PROJECT REPORT – 4

1. Project title and summary

 SRM Institute of Science and Technology
College of Engineering and Technology
Department of ECE

AY 2018-2019
15EC496L -Major Project Details

Sl No Register No Students Name(s) Project Supervisor Project Title

RA1511004010107 Pooja Anand

Low Cost Digitalization (Industry 4.0) Solution for Siemens
1 Dr. P. Eswaran Sinumerik CNC System to Increase the Transparency and Utilization
RA1511004010059 Vinitha Lea Philip of the Machine.

RA1511004010553 kedar prasad karpe

RA1511004010511 Dhruv pant Scalable Cooperative Transport Strategy Using A Group Of Simple
2 DR. R .Kumar Robots
RA1511004010712 Nimish pastaria
RA1511004010654 jayati singh
 SRM Institute of Science & Technology
College of Engineering and Technology
Department of Electronics and Communication Engineering
Project Summary - 2018-2019

Sl Students Name Project Project Title Objective of the Realistic constraints Standards to be Multidisciplina Outcome
N Guide Project imposed referred/follow ry tasks
o ed involved
1 VINITHA LEA Dr. P. Low Cost To build a solution Safe place has to be Open user 1) Electrical Journal
PHILIP [Reg Eswaran Digitalization that is economical, can found in each machine interface design and Publication
No:RA151100401005 (Industry 4.0) be adopted by small to place the KIT & based on Electronics SNIP:0.354
9] Adaptors, so that no Engineering
Solution for and medium WinCC or Run
damage happens to the for utilizing
POOJA ANAND Siemens enterprises so that they MyHMI Raspberry
Sinumerik get a taste of how IoT kit or connecting
[Reg pi
CNC System can be adopted by cables, by machine Up to 10
No:RA151100401010 2) Computatio
operator or any other
7] to Increase the monitoring some of machining nal and IT
activity on the field for
Transparency the critical machine channels per
machine. programmin
and Utilization parameters thereby NCU
g the
of the Machine trying to reduce or Power fluctuations or The digital twin
raspberry pi
prevent breakdowns of failures may affect the – end-to-end using
machines and normal working of the development Python
associated kit. and new 3) Desktop
productivity. business models publication
for report.
2 KEDAR PRASAD Dr R. SCALABLE The develop a Calibration ISO 10218- Microcontroller IEEE
KARPE [Reg No: Kumar COOPERATI solution to the 1:2011 architecture for Conference
RA1511004010553] VE cooperative transport Sensors: handling (Malaysia)
TRANSPORT problem is based on an communication
DHRUV PANT [Reg STRATEGY articulated drive model Registration and motor IEEE
No: USING A where the group of control Conference
RA1511004010511] GROUP OF robots has leader and Modeling (Brasil)
SIMPLE multiple follower SPRINTER
JAYATI SINGH [Reg ROBOTS robots. NIST will produce robot which
No: robot models, datasets, represents all
RA1511004010654] software tools, and the separate
calibration artifacts components of
NIMISH PASTARIA that can lead to easily the mechanical
[Reg No: calibrated or even self- design
RA1511004010712] calibrating sensors and
robots.
PROJECT REPORT – 4
2. Project report
 SCALABLE COOPERATIVE TRANSPORT
STRATEGY USING A GROUP OF SIMPLE ROBOTS

A PROJECT REPORT

Submitted by

KEDAR PRASAD KARPE [Reg No: RA1511004010553]

DHRUV PANT [Reg No: RA1511004010511]
JAYATI SINGH [Reg No: RA1511004010654]
NIMISH PASTARIA [Reg No: RA1511004010712]

Under the guidance of

Dr. R. Kumar, Ph.D
(Professor, Department of Electronics and Communication Engineering)

in partial fulfillment for the award of the degree

of

BACHELOR OF TECHNOLOGY
in
ELECTRONICS AND COMMUNICATION
ENGINEERING
of

FACULTY OF ENGINEERING AND TECHNOLOGY

S.R.M. Nagar, Kattankulathur, Kancheepuram District

MAY 2019
SRM INSTITUTE OF SCIENCE AND TECHNOLOGY
(Under Section 3 of UGC Act, 1956)

BONAFIDE CERTIFICATE

Certified that this project report titled "SCALABLE COOPERA-

TIVE TRANSPORT STRATEGY USING A GROUP OF SIMPLE
ROBOTS" is the bonafide work of "KEDAR PRASAD KARPE [Reg
No: RA1511004010553], DHRUV PANT [Reg No: RA1511004010511],
JAYATI SINGH [Reg No: RA1511004010654], NIMISH PASTARIA
[Reg No: RA1511004010712]", who carried out the project work under
my supervision. Certified further, that to the best of my knowledge the
work reported herein does not form any other project report or disserta-
tion on the basis of which a degree or award was conferred on an earlier
occasion on this or any other candidate.

Date: Project Supervisor Head of the Department

Submitted for University Examination held on in the Department of Elec-

tronics and Communication Engineering, SRM Institute of Science and
Technology, Kattankulathur.

Date: Internal Examiner External Examiner

 DECLARATION

We hereby declare that the project titled "SCALABLE COOPERA-

TIVE TRANSPORT STRATEGY USING A GROUP OF SIMPLE
ROBOTS" to be submitted for the Degree of Bachelor of Technology is
our original work as a team and the dissertation has not formed the ba-
sis of any degree, diploma, associateship or fellowship of similar other
titles. It has not been submitted to any other University or institution for
the award of any degree or diploma.

Date:
Place:

Kedar Prasad Karpe

[Reg No: RA1511004010553]

Dhruv Pant
[Reg No: RA1511004010511]

Jayati Singh
[Reg No: RA1511004010654]

Nimish Pastaria
[Reg No: RA1511004010712]

iii
 ABSTRACT

The transport of goods is a commonly observed phenomena in factories

and warehouses, airports and waterways, and even in nature such as in
ant colonies. One of the major challenges encountered while transport-
ing an object is that the object may be too heavy or too big for a single
entity to transport. However, it is often observed in the nature that social
animals such as ants, bees, etc. often exploit cooperative behaviors to
overcome this problem. Besides being able to transport large and heavy
objects, such behaviors also offer the advantage of being robust to indi-
vidual failure, better planning of task and optimization of energy.

Cooperative transport using multi robot systems is a widely addressed

research topic due to two major advantages; It is scalable: If the weight
of the object to be transported increases, simply adding more robots to
the system overcome the weight barrier; It is fault tolerant: If one or two
robots fail, the others can still complete the assigned task. Sometimes, a
substitute robot can also join the workforce.

In this project, we develop a control strategy to perform a cooper-

ative transport of rigid objects from one point to another, by a group
of simple and miniature robots. The goal of this project is to improve
the state-of-the-art on cooperative transport which, at present, requires a
large number of complex robot features that prevent such a system from
being economical and scalable.
 ACKNOWLEDGEMENTS

I would like to express my deepest gratitude to my guide, Dr. Lorenzo

Sabattini, Associate Professor, DISMI, University of Modena and Reg-
gio Emilia, Italy for his valuable guidance and supervision towards this
project.

I also thank Dr. R. Kumar, Professor, and Dr. Dhanalakshmi Samiap-

pan, Associate Professor, of the Department of Electronics and Com-
munication Engineering, SRM Institute of Science for constant feed-
back and unconditional support. This project would have been impos-
sible without the assistance of Ayon Chatterjee and Pranav Srinivas of
Beeclust - Multi Robot Systems Lab. I thank Ayon for his help with the
mechanical design and Pranav for his valuable feedback on the hardware
design of the project.

Kedar Prasad Karpe

Dhruv Pant
Jayati Singh
Nimish Pastaria

v
 TABLE OF CONTENTS

DECLARATION iii

ABSTRACT iv

ACKNOWLEDGEMENTS v

LIST OF TABLES viii

LIST OF FIGURES ix

ABBREVIATIONS x

LIST OF SYMBOLS xi

1 INTRODUCTION 1

2 State of the Art 3

3 Problem Formulation 5

4 Simulation 7
4.1 Background on V-REP and MATLAB . . . . . . . . . . . . . . . . 7
4.2 E-Puck Robot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
4.3 System Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

5 Testbed Architecture 10

6 Robot Design 11
6.1 Mechanical Design . . . . . . . . . . . . . . . . . . . . . . . . . . 11
6.2 Electronics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
6.2.1 Architechture . . . . . . . . . . . . . . . . . . . . . . . . . 14
6.2.2 Fabrication and Assembly . . . . . . . . . . . . . . . . . . 14

vi
 6.2.3 Programming . . . . . . . . . . . . . . . . . . . . . . . . . 15
6.2.4 Cost Analysis . . . . . . . . . . . . . . . . . . . . . . . . . 15

7 Localisation 17
7.1 AprilTag . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
7.2 Calibration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

8 Experiments 20
8.1 Pseudogripper . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

9 Conclusion 22

10 Future Enhancement 23
 LIST OF TABLES

6.1 Cost Per Robot . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

viii
 LIST OF FIGURES

3.1 Visual representation of the proposed strategy. . . . . . . . . . . . . 5

3.2 Visual representation of the articulated drive model. . . . . . . . . . 6
3.3 Go-to-goal proportional controller of the follower robots. . . . . . . 6

4.1 E-puck robot as seen in V-REP robot simulator. . . . . . . . . . . . 8

4.2 Screenshot of the simulation experiment. The highlighted robot was
randomly assigned the role of the leader while the other robots were
assigned roles of followers. . . . . . . . . . . . . . . . . . . . . . . 8
4.3 Tracking data of the leader robot and the centre of mass of the object
being transported . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

5.1 The architecture of the multi-robot testbed used for physical experi-
ments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

6.1 Complete model of robots used for experiments. . . . . . . . . . . . 11

6.2 Exploded view of the robot. . . . . . . . . . . . . . . . . . . . . . . 12
6.3 Disassembled robot chassis depicting the two fragments of robot de-
sign. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
6.4 Architechture of the robot circuit board. . . . . . . . . . . . . . . . 14
6.5 The spring-loaded programmer. . . . . . . . . . . . . . . . . . . . . 15
6.6 Programming the microcontroller. . . . . . . . . . . . . . . . . . . 15
6.7 Atmel-ICE programmer with the 6 pin spring loaded ISP programming
interface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

7.1 Screenshot of the AprilTag tracking software, detecting the actual robots
on the testbed. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
7.2 AprilTag TagID = 0 from the Tag36h11 family. . . . . . . . . . . . 18
7.3 Screenshot of checkerboard calibration. . . . . . . . . . . . . . . . 19

8.1 Experiments perform on the testbed. . . . . . . . . . . . . . . . . . 20

8.2 Three robots placed into the pseudo gripper of an acrylic object. . . 21

ix
ABBREVIATIONS

x
 LIST OF SYMBOLS

α, β Damping constants
θ Angle of twist, rad
ω Angular velocity, rad/s
b Width of the beam, m
h Height of the beam, m
{f (t)} force vector
[K e ] Element stiffness matrix
[M e ] Element mass matrix
{q(t)} Displacement vector
{q̇(t)} Velocity vector
{q̈(t)} Acceleration vector

xi
 CHAPTER 1

INTRODUCTION

Multi-robot systems are a promising alternative to automate tasks that are beyond the
competency of single robot systems. Transporting big objects, surveillance of vast ar-
eas, or robot tasks that can be decomposed into smaller tasks so that they can be carried
out simultaneously by several robots are examples of application domains particularly
suited for multi-robot systems. In addition, multi-robot systems, comprised of many but
simple individuals, may be cheaper to build and easier to program than a complex robot
capable of performing similar tasks. Multi-robot systems are also potentially more re-
silient to a large variety of hardware or software failures; when one robot fails or makes
a mistake, the others can still complete the task successfully.

Cooperative multi-robot systems are generally employed when the object is too
heavy, too large, or has a complex shape so that it can not be transported by a single
robot. However, this is not a strict requirement; not all group members need to partici-
pate in the physical act of transport; carrying or pushing/pulling the object. Cooperation
can still be achieved when a single or few robots transport the object, and the others plan
the coordination and navigation of the transporters along a desired trajectory, or clear
the way from obstacles.

Cooperative transport strategies have been mainly classified into three fundamental
types:

1. Pushing: Pushing-only strategies are methods of mutually transporting items by

applying pushing forces on the item. These ways of strategies are primarily em-
ployed by robots that cannot pull objects, because they have no means to grasp
them. Pushing-only strategies may appear to be relatively simple methods of co-
operative transport. However, on top of the challenges common to all transport
strategies like the alignment of forces required to initiate the transport, pushing-
only strategies also require a significant amount of coordination of actions to bear
the transport.

2. Grasping: Grasping strategies involve methods in which the robots physically

attach themselves to an item to be able to collectively transport it. Thus, grasping
 strategies can only be exploited by robots which possess the ability and mecha-
nisms to grasp an object. There exist several mechanisms that allow a robot to
physically connect to an object, some of which allow the robots not only to grasp
but also to lift an item. Compared to pushing-only strategies, grasping strategies
provide a better control over the transported object, since once grasped, the object
can be either pushed or pulled.

