-term payoffs
are realized, and there are associated costs to the initial investments.
 9

CHAPTER – 2
LITERATURE REVIEW

In this chapter, various papers and research works related to quadruped robots
and their mechanics were reviewed. They were relating the designing and
simulation, calculation of inverse kinematics and gait algorithms, 3D printing for
parts assembly and OBC integration. This review helped in the identification of
the existing solutions and the gaps in them. This review also helps in the
foundational knowledge needed and inputs in the due course of the project.

2.1 DESIGN AND SIMULATION

2.1.1 3-Dimensional Design using SOLIDWORKS software

Hiroshi Kimura et al. (2016) stated that Quadruped robots are robotic systems
designed with four legs to traverse a variety of terrains, mimicking the
locomotion of animals. In recent years, the development of quadruped robots has
gained substantial momentum due to their superior adaptability in complex
environments, compared to wheeled or tracked robots. These robots are being
actively researched and developed for applications in areas such as disaster
rescue, military operations, space exploration, and industrial automation. Kumar
et al. (2020) study developed a robot model using the software and included
finite element analysis (FEA) to evaluate the structural integrity. The study
suggested the usage of aluminium and composite materials for a balance of
durability and agility. El-Dalatony et al. (2022) This research details the design
and control of a bio-inspired quadruped robotic leg with integrated actuation and
sensors. Using a three DOF kinematic model and cascaded controllers, the
lightweight leg achieves gait trajectory tracking, vertical jumping, and supports
up to 6.54 kg, validated through experimental testing.
 10

2.1.2 Simulation using ROS and R-Viz

Gennaro Raiola1 et al. (2012) developed the “The WoLF” software can be used
for the controlling of mounted devices such as arms and pan tilt cameras, jointly
with quadrupedal platform, The paper the authors discuss about the main features
of WoLF and its overall software architecture. The paper throws light into the
simulation and analysis of the 4-legged robot of different length and sizes, the
paper also helps us understand the basic ROS (used for simulation), and ROS
control by providing open-source materials in GitHub that can be accessed.

Siregar et al. (2023) introduces the mapping results, visualized in real-time using
RViz, demonstrating an impressive 97% accuracy rate, showcasing the robot’s
ability to navigate and map environments effectively. This approach highlights
the potential of quadruped robots in applications such as environmental
monitoring, disaster response, and industrial automation.

Müller et al. (2020) explored physics-based quadruped locomotion in ROS-

Gazebo, using RViz for real-time feedback. The research highlighted that RViz’s
interactive markers allowed dynamic adjustments of joint angles, enabling
precise gait optimization. The study further emphasized the importance of URDF
and SDF models in improving simulation fidelity

2.2 INVERSE KINEMATICS AND GAIT ALGORITHM

Muhammed Arif Sen et al. (2020) elaborates the concepts of Kinematics

analysis is the main problem in manipulators and robots. Dynamic and kinematic
structures of quadruped robots are very complex compared to industrial and
wheeled robots. In this study, inverse kinematic solutions for a quadruped robot
with 3 degrees of freedom on each leg are presented. Denavit-Hartenberg (D-H)
methods are used for the forward kinematic. The inverse kinematic equations
obtained by the geometrical and mathematical methods are coded in MATLAB.
And thus, a program is obtained to calculate the legs joint angles corresponding
to desired various orientations of the robot and endpoints of legs. Also, the
 11

program provides the body orientations of robots in graphical form. The angular
positions of joints obtained corresponding to desired different orientations of the
robot and endpoints of legs are given in this study. In this study, forward and
inverse kinematic analysis of a quadruped robot is investigated. The Denavit-
Hartenberg method is used for forward kinematics, analytical solutions are used
for inverse kinematics. The obtained kinematic equations are transferred to the
MATLAB environment and a program is developed to obtain the angular
position of joints of each leg corresponding to the different robot orientations
and different coordinate systems of the endpoint of legs. To test the program, 3
different inverse kinematic analyses are performed and the results obtained from
the program are presented in graphical form.