3. Caging: Cooperative transport by caging is a special case of the previously con-

cluded pushing-only strategy where robots intentionally entrap the object to en-
sure the object follows the group movements. In the caging strategy, robots ar-
range themselves around the object to form a “closure” that traps the object. The
closure must be maintained during transport to ensure the object does not escape
from the robots’ cage. In cooperative transport based on a caging strategy, the
object’s shape and size are particularly important features since they bear upon
the minimum number of robots required to surround the object. In generally used
caging strategy the major disadvantage is the fact that if the object to be trans-
ported is convex in nature then caging fails.

2
 CHAPTER 2

STATE OF THE ART

A significant amount of research has been conducted in the field of cooperative transport
by robotic systems [4], [5], [6]. Most of the work [2], [8] has addressed solutions
related eliminating communication dependency from such a strategy to make multi-
robot transport more viable in a real time environment. Some work [1] [3] has also
discussed manipulation strategies for specific type of objects as the shape posesses a
major barrier in cooperative transport. This project will focus on optimizing the net
forces such that their contribution towards the translation of the object being transported
is maximized.

In this chapter, we shall study three different strategies of multi-robot transport and
discuss their implications with respect to the shape of object that can be transported, the
dependency on explicit communication and the force optimization in the system.

[8] discusses a communication-less decentralised control strategy for multi-robot

transport. It employs a single leader system in which the object being transported is
used as a means of extrapolating system states. The dynamics of the leader robot are
modelled using explicit path planning whereas the dynamics of follower robots con-
verge using the well known consensus protocol. Force feedback measurements are
used to model a force controller for the robots which is used as a means of positional
manipulation.

[2] proposes a decentralised strategy in which a large group of robots push an object
at points where the direct line of sight to the goal is occluded. Though this strategy does
not require explicit communication or manipulation mechanisms, it largely depends on
robot hardware for the visual computation. Though it is proved that pushing the object
at such discrete points eventually leads it to the goal, the author’s make an assump-
tion that the object being transported must be convex in shape. In addition, the author
makes an assumption the environment being an uncluttered one. In case of a cluttered
environment, the author presents an experimental analysis using a teleoperate robot as
a temporary goal location.

[7] exploits a decentralised behaviour to achieve transport without explicit coordi-

nation. The authors compares the dynamical model to the behaviour of ant colonies. A
leader robot points to the direction of the goal, similar to a compass. All followers con-
verge to the same direction and apply parallel force vectors to this. Using this method,
it is proven that the centre of mass of the object moves in a straight line towards the
goal. The author experimentally validates the hypothesis using multiple test cases.

In this project, we develop a strategy which is dynamic to the shape of the object
being transported, while at the same time, it optimises the net force acting on the object.

4
 CHAPTER 3

PROBLEM FORMULATION

The proposed solution to the cooperative transport problem is based on an articulated

drive model. We divide the group of robots into a leader and multiple follower robots.
The leader is chosen randomly and does not affect the dynamics of the system. Let us
assume that the leader robot has the knowledge of the path to the goal. According to the
proposed strategy, if the follower robots apply forces in the direction of the leader robot,
then the object can be successfully transported between the start and the goal locations.
Figure 3.1 shows the visual representation of the proposed strategy.

Figure 3.1: Visual representation of the proposed strategy.

Furthermore, we predict that if the model follows the proposed algorithm, then the
system behaves similar to an articulated drive robot/vehicle. The articulate joint exists
between the leader robot and the object being transported and if the heading of the leader
robot changes, then the heading of the object slowly converges to the same value.

Figure 3.2: Visual representation of the articulated drive model.

While the leader robot’s controller follows a predefined path plan, the follower
robots’ use a simple go-to-goal controller, where the goal is defined as the centroid
of the leader robot. The follower robots’ controller can be seen in Figure 3.3

Figure 3.3: Go-to-goal proportional controller of the follower robots.

6
 CHAPTER 4

SIMULATION

In this chapter, we discuss about the simulation experiments that were performed to
verify the hypothesis made in the previous section.

To validate the predictions of the porposed control strategy, we performed physical

simulations using the well known Open Dynamics Engine. The control strategy was
implemented on V-REP robot simulator while analysis was performed on MATLAB
client.

4.1 Background on V-REP and MATLAB

V-REP or Virtual Robot Experimentation Platform, provides an integrated development

environment for robotics and physics simulations. It offers the Open Dynamics Engine
for fast and customisable dynamics calculations. V-REP allows six different program-
ming modes. It can be programmed using MATLAB in remote API mode. We used this
server-client configuration for performing the simulations where MATLAB was used as
an analysis tool.

4.2 E-Puck Robot

The e-puck is a differential drive robot developed by the ASL laboratory at EPFL,Switzerland.
The robots hardware and firmware are both open-source. We used the e-puck robots to
perform simulations in V-REP due to the similarity in their dynamics to the robots used
in the physical experiments of this project.
 Figure 4.1: E-puck robot as seen in V-REP robot simulator.

4.3 System Analysis

To analyse the predicted behavior of the system, we performed simulations with five
e-puck robots. A random robot was chosen as a leader, while the rest were assigned
as the follower roles. A CAD model of a dummy object was created and imported to
V-REP for the experiment. As discussed in the previous sections that the leader has
complete knowledge of the path between the start and destination points, the leader was
made to follow a line on the simulation floor. The follower controller, discussed earlier
was implemented in each follower robot.

Figure 4.2: Screenshot of the simulation experiment. The highlighted robot was ran-
domly assigned the role of the leader while the other robots were assigned
roles of followers.

Both the leader’s position and the position of the center of mass of the object were
tracked during the length of the simulation by the MATLAB client. Figure 4.3 shows
the heading data of the system. We can notice from the figure that the object’s position
and the leader’s heading vary widely till about 16-17 seconds of simulation time. This
phenomenon occurs intilailly till both the headings are not aligned. After this point,
both the headings start following each other until the end of the simulation. A very

8
minute tracking error is induced in the system each time the leader takes a turn, but his
error slowly converges to zero in a finite time.

This behavior can be compared to an articulated drive model where the active joint
(which is driven by the motors) follows the heading of the passive joint (which is being
pulled). As predicted by the hypothesis, it can be empirically noted that the system
behaves like an articulated drive robot/vehicle.

Figure 4.3: Tracking data of the leader robot and the centre of mass of the object being
transported

9
 CHAPTER 5

TESTBED ARCHITECTURE

A multi-robot testbed was developed to perform the physical experiments. Figure 5.1
shows the different parts of the multi-robot testbed. It consists of the following compo-
nents:

1. Robots: The testbed uses differential drive robots which were developed for the
physical experiments due to the ease in modelling such systems. Chapter 6 de-
scribes the robot design in detail.

2. Tracking System: The tracking system uses AprilTag fiducial markers to localise
the robots on in the arena. A generic USB webcam was used to track the robots.

3. Computer: The computer is used as a bridge between the robots and the tracking
system. It processes the tracking data and sends it to the RF transceiver.

4. RF Transceiver: The RF transceiver is used to establish wireless communication

between the computer and robots.

5. Robot Arena: The robot arena is constructed out of wooden logs. It was no-
ticed that lighter colored background improved tracking distance and hence it is
recommended that the testbed is white colored.

Figure 5.1: The architecture of the multi-robot testbed used for physical experiments.
 CHAPTER 6

ROBOT DESIGN

In this chapter, we will discuss the hardware and mechanical aspects of the robot design.
The robots used in the experiments were indigenously developed to accomodate the
limited size of the tracking system and also to save costs. A differential drive robot was
developed for this purpose due to the effective ease of modelling and controlling such a
system.

Figure 6.1: Complete model of robots used for experiments.

6.1 Mechanical Design

The chassis of the robot was designed as a two-fragment structure for easy access to
the programming interface and to the battery. The base chassis, holds the motors and
the balancing castor wheels, and the diffuser cap which houses the circuit board of the
robot. The chassis was 3D printed using PLA. Figure 6.2 shows the exploded view of
the robot.

1. Diffuser Lid: 3D printed cap, which houses the printed circuit board. The dif-
fuser is used to indicate robot states using the RGB light emitting diode on the
circuit board.
Figure 6.2: Exploded view of the robot.

12
 2. Printed Circuit Board: 1.2mm FRC circuit board, which houses the electricals
of the robot. Includes 6 proximity sensors along the perimeter for obstacle avoid-
ance and SLAM.

3. LiPo Battery: 3.7V 380mAh single cell lithium ion battery. On board power
management can charge the battery without the need of an external charger.

4. Motor Lid: 3D printed cap to prevent vibrations from motor movements.

5. Rubberised Wheels: 12mm 3D printed wheels wrapped with rubber rings to

avoid slippage.

6. Motors: 10mm 3V Planetary Gear DC motors. 1: 150 reducer gearbox produces

68 RPM at 3V.

7. Castor Wheel: 8mm steel castor ball for balancing the robot.

8. Main Chassis: 50mm diameter, 3D printed chassis which houses the motors and
the battery.

Figure 6.3: Disassembled robot chassis depicting the two fragments of robot design.

6.2 Electronics

The robot uses a single printed circuit board which houses all the electronic compo-
nents. The circuit is powered by a 3.7V 380mAh lithium ion battery which can be
charged with the on-board charging circuit. The circuit is based on the ATmega328P
microcontroller and uses nRF24L01+ for communicating with the base station and other
robots.

13
6.2.1 Architechture

The robot is controlled by an 8-bit ATmega328P microcontroller from the Atmel AVR
family. Figure 6.4 shows the architechture of the on-board electronics. The circuit
consists of the following components:

Figure 6.4: Architechture of the robot circuit board.

1. Microcontroller: The controller uses an ATmega328P running at 3.3V, 8MHz

using the internal clock.

2. Battery Management: AMS1117-3.3 linear voltage regulator is used to regulate

the battery voltage at a constant 3.3V. MCP73831T-2ACI/OT is used as a battery
charge controller for on-board battery charging.

3. Radio: A 2.4GHz nRF24L01+ module connected to the SPI bus of the micro-
controller provides communication capabilities to the robot.

4. Proximity Sensing: Six proximity sensors along the outer perimeter of the printed
circuit board can be utilized for obstacle avoidance or mapping and localisation.

5. Other: Other peripherals include an RGB light emitting diode for state indica-
tion, ambient light sensing for infrared sensor calibration, and a 12-bit Analog-
to-Digital converter for low voltage indication.

6.2.2 Fabrication and Assembly

The printed circuit board uses 0402 and 0603 packages for passives. Hot air or re-
flow soldering is recommended given the size of the components. The SN-63/PB-37

14
solderpaste was spread using on the circuit board using a custom laser cut steel stencil.
The board was reflowed in a modified OTG-Reflow Oven with the recommended solder
profile.

6.2.3 Programming

The robot can be programmed throught the ISP interface provided on the printed circuit
board. The ISP interface is in the form of contact pads and needs a special spring loaded
contact programmer for programming. An ATMEL-ICE debugger was used to program
the microcontroller. Figure 6.5 shows the custom 6 pin spring loaded contacts that were
created to program the board. Figure 6.6 depicts a board being programmed using the
ISP interface. Figure 6.7 is the ATMEL-ICE debugger that was used in programming
the robots.

Figure 6.5: The spring-loaded program- Figure 6.6: Programming the microcon-
mer. troller.

6.2.4 Cost Analysis

This section is very crucial to the development of such a system as the cost per robot di-
rectly affects the scalability of the system. Our goal was to develop a differential drive
robot within in budget of $30 but the final prototype was produced only for a mere
$21. If produced in larger quantities, the costs could be brought down even further and
scalability of the testbed imrpoved manifold. Table 6.1 depicts the cost incurred in man-
ufacturing a single robot. Majority of the costs shown are attributed to the electronics
of the robot.

15
Figure 6.7: Atmel-ICE programmer with the 6 pin spring loaded ISP programming
interface.

Parameter Price Comment

Circuit Board $10 ATmega328P, nRF24L01+, etc.
Battery $3 7.4v lithium-ion battery
Mechanical Parts $8 Motors, omniwheels, 3D printed chassis, misc
Total $21

Table 6.1: Cost Per Robot

16
 CHAPTER 7

LOCALISATION

In this chapter, we delve into the details of the fiducial tracking system used to localise
the robots. The proposed strategy states that the follower robots push heading towards
the leader robot. To implement this strategy, it is necessary that the robots have knowl-
edge of the leader’s position and their own position. Due to the simplicity of the robot
design, a decentralised localisation method, where each robot localises itself, cannot
implemented. Thus, we used a centralised fiducial tracking system using an overhead
camera to localise the robots.

The testbed used AprilTag fiducial markers to track the relative position of the robots
in the camera’s reference frame. While using such vision based markers, it is also
necessary to callibrate the camera. The next sections brief about the AprilTags and the
camera calibration process.

Figure 7.1: Screenshot of the AprilTag tracking software, detecting the actual robots
on the testbed.
7.1 AprilTag

AprilTag is a visual fiducial system, useful for a wide variety of tasks including aug-
mented reality, robotics, and camera calibration. It uses are 2D barcodes which can
be created from an ordinary printer, and the AprilTag detection software computes the
precise 3D position, orientation, and identity of the tags relative to the camera. The
AprilTag library is implemented in C with no external dependencies. It is designed to
be easily included in any applications, as well as be portable to embedded devices.

Figure 7.2: AprilTag TagID = 0 from the Tag36h11 family.

The AprilTag software produces tag ID, coordinate data and, homography in the
form of a rotation matrix. The existing AprilTag library was modified to calculate the
yaw angle, which is required to implement the proposed algorithm. Given the rotation
matrix R(α, β, γ) in the form:

R(α, β, γ) = Rz (α)Ry (β)Rx (α) (7.1)

 
r r r
 11 12 13 
= r21 r22 r23  (7.2)
 
 
r31 r32 r33

18
 then the yaw angle α can be calculated from:

α = arctan(r21 /r11 ) (7.3)

7.2 Calibration

Geometric camera calibration is used to estimate the parameters of a lens and image sen-
sor of an image or video camera. These parameters are used to correct lens distortion,
measure the size of an object in world units, or determine the location of the camera
in the scene. We performed camera calibration using a black and white checkerboard.
Figure 7.3 shows a screenshot of the calibration process.

Figure 7.3: Screenshot of checkerboard calibration.

19
 CHAPTER 8

EXPERIMENTS

In this chapter, we discuss about the physical experiments performed using the tesbed.
While the simulation results provided some initial proof of the predictions made by the
control strategy, various experiments were designed to verify the different properties
expected from the physical model.

Figure 8.1: Experiments perform on the testbed.

The experiments tested the following properties: agnosticity to shape, scalability

(Group size increase, Proportional mass to the number of robots), influence of attach-
ment points of the robots, initial orientation of the object, influence of the distribution
of mass, adaptability to a change in goal location and finally moving an active object.

8.1 Pseudogripper

While this strategy is robust to shape, it also assumes that the robots maintain contact
with the object throught the experiment. Due to the limited size of the robot, it was not
possible to incorporate a mechanical gripper or a manipulator on the chassis. Thus, it
would be difficult to maintain contact with the object at all times. Thus, we proposed
a method where the gripper would be incorporated into the object rather that the robot
itself. This concept can be termed as a pseudogripper and can effectively replace con-
ventional mechanical gripper in a scenario where the robot size or shape is a constraint.
Figure ??

Figure 8.2: Three robots placed into the pseudo gripper of an acrylic object.

21
 CHAPTER 9

CONCLUSION

We have presented a study on cooperative transport by presenting a simple physical

model describing the rigid body dynamics of such a system. This model predicts that
the proposed strategy is successful in transporting objects to the desired goal, regard-
less of the object size, shape and weight as long as a minimum number of robots are
engaged in transport. Furthermore, using this physical model, we can compare the sys-
tem dynamics to an articulated drive system, wherein, we empirically prove that any
input that causes the angle at the articulate joint to change, when ceased, will cause the
system to return to a straight angle at the articulate joint in a finite amount of time

The predictions of the model have been studied through simulations on the Open
Dynamics Engine, wherein the strategy was implemented using a group of 3 follower
e-Puck robots and 1 leader e-Puck robot. The tracking analysis has been presented in
this report. Furthermore, a system consisting of differential drive robots was developed
to physically ascertain the hypothesis. A localisation system was implemented using
AprilTag fiducial markers and the physical experiments were performed on the testbed.
Overall, this strategy seems to be a highly robust and yet simple given the miniscule
sensing and computational requirement of the robots.
 CHAPTER 10

FUTURE ENHANCEMENT

The proposed control strategy seems to perform pretty well at accomplishing the prop-
erties mentioned at the beginning of this report. The object seems to always reach its
desired goal irrespective of its shape and size. While this result is fascinating, to im-
plement such a strategy in real time would require getting rid of some of the realistic
constraints posed during the experiments. In order to accomplish a real-time behavior, a
better localisation methodology needs to be implemented which would be feasible in an
environment like warehouses. In addition, to impart complete autonomy to the system,
the algorithm should also consider distributing the robots along the object perimeter
from their stochastic initial positions. An alternative method to pseudogrippers is re-
quired to ensure contact with the object at all times. Some of the future work proposed
to improve the existing strategy is as follows:

1. An initialisation phase is proposed to be developed where stochastically scat-

tered robots can autonomously re-orient around the object by forming a closure.

2. Mechanical or pseudo-mechanical grippers are proposed to be developed for

the robot to maintain contact with the object during the entire pushing process.

3. A multi-robot SLAM (Simultaneous Localisation and Mapping) algorithm can

be developed to replace the expensive centralised fiducial localisation.
 REFERENCES
[1] Chen, J., Gauci, M., and Groß, R. (2013). “A strategy for transporting tall objects
with a swarm of miniature mobile robots.” 2013 IEEE International Conference on
Robotics and Automation, IEEE. 863–869.

[2] Chen, J., Gauci, M., Li, W., Kolling, A., and Groß, R. (2015). “Occlusion-based
cooperative transport with a swarm of miniature mobile robots.” IEEE Transactions
on Robotics, 31(2), 307–321.

[3] Groß, R. and Dorigo, M. (2004). “Cooperative transport of objects of different

shapes and sizes.” International Workshop on Ant Colony Optimization and Swarm
Intelligence, Springer. 106–117.

[4] Kube, C. R. and Bonabeau, E. (2000). “Cooperative transport by ants and robots.”
Robotics and autonomous systems, 30(1-2), 85–101.

[5] Mellinger, D., Shomin, M., Michael, N., and Kumar, V. (2013). “Cooperative
grasping and transport using multiple quadrotors.” Distributed autonomous robotic
systems, Springer, 545–558.

[6] Miyata, N., Ota, J., Arai, T., and Asama, H. (2002). “Cooperative transport by
multiple mobile robots in unknown static environments associated with real-time
task assignment.” IEEE transactions on robotics and automation, 18(5), 769–780.

[7] Rubenstein, M., Cabrera, A., Werfel, J., Habibi, G., McLurkin, J., and Nagpal, R.
(2013). “Collective transport of complex objects by simple robots: theory and ex-
periments.” Proceedings of the 2013 international conference on Autonomous agents
and multi-agent systems, International Foundation for Autonomous Agents and Mul-
tiagent Systems. 47–54.

[8] Wang, Z. and Schwager, M. (2016). “Kinematic multi-robot manipulation with

no communication using force feedback.” 2016 IEEE International Conference on
Robotics and Automation (ICRA), IEEE. 427–432.

24
PROJECT REPORT – 4
3. Publication
 Proceedings of the 2018 IEEE
International Conference on Robotics and Biomimetics
December 12-15, 2018, Kuala Lumpur, Malaysia

Perturbation Analysis of Decentralized Estimators

Kedar Karpe, Dhanalakshmi Samiappan, Kumar Ramamoorthy, and Lorenzo Sabattini

Abstract— Decentralized control strategies have been widely estimation scheme to estimate the states of robots, which
studied over the past few years. Decentralized control of multi- otherwise may not be available directly due to the restricted
leader robot networks often employs the use of decentralized communication topology. The estimation scheme presented
state estimation schemes to overcome the constraint of limited
communication. This paper presents the robustness analysis in [7] is a propitious methodology for overcoming the
of such an estimation scheme, towards robot breakdown or constraints posed by local interaction but only discusses the
communication link failure. The methodology discussed in this aspects of theoretical networks defined by intransigent graph
paper asserts that a set of independent robots estimate the Laplacian.
state of dependent robots to produce a control input for the In contrast to theoretical networks where graph connectiv-
network. To produce accurate estimates of the states of the
robots, we make an assumption that the estimator has complete ity is inflexible, physical networks, rely strongly on wireless
knowledge of the network topology. In this paper, we analyze communication for the exchange of information amongst
the dynamics of the estimation scheme in absence of this robots. The robots in such a network and the wireless links
assumption. We present a false topology to the estimator, which between them are prone to failure. Failure occurs mainly due
is symbolic of introducing robot breakdowns or link failures. to noise and external disturbances. [4], [8]–[12] discuss the
Using simulations performed by random sampling methods,
we conclude to an empirical proof about the robustness of the criticality of maintaining connectivity in multi-robot systems
estimator. in order to minimize faults. [13] proposes a concept of global
connectivity which, to some extent, overcomes the difficulty
I. INTRODUCTION of limitations of communication range, however, it does not
Cooperative robotics is an emerging branch of autonomous address non-trivial difficulty posed by robot failure or link
robotics with the prospect of simplifying complex tasks failure. Thus, physical factors such as limited range and
using the idea of decentralized behaviors. Applications like interference exhibit the necessity for fault tolerant control
distributed sensing [1], disaster recovery [2], or warehouse schemes robust to a certain amount of perturbations in their
management [3] demand the necessity of cooperative systems topology.
where a single robot is incapable or inefficient for performing This paper provides a novel analysis of the decentralized
a large distributed task. The advantages of such systems estimation scheme presented in [7] such that it guarantees
are vast. To name a few, such systems provide excellent complete state estimation even in the presence of robot
adaptability to tasks, resilience to failures, and scalability failures and network link failures. We examine the dynamics
of labor. of the estimation scheme by adding or deleting random
To achieve such complex distributed tasks defined only links to signify robot or link failure. We further examine
using local interaction rules, it is critical to maintain con- the constraints in implementing the estimation scheme and
nectivity amongst the robots to ensure an uninterrupted flow provide empirical proof of its robust characteristics.
of information for the distributed consensus [4]. This is The motivation for such a fault tolerant scheme is pre-
essential because we need complete information about the sented in [14] where a consensus between the robots in the
states of each robot in the network to model the dynamics of network has to be achieved. An application of this scheme
such a multi-robot system. We’re interested in the approach is achieving rendezvous, that is attaining a location-based
presented particularly in [5]–[7], which describes a control consensus between the robots of a network. [15] describes
strategy that exploits locally available information to achieve the difficulties in achieving consensus with time variant
a cooperative goal for the network. topologies. Most of the solutions to this problem involve
The network definition in [7] is made using the well known maintaining the connectivity of the network [16] and [17],
consensus protocol. The network is divided into two sets, by imposing constraints on the consensus locations. Thus,
dependent robots, which evolve by local interaction, and robust estimation schemes can prove a better solution to this
independent robots, which inject the control input in the problem.
network. The interaction rule makes use of a decentralized
II. DYNAMICS FORMULATION
K. Karpe, D. Samiappan and K. Ramamoorthy are with Department of Consider a set V = {1, 2, 3..., N } of robots in a network
Electronics and Communication Engineering, SRM Institute of Science and
Technology, Kattankulathur, India system, whose interconnection topology is defined by an
kedarprasad [email protected] undirected graph G, which can either be weighted or non-
{dhanalakshmi.s, kumar.r}@ktr.srmuniv.ac.in weighted. With N being the total number of robots, we
L. Sabattini is with the Department of Sciences and Methods for
Engineering (DISMI), University of Modena and Reggio Emilia, Italy divide the network into two sets, a set of dependent robots
[email protected] VD ⊂ V and the other set of independent robots VI ⊂ V .

978-1-7281-0377-8/18/$31.00 © 2018 IEEE 2373

The notation ND is used to denote the number of dependent robots have complete access to output y of the dependent
robots whereas NI = N − ND is used to denote the robots in order to produce the state estimates. Thus, we can
number of independent robots. The independent robots are write the output equation of dependent robots such that it
responsible for injecting control input into the network to ensures observability of the robot states:
achieve the cooperative goal. The dependent robots, on the
other hand, evolve their states based on local interaction. Let y = B  xD (7)
the state of each robot follow a single-integrator kinematic
model such that: III. DECENTRALIZED STATE OBSERVER
In this section, we discuss the need for an estimator in
ẋi = μi (1) our model of a networked robot system. We introduce the
where xi ∈ Rm is the state of the i-th robot in the network concept of decentralized estimation in multi-robot systems
and μi ∈ Rm is each robot’s control input. We assume the as in [7], to fulfill this need and manifest some constraints
robots are connected according to the well know consensus to make this estimation scheme stable.
protocol [14], thus their states converge to a common desire As defined in the previous sections, the i-th independent
value u. The consensus protocol articulates that the states of robot injects the control input component ui , i = 1, ..., NI
the robots evolve by nearest neighbor averaging. Thus, we at different points of the network to achieve the cooperative
can assume that the states of the dependent robots evolve goal. [19] discusses a control strategy which makes the use
according to the following interaction rule: of an exogenous signal to achieve such a cooperative goal.
We similarly define our input vector u following this control

ẋi = − (xi − xj ) (2) strategy as:
j∈Ni

Assume that the control input is injected at the independent u = F xD + ζ (8)

robots as: where, F ∈ RNI ×ND is the stabilizing term calculated
by pole placement, such that (A + BF ) is Hurwitz stable.
xi = ui (3) The vector ζ is an exogenous signal that defines by the
where ui is the ith component of control input u . [18] cooperative goal to be achieved.
defines a method to index the interconnection matrix in such In order to exploit the control strategy defined by equa-
a way that the last NI robots are independent robots and tion (8), it is critical for the independent robots to obtain
the first N − NI robots are the dependent robots. Thus, the the inputs calculated by other independent robots. By the
graph Laplacian associated with the network graph G can be assumption made in [7], the independent robots do not
decomposed as: communicate with each other. Since the independent robots
  do not have complete knowledge of the states xD due
A B to the limited communication, we exploit locally available
L(G) = − (4)
B C information to estimate the input components ui calculated
where A ∈ RND ×ND represents the interconnection by other independent robots.
among dependent robots and B ∈ RND ×NI represents the Assuming all the independent robots have complete
interconnection among dependent and independent robots. knowledge of the network topology, we define the control
We now introduce xD = [x1 , ..., xND ] ∈ RND as the state input ui which is an estimate of the actual input ui . This
vector of the dependent robots and xI = [xD+1 , ..., xN ] ∈ input is defined using the modified version of equation (8):
RNI as the state vector of the independent robots. From
equations (2) and (3), we define the dynamics of the network ûi = F x̂i + ζ (9)
using state space model as:
Each independent robot i, produces a state estimate xi as
   