Chang et al. (2020) published in the paper that one prominent approach in
quadruped robot development involves the use of the Robot Operating System
(ROS) to streamline the design and control process. The development typically
begins with kinematics modelling to understand the robot’s movements and
solve motion equations. Gait planning, crucial for effective navigation, uses
methods like Cartesian space trajectory planning to ensure precise foot
placement. Simulations conducted in tools like MATLAB validate these planned
movements before integration with ROS.

2.3 3D PRINTING AND ASSEMBLY

Kim et al. (2023) highlight the role of topology optimization in 3D-printed

robotic legs, improving strength-to-weight ratios while maintaining mechanical

flexibility. The authors argue that lattice-structured designs reduce material

consumption while maintaining load-bearing capabilities for robotic movement.

Despite the advantages of 3D printing, assembling a quadruped robot requires

precise alignment of mechanical and electronic components.

 12

García et al. (2021) discuss how modular joint design simplifies assembly and
maintenance. Their study suggests that pre-assembled servo mounts and snap-fit
connectors can improve structural reliability while reducing manual assembly
errors. Moreover, research by Tan et al. (2024) emphasizes the importance of
integrating embedded sensors within 3D-printed components. Their findings
suggest that incorporating strain gauges and inertial measurement units (IMUs)
during printing enhances real-time monitoring of robotic gait and stability.

2.4 INTEGRATING OBC WITH ROBOT

Yibin Li et al. (2011) develops the Legged robots excel in navigating non-
uniform terrains, making them ideal for tasks like exploration, monitoring, and
data collection in environments where other solutions face significant
limitations. However, the high cost and complexity of most quadruped robots
limit accessibility for students and young researchers. To address this, an open-
source, low-cost quadruped robot is proposed, utilizing the ROS framework to
equip learners with tools for developing basic autonomous capabilities.

Patel et al. (2022) developed an advanced quadruped robotic system integrated

with an On-Board Computer (OBC) to enhance real-time locomotion and
environmental perception. The system utilized a Raspberry Pi 4 and an NVIDIA
Jetson Nano to process sensor data from LiDAR, IMUs, and vision cameras,
enabling autonomous navigation and obstacle avoidance. Their research
demonstrated that real-time AI inference on the OBC significantly improved
terrain adaptability and motion planning.

Additionally, Müller et al. (2021) explored the use of Robot Operating System
(ROS) middleware for seamless hardware-software integration in quadruped
robots. The study revealed that ROS-based communication frameworks
improved sensor fusion efficiency and optimized multi-threaded control loops
 13

for gait execution. The integration of an AI-assisted motion control system

further enhanced stability and adaptability on uneven terrains.

Wang et al. (2023) highlighted key challenges in OBC integration, such as

thermal management, power efficiency, and computational latency. Their
findings suggested that using AI accelerators like the Google Edge TPU or
NVIDIA CUDA cores improved real-time decision-making capabilities while
reducing energy consumption. The study emphasized the importance of efficient
power distribution to sustain long-term autonomous operations in mobile
quadruped robots.
 14

CHAPTER 3
PROBLEM STATEMENT AND OBJECTIVES

3.1 PROBLEM STATEMENT

Traditional methods of exploration are often limited by accessibility to

remote or dangerous locations with traditional forms of exploration and thus
result in incomplete or inefficiently gathered data. They are heavily dependent
upon human intervention and prone to environmental influences out of their
control. A recent innovation in exploration entails quadruped robots designed for
more efficient exploration. But to ensure effective autonomy, issues like terrain
adaptability, power efficiency, obstacle detection, and reliable data transmission
must be taken care of.