A B ONI ,ND follows:
ẋ = x+ ν (5)
ONI ,ND ONI ,NI INI
x̂˙ i = Ax̂i + B ûi − Ki (y[i] − bi  x̂i ) (10)
where ν ∈ RNI is the input vector and x = {x1 , ..., xN } ∈
N
R is the state of the multi-robot system. From equation (5), where Ki ∈ RND ×mi is a matrix containing the weights
we can model the dynamics of independent robots and pertaining to the i-th robot of the filter matrix K and bi is
dependent robots respectively as follows: the i-th column of the interconnection matrix B. [7] defines
 the method to calculate the filter matrix K such that the
ẋI = ν estimation error has a stable dynamical response that is, it
(6)
ẋD = AxD + Bu converges to zero as time t → ∞. To find the filter matrix K,
As we are aware from definitions given so far, the relation we define a constraint on pole placement such that, λi,min ,
between the two robots is defined by the interconnection which is the minimum eigenvalue of the equation A+ki b i +
matrix B  , in addition, we assume that the independent BF , ∀i = 1, . . ., NI is given by:

2374
 and the input ûiρ is given by:
λi,min  −(INI ⊗ B)F̃E  (11)
ûiρ = F x̂iρ + ζ (13)
The operator ⊗ indicates Kronecker product of the
operand matrices. If (A + BF ) were Hurwitz stable, we For the error analysis, we collect estimates from all the
perform pole placement using this condition to obtain the independent robots as follows:
filter term Ki  for all independent robots, which gives us
the complete filter matrix. By this method, we can explicitly x̂Eρ = col(x̂iρ , i = 1, . . ., NI ) (14)
define the filter term Ki  such that the estimator has a stable
response, which is, it converges to zero as time t → ∞. The We now define the new estimator error ρ ∈ RNI ND as:
matrix F̃E in equation (11) is a block matrix where each
(i, j)-th block is a null matrix except the j-th row of matrix ρ = x̂Eρ − xE (15)
F . The structure of both the matrices F and F̃E is shown in where, xE is a vector collecting the state vector xD , NI
the following example: times.
  To simulate the perturbed estimation scheme, we define
F11 F12 F13
F = the exogenous input signal by means of periodic setpoints
F21 F22 F23
for each independent robot. The estimation error can be
⎡ ⎤ derived by comparing the actual states generated by these
F11 F12 F13 0 0 0 setpoint inputs with the estimated states produced by the
⎢ 0 0 0 F21 F22 F23 ⎥
F̃E = ⎢
⎣ F11
⎥ novel estimation scheme. The setpoint generation can be
F12 F13 0 0 0 ⎦ achieved by exploiting the states of an exosystem as a linear
0 0 0 F21 F22 F23 combination of n ∈ N harmonics as in [16]. We can define
IV. PERTURBATION TOPOLOGY the exosystem by the equation:
In this section, we introduce the main result of the paper. ˙ = Gξ(t)
ξ(t) (16)
In particular, the estimation scheme presented so far was
developed in [7] considering an ideal situation, in which fail- where ξ ∈ Rn̄ is the state vector with n̄ = 2n + 1. The
ures of robots and communication links are not considered. initial condition ξ(0) is assigned as ξ(0) = [1, 0, 1, ..., 1]
In a real robotic system, a link between the robots signifies and G ∈ Rn̄ × n̄ is an opportunely defined, block diagonal
a channel for information exchange. These communication matrix:
links facilitate the exchange of state information amongst ⎡ ⎤
the robots in the network. However, in a real-time system, G̃0 0 . . . 0
⎢ 0 G̃1 . . . 0 ⎥
these links are prone to failure, interrupting the information ⎢ ⎥
G=⎢ . .. .. .. ⎥
exchange. [20] and [21] analyze the cause and effect of ⎣ .. . . . ⎦
different kinds of such link failures in robotics. As such,
0 0 . . . G̃n
it is very important to analyze the performance of the
estimation scheme in such a realistic scenario, that is in the Each block G̃p is given by:
presence of failure. In particular, we want to analyze the ⎧
robustness of the estimation scheme to perturbations. For ⎨ 0,
⎪  if p = 0
this purpose, we will hereafter consider perturbations in the G̃p = 0 p 2π
topology, which directly affect the dynamics of the system. ⎪
⎩
T , if p = 1, . . ., n.
−p 2π
T 0
Specifically, we consider a perturbation as the interruption
of the communication among two robots. An edge matrix is The solution of this exosystem is given by:
used to define the interconnections between the robots where
the edge signifies the communication link. To simulate link 2π 2π 2π 2π
failure, we interpret perturbation as addition or removal of ξ(t) = [1 sin( t) cos( t) . . . sin(n t) cos(n t)]
T T T T
these edges to or from the network topology. (17)
Let us introduce the perturbed form of interconnection [22] describes how this exosystem can solve the setpoint
matrix A as Aρ and B as Bρ . Hereafter, we make the tracking problem. Thus, we can define the input u as follows:
following assumption:
u = F xD + (Γ − F Π)ξ (18)
Assumption 1: Post the introduction of the perturbation, the
¯ ¯
new topology defined by interconnection matrices (Aρ , Bρ ), where Γ ∈ RNI ×n̄ and Π ∈ RND ×n̄ are obtained
is also controllable. as solution of the regulator equations. From the setpoint
dynamics defined in [13] we can write these equations as:
Hence, according to the assumption, the dynamics of the

estimator, are modified as: AΠ + BΓ = ΠG
(19)
x̂˙ iρ = Aρ x̂iρ + Bρ ûiρ − Ki (y[i] − b
iρ x̂iρ ) (12) Π − J = OND ,n̄

2375
Fig. 1: Example topologies and their estimation errors in the presence of random network perturbations. The robots marked
with the prefix D represent the dependent robots whereas those marked with the prefix I represent the independent robots.
The estimation error as defined in equation (15) is presented in the graph.

Now, to empirically ascertain our hypothesis of the estima- topology. Table I outlines the attributes that were used to
tion scheme being robust, we use Monte Carlo method [23] perform the simulations. For each sample, we run the estima-
and [24]. We randomize the perturbations to a set of pre- tion scheme to generate the estimation error response of the
defined topologies and find the estimation error in presence perturbed topology. If we observe the complete convergence
of these perturbations. For this, we choose more than ten of dynamic response for a large number of random pertur-
topologies in varying order of connectedness. The Monte bations then we can conclude that the estimation scheme is
Carlo method is used to produce random samples, which robust one [25].
we exploit to represent the edges in the network. We then Figure 1 exhibits the estimation error for two different
use these samples to remove the corresponding edges from topologies which were perturbed with consistency to as-
the originally defined edge matrix and generate a perturbed sumption made in the previous sections. The robots 1 and

2376
 TABLE I: Simulation Attributes
Number of nodes 3→12
Number of controllable network graphs (weighted and 18
unweighted)
Number of random edge additions 25
Number of random edge removals 175
Number of unstable responses 0

8 in Topology 1 and the robots 10, 11, 12 in Topology 2

represent independent robots that provide the control inputs
to the network. We can observe that estimation error in
both cases achieves complete convergence in finite time. We
then perform the analysis of eighteen different topologies
with varying complexity of about two hundred different
perturbation samples. Figure 2 shows estimation error due to
the addition of edges from the original network topology. In
Figure 3 we can notice that the estimation scheme is stable
for all two hundred simulations. [25] analyses how results
of Monte Carlo simulations can be deemed as inferences.
Thus, we can perform simulations with a large number of
random inputs and then generalize our result to all kinds of
networks. With 200 general cases of random perturbations,
we can make a strong conclusion about the behavior of our
Fig. 2: Monte-Carlo simulation of addition of random edges simulations. From Figure 3, which represents the stability
from the network topology. The colored lines represent esti- of the perturbed estimator defined in equation (10), we can
mated states of each dependent robot by every independent infer the robustness of the estimation scheme.
robot in the network.
V. CONCLUSION
In this paper, we have empirically proven how the esti-
mation scheme for networked multi-robot systems is robust.
We have primarily defined a scenario in which perturbations
occur and what they signify. The interpretation for perturba-
tion has been defined as robot failure and communication
link failure in networked multi-robot systems. We have
then simulated a deducible number topologies subjected to
both link addition and link removal using Monte Carlo
method. Through the practical evidence of the stability of
the estimation error, we can thus infer that the estimation
scheme is a robust one.

R EFERENCES
[1] A. Howard, M. J. Matarić, and G. S. Sukhatme, “Mobile sensor
network deployment using potential fields: A distributed, scalable
solution to the area coverage problem,” in Distributed Autonomous
Robotic Systems 5, pp. 299–308, Springer, 2002.
[2] N. Michael, S. Shen, K. Mohta, V. Kumar, K. Nagatani, Y. Okada,
S. Kiribayashi, K. Otake, K. Yoshida, K. Ohno, et al., “Collaborative
mapping of an earthquake damaged building via ground and aerial
robots,” in Field and Service Robotics, pp. 33–47, Springer, 2014.
[3] P. R. Wurman, R. D’Andrea, and M. Mountz, “Coordinating hundreds
of cooperative, autonomous vehicles in warehouses,” AI magazine,
vol. 29, no. 1, p. 9, 2008.
[4] N. Michael, M. M. Zavlanos, V. Kumar, and G. J. Pappas, “Maintain-
ing connectivity in mobile robot networks,” in Experimental Robotics,
pp. 117–126, Springer, 2009.
[5] F. Boem, L. Sabattini, and C. Secchi, “Decentralized state estimation
for heterogeneous multi-agent systems,” in Decision and Control
Fig. 3: Monte-Carlo simulation of 200 random samples of (CDC), 2015 IEEE 54th Annual Conference on, pp. 4121–4126, IEEE,
edge addition and removal. 2015.
[6] F. Boem, L. Sabattini, and C. Secchi, “Decentralized fault diagnosis
for heterogeneous multi-agent systems,” in Control and Fault-Tolerant
Systems (SysTol), 2016 3rd Conference on, pp. 771–776, IEEE, 2016.

2377
 [7] F. Boem, L. Sabattini, and C. Secchi, “Decentralized state estimation in single-leader, consensus networks?,” IEEE Control Systems, vol. 32,
for the control of network systems,” Journal of the Franklin Institute, no. 4, pp. 66–73, 2012.
2018. 10.1016/j.jfranklin.2017.11.028. [19] L. Sabattini, C. Secchi, M. Cocetti, A. Levratti, and C. Fantuzzi, “Im-
[8] J. Cortés, S. Martı́nez, and F. Bullo, “Robust rendezvous for mobile plementation of coordinated complex dynamic behaviors in multirobot
autonomous agents via proximity graphs in arbitrary dimensions,” systems,” IEEE Transactions on Robotics, vol. 31, no. 4, pp. 1018–
IEEE Transactions on Automatic Control, vol. 51, no. 8, pp. 1289– 1032, 2015.
1298, 2006. [20] R. Parasuraman, “Few common failure cases in mobile robots,” arXiv
[9] J. Panerati, M. Minelli, C. Ghedini, L. Meyer, M. Kaufmann, L. Sabat- preprint arXiv:1508.03000, 2015.
tini, and G. Beltrame, “Robust connectivity maintenance for fallible [21] J. Carlson and R. R. Murphy, “Reliability analysis of mobile robots,”
robots,” Autonomous Robots, 2018. in Robotics and Automation, 2003. Proceedings. ICRA’03. IEEE
[10] C. Ghedini, C. H. C. Ribeiro, and L. Sabattini, “Toward efficient International Conference on, vol. 1, pp. 274–281, IEEE, 2003.
adaptive ad-hoc multi-robot network topologies,” Ad Hoc Networks, [22] A. Saberi, A. A. Stoorvogel, and P. Sannuti, Control of linear systems
2018. with regulation and input constraints. Springer Science & Business
[11] C. Ghedini, C. Ribeiro, and L. Sabattini, “Toward fault-tolerant multi- Media, 2012.
robot networks,” Networks, vol. 70, no. 4, pp. 388–400, 2017. [23] S. Raychaudhuri, “Introduction to monte carlo simulation,” in Simu-
[12] M. Minelli, M. Kaufmann, J. Panerati, C. Ghedini, G. Beltrame, lation Conference, 2008. WSC 2008. Winter, pp. 91–100, IEEE, 2008.
and L. Sabattini, “Stop, think, and roll: Online gain optimization for [24] C. B. Borkowf, “Random number generation and monte carlo meth-
resilient multi-robot topologies,” in Proceedings of the International ods,” Technometrics, vol. 42, no. 4, p. 431, 2000.
Symposium on Distributed Autonomous Robotic Systems (DARS), [25] C. Beisbart and J. D. Norton, “Why monte carlo simulations are infer-
(Boulder, CO, USA), 2018. ences and not experiments,” International Studies in the Philosophy
[13] L. Sabattini, C. Secchi, N. Chopra, and A. Gasparri, “Distributed of Science, vol. 26, no. 4, pp. 403–422, 2012.
control of multirobot systems with global connectivity maintenance,”
IEEE Transactions on Robotics, vol. 29, no. 5, pp. 1326–1332, 2013.
[14] R. Olfati-Saber, J. A. Fax, and R. M. Murray, “Consensus and
cooperation in networked multi-agent systems,” Proceedings of the
IEEE, vol. 95, no. 1, pp. 215–233, 2007.
[15] R. Olfati-Saber and R. M. Murray, “Consensus problems in networks
of agents with switching topology and time-delays,” IEEE Transac-
tions on automatic control, vol. 49, no. 9, pp. 1520–1533, 2004.
[16] M. M. Zavlanos, M. B. Egerstedt, and G. J. Pappas, “Graph-theoretic
connectivity control of mobile robot networks,” Proceedings of the
IEEE, vol. 99, no. 9, pp. 1525–1540, 2011.
[17] M. A. Hsieh, A. Cowley, V. Kumar, and C. J. Taylor, “Maintaining
network connectivity and performance in robot teams,” Journal of
Field Robotics, vol. 25, no. 1-2, pp. 111–131, 2008.
[18] M. Egerstedt, S. Martini, M. Cao, K. Camlibel, and A. Bicchi,
“Interacting with networks: How does structure relate to controllability