3.2 OBJECTIVES

This project is to design and develop a Raspberry Pi based quadruped

robot that can navigate varied terrains The robot will integrate mechanical
design, robust electronics systems, and efficient software algorithms to achieve
mobility and functionality.
 15

CHAPTER 4
MATERIALS AND METHODS

4.1 WORKING OF THE PROJECT

This project is to build a semi-automated 4-legged robot that can traverse

any type of terrain. This quadruped robot is an advanced system that can be used
for critical task handling. It integrates mechanical, electronic and computational
subsystems, which is handled by the Robotics Operating software framework
which is proven to be highly effective for seamless communication and quick
task execution.

At startup, the onboard computer in which ROS is loaded on to, performs a self-
diagnostics check to ensure that all components, including the servo motor,
LiDAR module and Inertial Measurement units, i.e. the Gyroscopes, the
Accelerometers, the Magnetometer are all to the functional standard.

The command for the robot is given either from a computer or a joystick, based
on which the onboard computer commands the related subsystems to action. The
LiDAR sensor scans the environment for providing a real-time map which is
used to identify obstacles and plot optimal paths. The IMU stabilizes the robot
properly so that it doesn’t lose its balance or fall while going through some
stringent movements.

The robot uses inverse kinematics calculations to make the required joint angles
for its legs, enabling seamless locomotion. Once the task is completed, the robot
returns to its HOME position.

4.2 PROJECT FLOW

● Before turning the robot ON, its external structural conditions are to be
assessed to ensure that there isn’t any damage or any failure causing
obstructions.
 16

● On powering ON, the onboard computer runs a self-diagnosis.

● All the nodes of the ROS framework are initialized to create a
communication between all the components of the robot.
● Commands such as walking, trotting are specified by the remote
controller, the computer or joystick, to the robot. This is sent to On board
computer via the ROS framework.
● The IMU constantly captures the orientations of the robot, to create a real
time feedback loop for the Onboard computer to adjust the movements
accordingly.
● The Leg movement is achieved using Inverse Kinematics algorithms
which calculates the joint angles.
● After the completion of the given command, the robot returns to its
HOME configurations.

Figure 4.1: Block diagram

 17

4.4 CALCULATION

𝑓𝑜𝑜𝑡 = 0.185 𝑚
ℎ𝑖𝑝 = 0.075 𝑚
𝑓𝑜𝑜𝑡 = ℎ𝑖𝑝 = 0.23 𝑚
Elbow length (EL) = 0.106 m
Wrist length (WL) = 0.145 m
Shoulder length (SL) = 0.055 m

4.4.1. INVERSE KINEMATICS

Shoulder Angle(ϴ)
𝜃 = 𝑡𝑎𝑛2 (𝑓𝑜𝑜𝑡 − ℎ𝑖𝑝 , 𝑓𝑜𝑜𝑡 − ℎ𝑖𝑝 ) (4.1)

𝜃 = 𝑡𝑎𝑛2 (0.185 − 0.075 , 0.23 − 0.23)

𝜃 = 𝑡𝑎𝑛2 (0.11,0) ⇒ radians

Shoulder Angle (SA) 𝜃 = 𝑜𝑟 90 ˚

Elbow Angle (α)

𝐸𝐿 + 𝑊𝐿 − 𝑑
𝛼 = 𝑐𝑜𝑠
2 × 𝐸𝐿 × 𝑊𝐿
(4.2)

𝑑= ((𝑓𝑜𝑜𝑡 − ℎ𝑖𝑝 ) + 𝑓𝑜𝑜𝑡 − ℎ𝑖𝑝 ) (4.3)

𝑑= ((0) + (0.11) ) = 0.11

 18

0.106 + 0.145 − 0.11

𝛼 = 𝑐𝑜𝑠
2 × 0.106 × 0.145

= 49.015 ˚ ≅ 49˚

Wrist Angle (WA) = 𝑟𝑎𝑑𝑖𝑎𝑛𝑠 𝑜𝑟 49 ˚

Wrist Angle(β)

𝑓𝑜𝑜𝑡 − ℎ𝑖𝑝 − 𝑆𝐿 × 𝑠𝑖𝑛 (𝑆𝐴)

𝛽 = 𝑡𝑎𝑛2 𝑓𝑜𝑜𝑡 − ℎ𝑖𝑝 − 𝑆𝐿 × cos(𝑆𝐴)
,
(4.4)
𝑆𝐴 − 𝐸𝐴

Substituting the values we get

0.11 π 4π
β = 𝑡𝑎𝑛2 , −
0 − 0.55 X 0 2 15

𝑡𝑎𝑛2 0, −90˚ − 49˚

𝑡𝑎𝑛2 0, −139˚ = 0˚ 𝑜𝑟 0 𝑟𝑎𝑑𝑖𝑎𝑛𝑠

 19

4.4.2. FORWARD KINEMATICS

Elbow Joint

𝑒𝑙𝑏𝑜𝑤
= ℎ𝑖𝑝 + 𝑠ℎ𝑜𝑢𝑙𝑑𝑒𝑟 𝑙𝑒𝑛𝑔𝑡ℎ × 𝑐𝑜𝑠 (𝜃) (4.5)

𝑒𝑙𝑏𝑜𝑤
= ℎ𝑖𝑝 + 𝑠ℎ𝑜𝑢𝑙𝑑𝑒𝑟 𝑙𝑒𝑛𝑔𝑡ℎ × s𝑖𝑛 (𝜃) (4.6)

𝑒𝑙𝑏𝑜𝑤
= 0.23 + 0.055 × 𝑐𝑜𝑠 =0

𝑒𝑙𝑏𝑜𝑤
𝜋
= 0.75 + 0.055 × 𝑠𝑖 𝑛 = 0.625
2

Elbow Position (0.23,0.625)

Wrist Joint

𝑤𝑟𝑖𝑠𝑡
= 𝑒𝑙𝑏𝑜𝑤 + 𝑒𝑙𝑏𝑜𝑤 𝑙𝑒𝑛𝑔𝑡ℎ × 𝑐𝑜𝑠 (𝛼) (4.7)

𝑤𝑟𝑖𝑠𝑡
= 𝑒𝑙𝑏𝑜𝑤 + 𝑒𝑙𝑏𝑜𝑤 𝑙𝑒𝑛𝑔𝑡ℎ ×𝑠𝑖𝑛 𝑠𝑖𝑛 (𝛼) (4.8)

𝑤𝑟𝑖𝑠𝑡
= 0.23 + 0.106 × 𝑐𝑜 𝑠(49˚) = 0.299

𝑤𝑟𝑖𝑠𝑡
= 0.625 + 0.106 × 𝑠𝑖𝑛 (49˚) = 0.704

Wrist Position (0.299, 0.704)

 20

Foot

𝑓𝑜𝑜𝑡
= 𝑒𝑙𝑏𝑜𝑤 + 𝑤𝑟𝑖𝑠𝑡 𝑙𝑒𝑛𝑔𝑡ℎ × 𝑐𝑜𝑠 (𝛽) (4.9)
𝑓𝑜𝑜𝑡
(4.10)
= 𝑒𝑙𝑏𝑜𝑤 + 𝑤𝑟𝑖𝑠𝑡 𝑙𝑒𝑛𝑔𝑡ℎ × 𝑠𝑖𝑛 (𝛽)

𝑓𝑜𝑜𝑡 = 0.299 + 0.145 × 𝑐𝑜 𝑠(0˚) = 0.44

𝑓𝑜𝑜𝑡 = 0.704 + 0.145 × 𝑠𝑖𝑛 (0˚)=0.70

Foot Position (0.44, 0.70)

 21

4.5 COMPONENTS USED

A list and description of all components used are mentioned below:

(a) (b)

(c) (d)

(e) (f)
 22

(g)
Figure 4.2: Overview of components used

4.5.1 Raspberry Pi 4:

Serves as the main processing unit of the quadruped robot, responsible

for executing control algorithms, handling sensor data, and managing
communication between various modules. As shown in Figure 4.2 (a) It is
chosen for its ability to process complex tasks like gait generation, real-time
feedback, and motion planning, all essential for smooth robotic locomotion.
With built-in Wi-Fi and USB support, it also allows for easy debugging and
external control.