2378
2019 19th International Conference on Advanced Robotics (ICAR)
Belo Horizonte, Brasil, December 02-06, 2019

SPRINTER: A Discrete Locomotion Robot for Precision

Swarm Printing
Kedar Karpe, Ayon Chatterjee, Pranav Srinivas,
Dhanalakshmi Samiappan, Kumar Ramamoorthy and Lorenzo Sabattini

Abstract— This paper presents SPRINTER, a system and amongst a group of mobile robots. Also, we try to overcome
method for multi-robot printing. In this paper, we discuss the the constraint of the maximum printing area in conventional
design of a quasi-holonomic mobile robot and present a method printers by exploiting the mobility of such robot printers.
which uses a group of such robots to distributively print a
large graphical image. In the distributive printing method, we The motivating work for this research is [14], where
introduce the concept of image cellularization for segmenting the authors develop a mobile printing system which can
the graphic into a group of smaller printing tasks. We then draw graphics on a given surface. The authors present two
discuss a centralized method to allocate these tasks to each methods, one centralized and one decentralized algorithm,
robot and execute the printing process. In summary, we present to dynamically distribute tasks among a group of robots.
a multi-robot printing system which enhances the printing
speed and maximizes the printing area of traditional industrial While the presented work demonstrates positive results to-
printers. wards multi-robot printing, the assumption that the print data
constitutes of only straight lines creates a major roadblock
I. I NTRODUCTION in scaling such a system to standard RGB graphics. The
Multi-robot systems present a fascinating bio-inspired paper fails to discuss task allocation for complex RGB
methodology of controlling a large group of simple robots graphics. Besides, given the ability of mobile robots to print
to perform complex coordinated tasks [1]–[3]. Such systems, immensely large data, the paper doesn’t discuss methods to
consisting of multiple robots coordinating together, have correct positional errors introduces by the dead-reckoning
been considered a promising research problem, mainly due system [15].
to their dynamicity to faults and failures [4]–[7]. Though In this paper, we present a discrete locomotion printing
there has been a propitious development in this field, most of robot, the SPRINTER, and a centralized multi-robot printing
the outcome has been theoretical or research-oriented. Multi- system which closely resembles a parallelized inkjet printer.
robot platforms like the Kilobots [8], the GRITSbot [9] and The custom omni-wheeled robot is developed to achieve a
the E-Puck robots [10] were designed mainly for studying high degree of miniaturization so that the robots can be
collective behaviors and thus possess low modularity for easily scaled in number. Unlike other mobile robots [8]–
practical applications. Very little work has been done which [10] whose primary objective is to study collective behaviors,
prompts the usage of multi-robot systems commercially. the SPRINTER is primarily developed and optimized for
Only recent developments in warehouse logistics [11]–[13] the specific task of printing and, at the same time, is also
have been able to achieve a certain degree of commercial- modular, making it proficient at performing other tasks.
ization. These developments have showcased that indeed, The system consists of a base station which performs the
multi-robot systems can be effectively used to solve practical centralized image segmentation and the task allocation steps,
problems and have prompted the growth of commercial the SPRINTER robots which perform the inkjet printing,
multi-robot systems. and an optical tracking system which provides global pose.
In our work, we specifically improvise the traditional The architecture of the proposed system overcomes the
graphic printing technology and introduce cooperative print- major drawbacks in [14] which we will address one at a
ing methods to reduce printing times of industrial printers. time. The major improvement that the SPRINTER system
The prime motivation for this work is the dependency of introduces is the method of addressing and distributing tasks.
printing time on the size of the graphic in conventional The proposed architecture divides an image into smaller
printing systems. The work presented in this paper aims at cells whilst maintaining an equal cost function of each cell.
reducing this dependency by distributing the printing task The robots, which receive numerical commands from the
base station [16], addresses an image with its constituent
K. Karpe, A. Chatterjee, P. Srinivas, D. Samiappan and K. pixels, hence, the algorithm can be easily scaled to various
Ramamoorthy are with Department of Electronics and Communication applications, of which additive printing has a very promising
Engineering, SRM Institute of Science and Technology, Kattankulathur,
India {kedarprasad pra, ayon santoshkumar, future. The cellularization of the image and the centralized
pranavsrinivas murali}@srmuniv.edu.in architecture ensure collision-free paths, thus eliminating the
{dhanalas, kumarr}@srmist.edu.in need for a dynamic collision avoidance mechanism. The
L. Sabattini is with the Department of Sciences and Methods for
Engineering (DISMI), University of Modena and Reggio Emilia, Italy SPRINTER architecture also employs an optical tracking
[email protected] system to correct pose errors caused by dead-reckoning.

978-1-7281-2467-4/19/$31.00 ©2019 IEEE 733

 While printing is an interesting problem per se, the results
described in this paper can be easily extended to further ap-
plications. In particular, due to its similarity to printing sys-
tems, the additive manufacturing (which is seen to be one of
the prospective applications of collective fabrication) seems
a promising domain. At present, the control of collective
additive manufacturing robots is a conceptual idea, but initial
studies have shown positive evidence for employing multi-
robot systems in additive manufacturing. For instance, [17]
has combined methodologies in aerial robotics and additive
manufacturing to develop a 3D printing quadcopter. A proof
of concept is presented in [18], for using swarms of robots
in complex additive manufacturing. It is noticeable that Fig. 1: The architectural overview and the process flow of
most of these advances towards collective manufacturing the SPRINTER system.
have been curtailed by the lack of knowledge on material
tunability. At present, it is not very clear how the control
of the physical properties of materials could be achieved δ(N ) which clusters pixels based on euclidean distance from
in a collective fabrication process. Preliminary results on centroids ci :
material tunability specifically for collective manufacturing
are presented in [19]. Thus, advances in material fusion and δ(N ) = {pm ∈ Ci | dmin (pm , ci ) ≤ dmin (pm , cj ),
tunability proffers guaranteed anticipation for a system like
n(Ci ) ∼ n(Cj ), ∀i 6= j} (1)
the SPRINTER.
The rest of the paper is organized as follows. The overall Here dmin (pm , ci ) is the shortest distance of pixel pm from
architecture of the SPRINTER system is presented in Sec- the centroid ci and n(Ci ) is the number of pixels in cell Ci .
tion II. The design of the SPRINTER robots is detailed in In the next step we update the centroids ci :
Section III. Subsequently, Section IV presents the motion
control algorithm implemented on the SPRINTER robots. N
1 X
Simulation results are then summarized in Section V. Finally, ci = pi (2)
concluding remarks are reported in Section VI. | Ci | i=1

II. S YSTEM OVERVIEW If we iterate over the above steps sufficent number of
A. Architectural Overview times, then the segmented image will contain cells with
almost equal number of black pixels. Additionally, this
Fig. 1 depicts the architectural overview of the SPRINTER algorithm will also guarantee that each robot takes almost
system. In the following sections, we introduce the con- equal time to print each cell, assuming only the black cells
cept of image cellularization, that is dividing an image contribute to the printing time. Fig.2 shows a sample image
into smaller cells of equalized cost function. The proposed cellularized using the proposed algorithm.
system performs distributed printing based on this principle
of segmenting an image. Each cell is then assigned to an
individual robot using a distributed path planning and task
allocation algorithm and hence a distributed printing process
is performed by the robots.
B. Image Cellularization
Let us assume that we have N SPRINTER robots available
for an experiment, we need to divide the print graphic in
N cells. For simplicity of the prototype development, the
SPRINTER robot was equipped with only a black inkjet Fig. 2: A sample monochrome image segmented into 5 cells.
cartridge. Hence, we will consider a 1-bit monochromatic The dashed lines represent cell boundaries and the circular
image that needs to be cellularized. This methodology can markers represent centroids of the cells.
be easily extended to more complex RGB images by simply
considering the average of the color components.
We use a modified version of k-means clustering to
C. Task Allocation
define the image cells. First, define N centroid seeds
{ci , i = 1, 2, ..., N } and their corresponding cells {Ci , i = Task allocation is critical to ensure completion of the over-
1, 2, ..., N } using k-means++ algorithm. We consider black all objective while guaranteeing collision free trajectories for
pixels {pm , m = 1, 2, ..., M } where M is the total number the homogeneous robots. We have a problem of allocating
of black pixels in the image. Define an objective function N robots to N goal locations in a 2-dimensional Euclidean

734
space1 , where each location can be occupied only by a single
robot. Here, we consider the goals location to be the center
of the cell ci , as defined in the previous subsection. The
initial positions, however, can be random locations on the
print area.
Finding collision free trajectories is crucial because inter-
robot collisions can cause catastrophic printhead misalign-
ment resulting in a malformed print. To solve this assignment
problem, we use a centralized algorithm [20] based on
minimal velocity trajectory. The algorithm presented in [20]
ensures collision free trajectories in a finite time under
mild assumptions about clearance requirements between the
robots. For this method we define an N × N assignment Fig. 3: The basic hardware architecture of SPRINTER Robot.
matrix φ such that: It uses dual microcontroller architecture for handling com-
munication and motor control separately.
φi,j ∈ {0, 1}, ∀ i, j ∈ N (3)
Define now the position of the i − th robot as xi ∈ R2 , and [20] then shows that the trajectories γ(t) can be obtained
the position of the j − th goal as cj ∈ R2 . Stacked vectors from (9) such that:
of positions X(t), C ∈ R2N can then be defined as follows:  
t − t0
| |
X(t) = [x1 (t) , x2 (t) , ..., xN (t) ]|
|
(4) γ(t) = 1 − X(t0 )
t − tf
 
| t − t0
C = [c1 | , c2 | , ..., cN | ] (5) + (ΦC + (IN n − ΦΦ| )X(t0 )) (10)
t − tf
as well as an extended assignment matrix Φ = φ⊗In , where
Define now R > 0 as the radius of the circle circumscribing
In is an identity matrix in n dimensions. We try to find N
the outline of the robot. It is further proved in [20] that, if
finite time trajectories defined by:
we ensure√the distance between any two initial locations is
γ(t) : [t0 , tf ] → X(t) (6) at least 2 2R and, similarly, the√distance between any two
goal locations is also at least 2 2R, then, the trajectories
where initial and final times t0 and tf are explicitly defined generated from (9) are collision free. The task allocation
considering the speed limitations of the robot. The boundary is performed in the initialization phase following which the
conditions for this problem are: robots traverse to their assigned cells. The optical tracking
system provides the initial positions of all the robots and the
γ(t0 ) = X(t0 )
(7) image segmentation provides the goal locations as the center
Φγ(tf ) = C of each cell.
φ| φ = IN (8) D. SPRINTER Robot
These boundary conditions ensure the feasibility of the The SPRINTER robot is a custom mobile robot designed
obtained trajectories: namely, the trajectories start, at time specifically for multi-robot printing application. It mainly
t0 , in the current position of the robots, and end, at time features an open-loop precision locomotion system and an
tf , in the position of the goals. Eq. 8 ensures each robot is in-built inkjet cartridge which we exploit to obtain graphical
mapped only to a single goal. prints. We discuss the details about the SPRINTER robot in
In this solution to the assignment problem, we do not Section III.
consider clearance requirements for the trajectories. Rather,
E. Fiducial Localization
we set limitations on the initial and goal locations such that
the trajectories are always guaranteed to be collision free. The SPRINTER system uses an optical tracking system for
Under the assumption that clearance requirements do not global localization, using the well known AprilTag fiducial
exist, the solution to the assignment problem shown in [20] markers [21]. A USB webcam connected to the server tracks
simplifies to: the markers and obtains the pose and homography of each
Z tf robot. The optical tracking system is used mainly to obtain
minimize Ẋ(t)| Ẋ(t)dt the initial pose of the robots for task allocation and nozzle
φ t0 (9) alignment before printing.
subject to (3), (7), (8)
III. THE SPRINTER ROBOT
1 Since,in this paper, we are considering a printing problem, we focus A. Hardware Architecture
on a 2-dimensional environment. However, everything can be trivially
extended to higher-dimensional environments, for instance to consider In this section, we discuss the hardware design of the
additive manufacturing cases. SPRINTER robot. We also discuss certain the electronic