4.5.2 11.1V LI- PO Battery

Provides the primary power source for the robot, supplying sufficient
current to drive multiple high-torque servo motors and electronic modules. The
11.1V output is ideal for mobile robotics, offering a balance between power and
 23

weight, and ensuring stable operation throughout the robot’s motion [Link]
shown in Figure 4.2 (b)

4.5.3 MG995 Servo Motor

Used for the actuation of the quadruped robot’s legs, enabling continuous
rotational motion to produce forward, backward, or turning movement. as shown
in Figure 4.2 (c). The high torque and robust build of the MG995 make it suitable
for supporting the mechanical load of each limb while ensuring precise and
smooth locomotion.

4.5.4 LM2596HVS DC-DC Buck Converter (4.5–50V TO 3–40V)

As shown in Figure4.2 (d) This adjustable step-down voltage regulator is

used to convert the 11.1V output from the Li-Po battery to safe operating
voltages (such as 5V or 6V) for the Raspberry Pi, servo motors, and other
electronic modules. It ensures stable voltage levels, protecting components from
over-voltage damage.

4.5.5 MPU9250 9-Axis Sensor Module

Used for sensing orientation, acceleration, and rotational velocity of the

robot in real time. The MPU9250 combines a 3-axis gyroscope, 3-axis
accelerometer, and 3-axis magnetometer, making it ideal for maintaining balance
and adjusting the posture of the quadruped on uneven terrain. as shown in Figure
4.2 (e).
 24

4.5.6 16-channel 12-bit PWM/ Servo Driver (PCA9685)

As shown in Figure 4.2 (f) This module is responsible for generating

precise PWM signals to control up to 16 servo motors simultaneously. It is
interfaced with the Raspberry Pi to offload the PWM generation task, allowing
better timing control and reducing CPU load. This ensures synchronized leg
movements in the robot.

4.5.7 3-D Printed Parts

As shown in Figure 4.2 (g) The structural components of the quadruped

robot, such as limb segments, brackets, and the main frame, were designed using
SolidWorks and Fusion 360 CAD software for precision modelling and assembly
simulation. These parts were fabricated using PLA (Polylactic Acid), a
biodegradable thermoplastic commonly used in 3D printing due to its ease of
processing, lightweight nature, and adequate mechanical strength for non-load-
bearing robotic structures
 25

CHAPTER 5
RESULT AND DISCUSSION

5.1 MODEL ARCHITECTURE

A schematic layout of the developed quadruped robot is illustrated in

Figure 5.1, showcasing the mechanical frame, control electronics, and power
module. The system’s core functionality is driven by servo motors at each leg
joint, controlled through a Programmable Automation Controller (PAC) module
interfaced with a Raspberry Pi 4. The Raspberry Pi functions as the on-board
computer (OBC), executing predefined walking algorithms and transmitting
control signals to the servo motors for leg movement.

The robot was designed and modelled using SolidWorks, ensuring accurate joint
placement and stability analysis. Each leg is programmed for coordinated
movement based on a gait pattern pre-coded into the Raspberry Pi, offering
walking capability on flat terrain.

Figure 5.1: Final Prototype Model of Quadruped

Robot
 26

5.2 ANALYSIS AND DISCUSSION

The prototype was tested on various flat surfaces, including tiled, wooden,
and concrete flooring. The control system successfully achieved basic
locomotion such as forward walking, standing, and turning.

Motor Performance and Control: The servo motors showed consistent and
responsive movement under control of the PAC module. The use of the PAC
improved signal integrity and reduced latency between command and action,
which was essential for maintaining balance. As shown in figure 5.2

Figure 5.2: Servo Angle vs Time Graph

Stability Observations: As shown in figure 5.3The quadruped robot maintained

good static stability during slow gait sequences. Dynamic stability under faster
motion requires further refinement in control algorithms, which is expected due
to the absence of sensors like gyroscopes or inertial measurement units (IMUs).
 27

Figure 5.3: Stability diagram of Quadruped

Robot

Limitations Identified: No obstacle detection or terrain adaptation due to lack

of LiDAR or other environment mapping tools. Manual tuning of gait is required
for different surface types.. Limited autonomy, as the robot executes pre-
programmed movement sequences without external feedback

Strengths of the Prototype: As shown in figure 5.4, Compact and modular design
for ease of assembly and maintenance. Cost-effective use of Raspberry Pi and
off-the-shelf servo motors .Scalable platform for future integration of sensors
and AI modules

Figure 5.4: GAIT trajectory of Quadruped

Robot
 28

Table 5.1: Gait Test Results on Various Terrains

Stability
Surface Type Successful Walk Cycles
Rating(1–5)

Tiled Floor
7/10 5

Wooden Floor
6/10 4

Concrete Surface
6/10 4

Carpeted Floor
4/10 3

Table 5.2: Power Consumption Summary

Avg
Activity Current Draw Voltage (V) Power (W) Duration (min)
(A)
Idle State 0.3 5.0 1.5 -
Standing Posture 0.5 5.0 2.5 Continuous
Walking (forward
1.0 5.0 5.0 ~10 min/test
gait)
Turning
1.2 5.0 6.0 ~5 min/test
Maneuverer

Additionally, increasing the payload capacity and incorporating advanced AI

algorithms could expand the robot’s applicability in tasks such as logistics,
search and rescue, and industrial automation. Overall, the results demonstrate the
robot's effectiveness as a versatile and reliable solution for complex operations.
 29

CHAPTER 6
CONCLUSION

6.1 CONCLUSION

The development of the quadruped robot demonstrates a significant

advancement toward agile and terrain-adaptive robotic systems. Unlike
traditional wheeled or tracked robots, this quadruped robot exhibits superior
stability and mobility across uneven and unpredictable surfaces. The integration
of servo motors through the PAC module to the Raspberry Pi 4 enabled precise
control and real-time responsiveness, laying the groundwork for dynamic
locomotion. Although LiDAR-based mapping and vision systems have not yet
been integrated, the current configuration provides a reliable platform for future
enhancements. The design, simulation, and partial automation of the robot pave
the way for practical applications in sectors such as logistics, surveillance, and
search-and-rescue. This project provides a scalable and modular foundation for
continued innovation, where improvements such as sensor integration, vision-
based navigation, and enhanced gait algorithms could further elevate the robot's
capability in real-world operations.

6.2 SCOPE OF FUTURE WORK

Although the existing prototype achieves terrain-adaptive locomotion by

controlling servo motors and onboard processing via Raspberry Pi 4, several
improvements can be made to enhance functionality, autonomy, and resilience:

LiDAR and Camera System Integration: Adding sensors like LiDAR and stereo
cameras is possible, allowing for obstacle detection, mapping of the
environment, and visual navigation, which will hugely enhance autonomy.

SLAM and AI Algorithm Implementation: Future research can involve

Simultaneous Localization and Mapping (SLAM) for real-time navigation and
terrain analysis using AI for adaptive gait control in changing environments.
 30

Wireless Communication and Remote Control: Integrating wireless modules like

Wi-Fi, Bluetooth, or RF can enable remote command execution, telemetry
feedback, and coordination among multiple robots.

Mechanical Optimization: Structural design improvements with the use of light,

high-strength materials and advanced 3D-printed joints can improve mobility
and decrease energy usage.