735
design aspects such as the inkjet driver, which manages
the ink firing mechanism, and the power management unit,
which supplies power to different blocks of the system. Fig. 6
shows the robot which was used in the work presented.
1) Microcontroller: The robot uses a dual microcontroller
architecture as in [9]. The primary microcontroller is an
ESP8266 operating at 3.3V/80MHz while the auxiliary mi-
crocontroller is an ATmega328P operating at 5V/16MHz.
SPRINTER uses the ESP8266 to exchange local messages
with other robots, and global messages with the central
server.
We use a modified version of G-code, a low level nu-
merically controlled programming language, to convert the
graphical data into respective robot commands. The base
station sends these G-code packets to each robot over the
local network. The primary microcontroller receives and
processes these messages, while the auxiliary microcontroller
controls the robot.
The ATmega328P microcontroller primarily manages mo-
tor control and inkjet control. A UART message arrives in
the form of the robot’s target pose, and a 4-bit binary-coded-
decimal value which determines the sequence of nozzles to
be fired at the target pose. The stepper motor control is
achieved by two dual H-bridge stepper motor drivers.
2) Inkjet Driver/Printer: The robot uses a standard
generic inkjet cartridge for printing. A multiplexed high cur-
rent driver sprays ink from the cartridge nozzles. A set of 12
individually addressable nozzles are fired by the multiplexer
to spray ink in a specific pattern. Image resolutions of up to
96dpi can be achieved from this setup.
3) Power Requirements: The control circuit operates at
four distinct voltages, hence power management is quite
a complex aspect of the hardware design. The ESP-12
microcontroller operates at 3.3V, whereas the ATmega328P
operates at 5V. The firing voltage of the inkjet driver is rated
Fig. 4: The exploded view of the SPRINTER robot which
at 20-22V. Considering the size of the robot as our main
represents all the separate components of the mechanical
constraint, we selected a 11.1V 3s LiPo battery as the power
design.
source. The battery voltage is sampled at 0.1Hz and a low
voltage is notified to the base station.
B. Mechanical Design • Top layer: This layer is a 3D printed lid which shelters
The structure of the SPRINTER robot, shown in Fig. 4, the electronics and also accommodates the AprilTag.
can be spit into three separate layers.
IV. ROBOT DYNAMICS
• Bottom layer: This layer is the main robot chassis which
is 3D printed using PLA. The chassis, which measures The SPRINTER robot features a quasi-holonomic drive.
100mm x 100mm x 120mm, is square-shaped due to By this, we mean that the robot has a holonomic omni-wheel
the ease of alignment of the cartridge nozzle. It houses drive but, the firmware of the robot allows linear motion
the mechanical drive which includes the four stepper only in the (x, y) directions. After the task allocation step,
motors and the four omniwheels. The chassis also holds the rotation of the robot along θ is restricted to avoid nozzle
the Li-Po battery and the inkjet cartridge. The 48mm misalignment. We use a modified version of Bresenham’s
omniwheels lift the cartridge 2.5mm above the print Algorithm [22] to define the positional dynamics of the
surface. robot. Bresenham’s Algorithm defines a method to obtain
• Intermediate layer: This layer is the controller circuit nearly continuous trajectories using discrete motion. The
board which houses all the electronics of the robot. modified version of this algorithm assumes the start and end
Flex cables interface the board with electronics in the coordinates to be integral multiples of the step size of the
bottom layer. The circuit board has overall dimensions motor. Consider an initial point (x1 , y1 ) and destination point
of 80.2mm x 80.2mm x 20.5mm. (x2 , y2 ) in the coordinate space represented as a grid having

736
cell dimensions equal to that of the step size and the variables a single SPRINTER robot) and the SPRINTER system
∆x = x2 − x1 , and ∆y = y2 − y1 . We define the driving consisting N = 6 SPRINTER robots. The plots represent the
axis DA as the axis being tracked and passive axis P A as timing characteristics of the system. The line plot represents
the axis which evolves automatically. x-axis is chosen as the N = 1 robot whereas the box plot represents a system with
driving axis if |∆x| ≥ |∆y|, and y-axis if |∆x| < |∆y|. N = 6 robots. There is a notable acceleration of printing
Next, we define error  as the negative distance, from any speed using the proposed system. Also, given the near-
point on the line joining initial and goal locations to the infinite printing dimension, such a system can effectively
top edge of the grid cell at that point and ¯ = ∆P A. The replace conventional industrial printers.
Bresenham’s Algorithm keeps a track of  and increments
along the passive axis as  becomes greater than zero. This VI. C ONCLUSION
way, we’re able to achieve a discrete positional control of In this paper, we have discussed a system and a method
the SPRINTER robot. for multi-robot printing. The system uses the concept of
image cellularization with a fixed cost function. Each cell
Algorithm 1 Modified Bresenham’s Algorithm for integral is then assigned to the robots using a centralized task
coordinates allocation algorithm which also ensures collision avoidance.
1: Given (x1 , y1 ) and (x2 , y2 ) The system also uses an optical tracking mechanism for pose
2: ∆x = x2 − x1 and ∆y = y2 − y1 error correction and nozzle alignment. The SPRINTER robot
3: j = y1 is a quasi-holonomic robot which uses dead-reckoning for
4: ¯ = ∆x − ∆y precise localization. It employs a generic inkjet cartridge for
5: for i = x1 to x2 − 1 do printing. An essential feature in multi-robot systems would
6: Go to (i, j) be autonomous charging. But, as of now, the SPRINTER
7: if ¯ ≥ 0 then does not feature onboard charging of the batteries. The local
8: j+ = 1 communication feature is also mostly unused but can be used
9: ¯− = ∆x in the future to introduce decentralization, wherein the robots
10: end if can distribute tasks using dynamic coverage control. The
11: ¯+ = ∆y simulations section further presents an empirical comparison
12: end for of printing time improvement offered by SPRINTER.
As expected, the SPRINTER system outperforms tra-
ditional printing systems, in terms of printing time and
V. S IMULATIONS
maximum print size. Future work will aim at realizing an
So far we have discussed the design of the SPRINTER extensive evaluation campaign, considering more complex
robot and the methodology of the system. In this section, we images, possibly with colors, and larger groups of robots.
present the simulation results of the proposed algorithm.
We carried out simulations to compare the performance ACKNOWLEDGEMENT
of the proposed multi-robot architecture to that of a conven- This research was funded by the SRM Research Institute,
tional single printhead printers. As discussed in the previous SRMIST. We are also obliged to ARS Control Laboratory,
sections, we considered only monochromatic images due to Department of Sciences and Methods for Engineering, UNI-
hardware limitations for experiments. By varying the image MORE, Italy for their guidance and support.
size, we obtained different print areas, whereas, by varying
the image samples for a fixed size, we obtained different R EFERENCES
cell patterns. The difference in cellularization causes minor [1] M. A. Hsieh, Á. Halász, S. Berman, and V. Kumar, “Biologically
variations in printing time as the distribution of black pixel inspired redistribution of a swarm of robots among multiple sites,”
varies. We considered two different scenarios: Swarm Intelligence, vol. 2, no. 2, pp. 121–141, 2008.
[2] M. Brambilla, E. Ferrante, M. Birattari, and M. Dorigo, “Swarm
1) A system composed of a varying number of robots, robotics: a review from the swarm engineering perspective,” Swarm
from N = 2 to N = 10, for fixed image size. Intelligence, vol. 7, no. 1, pp. 1–41, 2013.
2) A fixed system size with N = 6 robots, for varying [3] J. Cortés and M. Egerstedt, “Coordinated control of multi-robot
systems: A survey,” SICE Journal of Control, Measurement, and
image sizes and a system with N = 1 robot. System Integration, vol. 10, no. 6, pp. 495–503, 2017.
Statistical analysis of the simulation results is shown in [4] C. Ghedini, C. Ribeiro, and L. Sabattini, “Toward fault-tolerant multi-
Fig. 5. In particular, Fig. 5a shows the achieved elapsed robot networks,” Networks, vol. 70, no. 4, pp. 388–400, 2017.
[5] C. Ghedini, C. H. C. Ribeiro, and L. Sabattini, “Toward efficient
printing time, for a variable number of robots. It can be adaptive ad-hoc multi-robot network topologies,” Ad Hoc Networks,
observed from the plot that the printing time significantly vol. 74, pp. 57 – 70, 2018.
decreases as the number of robots in the simulation increase, [6] M. Minelli, M. Kaufmann, J. Panerati, C. Ghedini, G. Beltrame,
and L. Sabattini, “Stop, think, and roll: Online gain optimization for
but the difference stabilizes as the number of robots are resilient multi-robot topologies,” in Proceedings of the International
increased further. An extended work on this topic would be Symposium on Distributed Autonomous Robotic Systems (DARS),
determining the optimal number of robots for a given image. (Boulder, CO, USA), oct. 2018.
[7] K. Saulnier, D. Saldana, A. Prorok, G. J. Pappas, and V. Kumar,
Fig. 5b shows the comparison of elapsed time between “Resilient flocking for mobile robot teams,” IEEE Robotics and
a conventional printer (assumed to be equivalent of using Automation Letters, vol. 2, no. 2, pp. 1039–1046, 2017.

737
 400
N=1 robot

700
350

600
300

Elapsed Printing Time

Elapsed Printing Time

500
250
400

200
300

150 200

100 100

0
50
2 3 4 5 6 7 8 9 10 1000 2000 4000 5000 7000 9,000 10,000 20,000 40,000 80,000
Number of SPRINTER Robots Total Print Area

(a) (b)
Fig. 5: (a) contrasts the printing times with varying number of SPRINTER robots for a fixed printing area, starting from
N = 2, all the way to N = 10. (b) shows the time comparison between N = 1 and N = 6 for 10 different image areas and
10 image samples for each area. The line plot represents a system with N = 1 robot and the boxplot represents a system
with N = 6 robots.

(a) (b) (c)

Fig. 6: (a) The actual SPRINTER robot used in the experiments. (b) The three layers of the robot. (c) The custom circuit
board that controls the electronics of the SPRINTER.

[8] M. Rubenstein, C. Ahler, and R. Nagpal, “Kilobot: A low cost scalable [15] Y. Zhou and G. S. Chirikjian, “Probabilistic models of dead-reckoning
robot system for collective behaviors,” in Robotics and Automation error in nonholonomic mobile robots,” in Robotics and Automation,
(ICRA), 2012 IEEE International Conference on, pp. 3293–3298, 2003. Proceedings. ICRA’03. IEEE International Conference on,
IEEE, 2012. vol. 2, pp. 1594–1599, IEEE, 2003.
[9] D. Pickem, M. Lee, and M. Egerstedt, “The gritsbot in its natural [16] P. Bézier, “Numerical control: mathematics and applications,” 1970.
habitat-a multi-robot testbed,” in Robotics and Automation (ICRA), [17] G. Hunt, F. Mitzalis, T. Alhinai, P. A. Hooper, and M. Kovac, “3d
2015 IEEE International Conference on, pp. 4062–4067, IEEE, 2015. printing with flying robots,” in Robotics and Automation (ICRA), 2014
[10] F. Mondada, M. Bonani, X. Raemy, J. Pugh, C. Cianci, A. Klaptocz, IEEE International Conference on, pp. 4493–4499, IEEE, 2014.
S. Magnenat, J.-C. Zufferey, D. Floreano, and A. Martinoli, “The e- [18] D. Cappelleri, D. Efthymiou, A. Goswami, N. Vitoroulis, and M. Za-
puck, a robot designed for education in engineering,” in Proceedings vlanos, “Towards mobile microrobot swarms for additive micro-
of the 9th conference on autonomous robot systems and competitions, manufacturing,” International Journal of Advanced Robotic Systems,
vol. 1, pp. 59–65, IPCB: Instituto Politécnico de Castelo Branco, 2009. vol. 11, no. 9, p. 150, 2014.
[11] P. R. Wurman, R. D’Andrea, and M. Mountz, “Coordinating hundreds [19] N. Oxman, J. Duro-Royo, S. Keating, B. Peters, and E. Tsai, “Towards
of cooperative, autonomous vehicles in warehouses,” AI magazine, robotic swarm printing,” Architectural Design, vol. 84, no. 3, pp. 108–
vol. 29, no. 1, p. 9, 2008. 115, 2014.
[20] M. Turpin, N. Michael, and V. Kumar, “Trajectory planning and
[12] L. Sabattini, M. Aikio, P. Beinschob, M. Boehning, E. Cardarelli, assignment in multirobot systems,” in Algorithmic foundations of
V. Digani, A. Krengel, M. Magnani, S. Mandici, F. Oleari, C. Reinke, robotics X, pp. 175–190, Springer, 2013.
D. Ronzoni, C. Stimming, R. Varga, A. Vatavu, S. Castells Lopez, [21] E. Olson, “Apriltag: A robust and flexible visual fiducial system,” in
C. Fantuzzi, A. Mäyrä, S. Nedevschi, C. Secchi, and K. Fuerstenberg, Robotics and Automation (ICRA), 2011 IEEE International Conference
“The pan-robots project: Advanced automated guided vehicle systems on, pp. 3400–3407, IEEE, 2011.
for industrial logistics,” IEEE Robotics Automation Magazine, vol. 25, [22] J. E. Bresenham, “Algorithm for computer control of a digital plotter,”
pp. 55–64, March 2018. IBM Systems journal, vol. 4, no. 1, pp. 25–30, 1965.
[13] V. Digani, L. Sabattini, and C. Secchi, “A probabilistic Eulerian
traffic model for the coordination of multiple AGVs in automatic
warehouses,” IEEE Robotics and Automation Letters, vol. 1, pp. 26–
32, jan. 2016.
[14] K.-H. Lee and J.-H. Kim, “Multi-robot cooperation-based mobile
printer system,” Robotics and Autonomous Systems, vol. 54, no. 3,
pp. 193–204, 2006.

738
PROJECT REPORT – 4
4. Evaluation Rubrics
 SRM Institute of Science & Technology
College of Engineering and Technology
Department of Electronics and Communication Engineering

EVALUATION PROCESS TO IDENTIFY BEST AND AVERAGE PROJECTS

The Major project is assessed and evaluated based on Program Outcomes achievement
which covers Problem analysis, Design component, Investigation Methodology, Usage of
contemporary tools, Project management and Presentation. Best and average project are
assesses using evaluation rubrics applied on Project Report, Presentation and Demonstration.
A. The Project Work will be assessed using the Assessment Rubrics given below
 Project goals and problems are clearly identified. The chosen solution was well
thought of.
 Design strategy development which includes, plan to solve the problem,
decomposition of work into subtasks, and development of a timeline using Gantt
chart.
 The implementation (also problem solving) is very systematic. Proper assumptions
made; results are correctly analysed and interpreted.
 Properly choose and correctly use all the techniques, skills, and modern engineering
tools for their project.
 Understanding on the impact of engineering solutions in a global, economic,
environmental, and societal context and he/she provides an in-depth discussion.
 Deep understanding of the professional issues involved and the ethical implications of
the project, system, etc.
 Information is presented in a logical, interesting way, which is easy to follow. Purpose
is clearly stated and explains the structure of work.
 Student can demonstrate effective project management skills and problem solving
techniques related to project management. Can apply the management principles such
as cost benefit analysis, strategic alignment and project portfolio management and
project performance analysis and metrics. Can deliver successful projects at a faster
pace in increasingly complex environments. Can demonstrate a strong understanding
of project finance and the various metrics associated with the monitoring of the
financial health of the project.
 Capability of doing research on his/her own, i.e. he/she can do a complete research
related to the project.
B. Project Report is assessed based on the assessment rubrics given in Table 1.
Table 1: Project Report Assessment Rubrics

Particulars Exceptional
Objective Objective complete and well-written; provides all necessary background
principles for the experiment

Content  Technically correct

 Contain in-depth and complete details of the project
 An engineer can recreate the project based on the report.
Language (Word  Sentences are complete and grammatical. They flow together easily
Choice,  Words are chosen for their precise meaning.
Grammar)  Engineering terms and jargon are used correctly.
 No misspelled words.
Experimental Well-written in paragraph format, all experimental details are covered
procedure
Numerical Usage  All figures, graphs, charts, and drawings are accurate, consistent with
and Illustrations the text, and of good quality. They enhance understanding of the text.
 All items are labeled and referred to in the text.
 All equations are clear, accurate, and labeled. All variables are defined
and units specified. Discussion about the equation development and
use is stated.
Results,  All important trends and data comparisons have been interpreted
Discussion and correctly and discussed, good understanding of results is conveyed.
Conclusions  All important conclusions have been clearly made, student shows good
understanding
Visual Format  Structuring the content to represent the logical progression
and  The doc. is visually appealing and easily navigated.
Organization  Usage of white space is used as appropriate to separate blocks of text
and add emphasis.
Use of references  Prior work is acknowledged by referring to sources for theories,
assumptions, quotations, and findings.
 Correct information for References.

Realistic  Incorporates appropriate multiple realistic constraints such as

constraints economic, environmental, social, political, ethical, health and safety,
manufacturability, and sustainability
 Analysis provides correct reasons as how this constraint affects the
design of the system, component, or process and contains in-depth
discussion.
Engineering Clear evidence of ability to use engineering principles to design
Standards components, devices or systems
C. Project Presentation is assessed based on the assessment rubrics given in Table 2.
Table 2: Project Presentation Assessment Rubrics

Particulars Exceptional
 Presentation contains all required components
Content  A complete explanation of major concepts and theories is provided
and drawn upon relevant literature
 Content is consistently accurate
Organization  Presentation is clear, logical and organized
 Audience can follow line of reasoning
 Presenters are comfortable in front of audience and his/her voice is
Professional audible
delivery  No reading from the notes or presentation
 Sentences are complete and grammatical, and they flow together
easily
Visual Aids  ability to understand the message
 grammar and choice of words
Conclusion of  Planned concluding remarks (not just “I guess that’s it.”)
presentation  Presented significant results
Responses to  Listened to questions without interrupting
questions  Began with general answer and then followed up with details

D. Project Demonstration is assessed based on the assessment rubrics given in Table 3.

Table 3: Project Demonstration Assessment Rubrics

Particulars Exceptional
 Clearly identifies and discusses focus/purpose of project.
Introduction  A complete explanation of major concepts and theories is provided
and drawn upon relevant literature.
 Presented the detailed design, including modelling, control design,
simulation, and experimental results, with diagrams and parameter
values.
Methodology  Compared simulation and experimental results. Compared achieved
performance with the design specification.
 Provided solid technical data, and presented it in an easily grasped
manner, using graphs where possible.
 Have all the materials required for the project demonstration
Organization &  All these materials are neatly organized so that the demonstration
Presentation runs smoothly
 Speech, confidence, knowledge and enthusiasm are inspirational
 Good eye contact and voice projection maintained throughout the
 entire presentation
 Group understands what they are doing and carries out the
demonstration as planned in an enthusiastic manner. There is a very
good understanding of the "how and why" of the project
Interest/Excitement  Demonstration was very interesting and captured the excitement of
all those viewing the presentation.
Professionalism  Respectable at all times. Shows extensive practice and preparation.
No safety issues during demonstration.
Social Impact and  The project has an authentic context, involves real-world tasks, tools,
Authenticity and quality standards, and makes a real impact on the world.
 Incorporates appropriate multiple realistic constraints such as
economic, environmental, social, political, ethical, health and safety,
Realistic manufacturability, and sustainability.
constraints  Analysis provides correct reasons as how this constraint affects the
design of the system, component, or process and contains in-depth
discussion.
Engineering  Incorporates appropriate engineering standards that defines the
Standards characteristics of a product, process or service, such as dimensions,
safety aspects, and performance requirements.
 Results are clearly explained in a comprehensive level of detail and
Results, Discussion are well-organized.
& Conclusion  Interpretations/ analysis of results are thoughtful and insightful
 Suggestions for further research in this area are provided and are
appropriate

E. Publications
Students are encouraged to publish their contribution of major project outcomes in reputed
indexed or non-indexed journals/ conferences. Based on their publication the outcome of the
project work is gauged. Students are advised to publish their research articles in
Scopus/SCI indexed Journals.
F. Best Practices in Major Project:
COMSPRO is the Major Project Design contest conducted every year in the department to
showcase the top 3 projects chosen from each domain by the respective project coordinators,
to the pre-final and second year students to motivate them to improve their design skills.
Judges were identified for the COMSPRO and were asked to select the winners of the
contest. The purpose of this design contest is to increase the student motivation,
engagement, confidence, self-perceptions and demonstration of the learning proficiency.
The preparatory work involved in the conduction of COMSPRO for the remaining
years say AY 2018-19 and 2017-18 are as follows:
 COMSPRO banner for wide publicity
 Evaluation Criteria for Judges
 Announcement of Winners
 Certificate for Best Project Award
 PROJECT REPORT – 4
5. Assessment record for Review 1,
2, 3 and CO & PO Mapping
 Review-I
NOVELTY/OBJECTIVE METHODOLOGY
PPT (5 marks)
(5 marks) (5 marks)
Mr. Gudide TOTAL TOTAL
S.No Domain Register.No Student Names Project Title Proj. Guide Dr.CT. Dr.CT.
Mr. AVM. Dr. Brajesh AVM. Dr. Brajesh Mr. AVM. Dr. Brajesh Dr.CT. 5 (15MARKS) (20MARKS)
Manimegal Manimega
Batch Manikandan Kumar Manikand Kumar Manikanda Kumar Manimegalai
ai lai
(5marks) (5 marks) an (5 marks) n (5marks) (5 marks) (5 marks)
(5 marks) (5 marks)
(5marks)

1 ESTVK 1 RA1511004010810 Nikhil Agarwal 4 4 4 4 4 4 4 4 4 4.6 12 16.6

2 ESTVK RA1511004010037 Vishwas Anandani Coal Mine safety Dr. 4 4 4 4 4 4 4 4 4 4.6 12 16.6
monitoring and control Chittaranjan
3 ESTVK RA1511004010812 Julian Jojo automation Nayak 4 4 4 4 4 4 4 4 4 4.6 12 16.6
4 ESTVK RA1511004010776 Piyush kumar Singh 2 3 3 3 3 3 2 3 3 4.6 9 13.6
5 ESTVK RA1511004010610 Abhijeet Kumar 4 4 4 4 4 4 4 4 4 5 12 17
6 ESTVK RA1511004010572 Ashmita Sarkar Landslide detection Dr. K 4 4 4 4 4 4 4 4 4 5 12 17
2
7 RA1511004010357 and prediction system Kalimuthu 4 4 4 4 4 4 4 4 4 5 12
ESTVK Sneha Dubey 17
8 ESTVK RA1511004010343 Chiranshi Baxi 4 4 4 4 4 4 4 4 4 5 12 17
9 ESTVK RA1511004010672 Shubhendu Ojas Tewary 5 4 5 5 4 5 5 4 5 4.8 14 18.8
10 ESTVK RA1511004010752 Aishwarya AR AGROBOT:Bot for Dr. Neelaveni 4 4 4 4 4 5 4 4 5 4.8 14 18.8
3
11 RA1511004010567 agriculture assistance Ammal 4 4 4 4 4 4 4 4 4 4.8 12
ESTVK Anoushka Shukla 16.8
12 ESTVK RA1511004010701 Ashish Sharma 3 4 4 3 4 4 3 4 4 4.8 12 16.8
13 ESTVK RA1511004010553 kedar prasad karpe 5 5 5 5 5 5 5 5 5 4.6 15 19.6
14 ESTVK RA1511004010511 Dhruv pant cooperative transport 3 4 4 3 4 4 4 4 4 4.6 12 16.6
4 using Multi-robot DR. R .Kumar
15 ESTVK RA1511004010712 Nimish pastaria system - 5 5 - 5 5 - 5 5 4.6 15 19.6
16 ESTVK RA1511004010654 jayati singh 3 4 4 3 4 4 4 4 4 4.6 12 16.6
17 ESTVK RA1511004010322 Ayushi Kandoi 5 5 5 5 5 4 5 5 4 5 15 20
18 ESTVK RA1511004010352 Pradyumna Hegade 5 5 5 5 5 4 5 5 4 5 15 20
Dr.Dhanalaks
5 Surveillance Robot
19 ESTVK RA1511004010320 Utkarsh Goel hmi 5 5 5 5 5 4 5 5 4 5 15 20
Venkata Harshavardhan
20 ESTVK RA1511004010306 5 5 5 5 5 4 5 5 4 5 15 20
Reddy Allu
21 ESTVK RA1511004010788 Botta venkata akhil 4 4 4 4 4 4 4 4 4 4.6 12 16.6
22 ESTVK RA1511004010774 Akhil rayana 4 4 4 4 4 4 4 4 4 4.6 12 16.6
Health Monitoring Dr.Diwakar R
6
23 ESTVK RA1511004010698 Bolugoddu sai dattathreya System Using IOT Marur 4 4 3 4 4 3 4 4 3 4.6 11 15.6

24 ESTVK RA1511004010805 Manchikanti tarun krishna 3 3 3 4 4 3 4 4 3 4.6 11 15.6

 NOVELTY/OBJECTIVE METHODOLOGY
PPT (5 marks)
(5 marks) (5 marks)
Mr. Gudide TOTAL TOTAL
S.No Domain Register.No Student Names Project Title Proj. Guide Dr.CT. Dr.CT.
Mr. AVM. Dr. Brajesh AVM. Dr. Brajesh Mr. AVM. Dr. Brajesh Dr.CT. 5 (15MARKS) (20MARKS)
Manimegal Manimega
Batch Manikandan Kumar Manikand Kumar Manikanda Kumar Manimegalai
ai lai
(5marks) (5 marks) an (5 marks) n (5marks) (5 marks) (5 marks)
(5 marks) (5 marks)
(5marks)