Power Management Systems: The use of efficient battery systems and energy
management circuits can increase operational time, making the robot ideal for
long-duration missions.

Modular Payloads: Future models can be equipped with payloads like

manipulators, environmental sensors, or delivery containers for application in
inspection, Défense, and disaster relief missions.

These improvements will assist in transforming the robot from a working

prototype to a field-deployable, smart robotic system
 31

REFERNCES

1. Biswal, Priyaranjan, and Prases K. Mohanty, Development of

Quadruped Walking Robots: A Review, 2014.
2. Caluwaerts, Ken, Atil Iscen, J. Chase Kew, Wenhao Yu, Tingnan
Zhang, Daniel Freeman, Kuang-Huei Lee, et al., Barkour:
Benchmarking Animal-Level Agility with Quadruped Robots, 2023.
3. Chang, Shuai, Gait Planning of Quadruped Robot Based on ROS,
2020.
4. El-Dalatony, Ahmed M., Tamer Attia, H. Ragheb, and A. M. Sharaf,
Development, Estimation and Control of a Bio-Inspired Quadruped
Robotic Leg, 2022.
5. García-Cárdenas, Facundo, Nelson Soberón, Oscar E. Ramos, and
Ruth Canahuire, Charlotte: Low-cost Open-source Semi-Autonomous
Quadruped Robot, 2020.
6. Han, Yangyang, Huaizhi Zong, Zhenyu Lu, Guoping Liu, Shengyun
Zhou, Liping Cao, and Junhui Zhang, Enhancing Quadruped Robot
Locomotion on Deformable Terrains Through Contact Perception and
Terrain Classification, 2024.
7. He, Tairan, Chong Zhang, Wenli Xiao, Guanqi He, Changliu Liu, and
Guanya Shi, Agile But Safe: Learning Collision-Free High-Speed
Legged Locomotion, 2024.
8. Kalyoncu, Mete, Inverse Kinematic Analysis of a Quadruped Robot,
2017.
9. Kimura, Hiroshi, Isao Shimoyama, and Hirofumi Miura, Dynamics in
the Dynamic Walk of a Quadruped Robot, 2016.
10. Kumar, S., H. Patel, and R. Sharma, Structural Optimization and
Material Selection for Quadruped Robot Design, 2020.
11. Li, Yibin, Bin Li, Jiuhong Ruan, and Xuewen Rong, Research of
Mammal Bionic Quadruped Robots: A Review, 2011.
 32

12. Raiola, Gennaro, Michele Focchi, and Enrico Mingo Hoffman,

WoLF: The Whole-body Locomotion Framework for Quadruped
Robots, 2012.
13. Rashtian, H., J. Zhang, and R. Dantu, Onboard Computing for
Autonomous Robots: Challenges and Opportunities, 2020.
14. Rong, Xuewen, Yibin Li, Jiuhong Ruan, and Bin Li, Design and
Simulation for a Hydraulic Actuated Quadruped Robot, 2011.
15. Sen, Muhammed Arif, Veli Bakircioglu, and Mete Kalyoncu, Inverse
Kinematic Analysis of a Quadruped Robot, 2020.
16. Siregar, Baihaqi, Hagel Bagustiawan, and Opim Salim Sitompul,
Environmental Mapping Automation by Quadruped Robot Using
Lidar Technology, 2023.
17. Sun, Yujin, Haoran Sun, Wenhao Yu, Tao Chen, Jie Tan, and Byron
Boots, Learning Agile Locomotion Skills with a New Dog-Like
Robot, 2025.
18. Tan, J., T. Zhang, and C. Liu, Learning Locomotion Control for
Quadrupedal Robots with Onboard Computing Constraints, 2021.
19. Yun, Z., X. Li, Y. Zhao, B. Zhao, and Z. Yang, A Study of
Lightweight, Low-cost Quadrupedal Robot Body Based on a Coaxial
Deformation Mechanism, 2024.
33
34
35