25 ESTVK RA1511004010593 MOHIT ANAND 4 4 4 3 3 3 4 4 4 4.4 12 16.4

26 ESTVK RA1511004010594 AKASH KUMAR CoalBased
IOT MineVehicle
safety Dr. 4 4 4 3 3 3 4 4 4 4.4 12 16.4
7 monitoring and control Dr.J.MANJUL
License Tracking Chittaranjan
27 RA1511004010590 automation A
Nayak 4 4 4 3 3 3 4 4 4 4.4 12
ESTVK CHIRAG BHATLA System 16.4
28 ESTVK RA1511004010588 ANCHAL SOHAL 4 4 4 3 3 3 4 4 4 4.4 12 16.4
29 ESTVK RA1511004010683 ANJIMA S NAIR 3 4 4 3 4 4 3 4 5 4.8 14 18.8
A Human Robot
30 ESTVK RA1511004010686 KRISHNA MANOJ Environment 5 5 5 5 5 5 5 5 5 4.8 15 19.8
Dr.K.VIJAYA
8 Interactive Reasoning
31 RA1511004010696 N 5 5 5 5 5 5 5 5 5 4.8 15
ESTVK D.SHRUTHI Mechanism For Object 19.8
Sorting Robot
32 ESTVK RA1511004010700 ROHAN SHARMA 5 5 5 5 5 5 5 5 5 4.8 15 19.8
33 ESTVK RA1511004010254 Syed Sharique Ahmad 3 3 3 3 3 3 3 3 3 5 9 14
34 ESTVK RA1511004010360 Anuroop Divakaran Calcolator for the 3 3 3 3 3 3 3 3 3 5 9 14
9 Ms.Srilekha
35 ESTVK RA1511004010338 Jaya Kumar Visually impared 3 3 3 3 3 3 3 3 3 5 9 14
36 ESTVK RA1511004010308 jhabar mal 3 3 3 3 3 3 3 3 3 5 9 14
37 EST-2 RA1511004010692 Visvender Singh 5 4 4 5 4 4 5 4 4 4.4 15 19.4
38 EST-2 RA1511004010684 Rishabh Verma 3 3 3 3 3 3 3 3 3 4.4 9 13.4
Mr. E.Siva
10 Multi Utility Drone
39 EST-2 RA1511004010671 Ritesh Kumar Kumar 4 4 4 4 4 4 4 4 4 4.4 12 16.4
Pawan Kumar
40 EST-2 RA1511004010215 - 2 2 - 2 2 - 2 2 4.4 6 10.4
Saini(eduvirtuoso)
41 EST-2 RA1511004010456 Sakshi awal 4 4 4 4 4 4 4 4 4 4.2 12 16.2
Design and
Goty muhammad implementation of IOT
42 EST -2 RA1511004010469 Mr. K. - 4 - - 4 - - 4 4 4.2 12 16.2
11 ishaq(cognizant) based actutator and
Ramesh
43 EST-2 RA1511004010473 Shaikh mohammad Rahil sensor sytem for 5 5 5 5 5 5 5 5 5 4.2 15 19.2
multistorage building
44 EST-2 RA1511004010459 Shayan ghosh 4 4 4 4 4 4 4 4 4 4.2 12 16.2
45 EST-2 RA1511004010400 Sushobhan Misra 4 4 4 4 4 4 4 4 4 4.8 12 16.8
IOT for Consumer Mr. K.
46 EST-2 12 RA1511004010475 Sangrila Joyesh 4 4 4 5 5 4 4 4 4 4.8 14 18.8
Appliance Application Ramesh
47 EST-2 RA1511004010509 Ashish Deshwal 5 5 4 4 4 5 5 5 5 4.8 15 19.8
N.Prasanth
48 EST-2 RA1511004010071 3 4 4 3 3 3 3 3 3 4.8 12 16.8
Narayan(infosys)
Highly Secured
49 EST-2 13 RA1511004010533 Piyush Gulati Biometric voting Mr. S Nivash 3 4 4 3 3 3 3 3 3 4.8 12 16.8
50 EST-2 RA1511004010022 Gollapudi Raj Bharath system 3 3 3 3 4 3 3 3 3 4.8 11 15.8
51 EST-2 RA1511004010136 Chikkala Surya Pradeep 3 3 3 3 3 3 3 4 4 4.8 11 15.8
 NOVELTY/OBJECTIVE METHODOLOGY
PPT (5 marks)
(5 marks) (5 marks)
Mr. Gudide TOTAL TOTAL
S.No Domain Register.No Student Names Project Title Proj. Guide Dr.CT. Dr.CT.
Mr. AVM. Dr. Brajesh AVM. Dr. Brajesh Mr. AVM. Dr. Brajesh Dr.CT. 5 (15MARKS) (20MARKS)
Manimegal Manimega
Batch Manikandan Kumar Manikand Kumar Manikanda Kumar Manimegalai
ai lai
(5marks) (5 marks) an (5 marks) n (5marks) (5 marks) (5 marks)
(5 marks) (5 marks)
(5marks)

Kariyavula Mounish
52 EST-2 RA1511004010029 4 4 4 4 4 4 4 4 4 4.8 12 16.8
Naidu
Coal Mine safety Dr.
Intelligent System for
53 EST-2 RA1511004010214 Kollipari Avinash monitoring and control Chittaranjan 4 4 4 4 4 4 4 4 4 4.8 12 16.8
Human Computer
14 automation Mr.Sriram
Nayak
Mummadi Ravi Teja Interface using Hand
54 EST-2 RA1511004010236 4 4 4 4 4 4 4 4 4 4.8 12 16.8
Reddy(congnizant) Gesture Recognition

55 EST-2 RA1511004010178 Chinthala Vineeth Reddy 4 4 4 4 4 4 4 4 4 4.8 12 16.8

56 EST-2 RA1511004010429 Abhishek Mukherjee 4 4 4 4 4 4 4 4 4 4.8 12 16.8
57 EST -2 RA1511004010379 Hetal Sahu Eye Blink Controlled 4 4 4 4 4 4 4 4 4 4.8 12 16.8
15 virtual keyboard using Mr.Bashyam
58 EST-2 RA1511004010377 Sudhanshu Kanth(377) Raspberry Pi 5 5 5 5 5 5 5 5 5 4.8 15 19.8
59 EST-2 RA1511004010455 Surendra 3 4 4 3 4 4 3 4 4 4.8 12 16.8
60 EST-2 RA1511004010527 Nikhil Datta(HPE) 4 4 4 4 4 4 4 4 4 4.2 12 16.2
61 EST-2 RA1511004010623 Gyaneshwar Pal Agriculture Land 4 4 4 4 4 4 4 4 4 4.2 12 16.2
Mrs. D.
16 Monitoring with Dry
62 Vijayalaksmi 4 4 4 4 4 4 4 4 4 4.2 12
EST-2 RA1511004010569 Sidharth Thakur Leaf Detection 16.2
63 EST-2 RA1511004010613 Shravan.S 3 3 3 3 3 3 3 3 3 4.2 9 13.2
64 EST-2 RA1511004010658 Samriddhi Shanker 5 5 5 5 5 5 5 5 5 5 15 20
65 EST-2 RA1511004010668 Garima Sharma 4 4 4 4 4 4 4 4 4 5 12 17
Autonomous way point Mrs. Anilet
17
66 EST-2 RA1511004010670 Aditya Kumar navigation with UGV Bala 4 4 4 4 4 4 4 4 4 5 12 17
67 EST-2 RA1511004010688 Vipul Notani(Continental) 5 5 5 5 5 5 5 5 5 5 15 20

68 EST-2 RA1511004010211 Prerana Roy(High radius) 4 4 4 4 4 4 4 4 4 4 12 16

Automated Detection
69 EST-2 RA1511004010383 Raveena Daawat(amazon) of white blood cell 4 4 4 4 4 4 4 4 4 4 12 16
18 Mrs. E. Chitra
cancer diseases
70 EST -2 RA1511004010189 Aswin kuamr s updated in IOT 4 4 4 4 4 4 4 4 4 4 12 16
71 EST-2 RA1511004010219 Kondagari Sai Nikhila 4 4 4 4 4 4 4 4 4 4 12 16

Prof. Incharge
HOD/ECE Project Coordinator
 Review-II
NOVELTY/OBJECTIVE METHODOLOGY
PPT (5 marks)
(5 marks) (5 marks)
Mr. Gudide TOTAL TOTAL
S.No Domain Batch Register.No Student Names Project Title Proj. Guide Dr.CT. Dr.CT.
Mr. AVM. Dr. Brajesh AVM. Dr. Brajesh Mr. AVM. Dr. Brajesh Dr.CT. 5 (15MARKS) (20MARKS)
Manimegal Manimega
Manikandan Kumar Manikand Kumar Manikanda Kumar Manimegalai
ai lai
(5marks) (5 marks) an (5 marks) n (5marks) (5 marks) (5 marks)
(5 marks) (5 marks)
(5marks)

1 ESTVK 1 RA1511004010810 Nikhil Agarwal 4 4 4 4 4 4 4 4 4 4.6 12 16.6

2 ESTVK RA1511004010037 Vishwas Anandani Coal Mine safety Dr. 4 4 4 4 4 4 4 4 4 4.6 12 16.6
monitoring and control Chittaranjan
3 ESTVK RA1511004010812 Julian Jojo automation Nayak 4 4 4 4 4 4 4 4 4 4.6 12 16.6
4 ESTVK RA1511004010776 Piyush kumar Singh 2 3 3 3 3 3 2 3 3 4.6 9 13.6
5 ESTVK RA1511004010610 Abhijeet Kumar 4 4 4 4 4 4 4 4 4 5 12 17
6 ESTVK RA1511004010572 Ashmita Sarkar Landslide detection Dr. K 4 4 4 4 4 4 4 4 4 5 12 17
2
7 RA1511004010357 and prediction system Kalimuthu 4 4 4 4 4 4 4 4 4 5 12
ESTVK Sneha Dubey 17
8 ESTVK RA1511004010343 Chiranshi Baxi 4 4 4 4 4 4 4 4 4 5 12 17
9 ESTVK RA1511004010672 Shubhendu Ojas Tewary 5 4 5 5 4 5 5 4 5 4.8 14 18.8
10 ESTVK RA1511004010752 Aishwarya AR AGROBOT:Bot for Dr. Neelaveni 4 4 4 4 4 5 4 4 5 4.8 14 18.8
3
11 RA1511004010567 agriculture assistance Ammal 4 4 4 4 4 4 4 4 4 4.8 12
ESTVK Anoushka Shukla 16.8
12 ESTVK RA1511004010701 Ashish Sharma 3 4 4 3 4 4 3 4 4 4.8 12 16.8
13 ESTVK RA1511004010553 kedar prasad karpe 5 5 5 5 5 5 5 5 5 4.6 15 19.6
14 ESTVK RA1511004010511 Dhruv pant cooperative transport 3 4 4 3 4 4 4 4 4 4.6 12 16.6
4 using Multi-robot DR. R .Kumar
15 ESTVK RA1511004010712 Nimish pastaria system - 5 5 - 5 5 - 5 5 4.6 15 19.6
16 ESTVK RA1511004010654 jayati singh 3 4 4 3 4 4 4 4 4 4.6 12 16.6
17 ESTVK RA1511004010322 Ayushi Kandoi 5 5 5 5 5 4 5 5 4 5 15 20
18 ESTVK RA1511004010352 Pradyumna Hegade 5 5 5 5 5 4 5 5 4 5 15 20
Dr.Dhanalaks
5 Surveillance Robot
19 ESTVK RA1511004010320 Utkarsh Goel hmi 5 5 5 5 5 4 5 5 4 5 15 20
Venkata Harshavardhan
20 ESTVK RA1511004010306 5 5 5 5 5 4 5 5 4 5 15 20
Reddy Allu
21 ESTVK RA1511004010788 Botta venkata akhil 4 4 4 4 4 4 4 4 4 4.6 12 16.6
22 ESTVK RA1511004010774 Akhil rayana 4 4 4 4 4 4 4 4 4 4.6 12 16.6
Health Monitoring Dr.Diwakar R
6
23 ESTVK RA1511004010698 Bolugoddu sai dattathreya System Using IOT Marur 4 4 3 4 4 3 4 4 3 4.6 11 15.6

24 ESTVK RA1511004010805 Manchikanti tarun krishna 3 3 3 4 4 3 4 4 3 4.6 11 15.6

 NOVELTY/OBJECTIVE METHODOLOGY
PPT (5 marks)
(5 marks) (5 marks)
Mr. Gudide TOTAL TOTAL
S.No Domain Batch Register.No Student Names Project Title Proj. Guide Dr.CT. Dr.CT.
Mr. AVM. Dr. Brajesh AVM. Dr. Brajesh Mr. AVM. Dr. Brajesh Dr.CT. 5 (15MARKS) (20MARKS)
Manimegal Manimega
Manikandan Kumar Manikand Kumar Manikanda Kumar Manimegalai
ai lai
(5marks) (5 marks) an (5 marks) n (5marks) (5 marks) (5 marks)
(5 marks) (5 marks)
(5marks)

25 ESTVK RA1511004010593 MOHIT ANAND 4 4 4 3 3 3 4 4 4 4.4 12 16.4

26 ESTVK RA1511004010594 AKASH KUMAR CoalBased
IOT MineVehicle
safety Dr. 4 4 4 3 3 3 4 4 4 4.4 12 16.4
7 monitoring and control Dr.J.MANJUL
License Tracking Chittaranjan
27 RA1511004010590 automation A
Nayak 4 4 4 3 3 3 4 4 4 4.4 12
ESTVK CHIRAG BHATLA System 16.4
28 ESTVK RA1511004010588 ANCHAL SOHAL 4 4 4 3 3 3 4 4 4 4.4 12 16.4
29 ESTVK RA1511004010683 ANJIMA S NAIR 3 4 4 3 4 4 3 4 5 4.8 14 18.8
A Human Robot
30 ESTVK RA1511004010686 KRISHNA MANOJ Environment 5 5 5 5 5 5 5 5 5 4.8 15 19.8
Dr.K.VIJAYA
8 Interactive Reasoning
31 RA1511004010696 N 5 5 5 5 5 5 5 5 5 4.8 15
ESTVK D.SHRUTHI Mechanism For Object 19.8
Sorting Robot
32 ESTVK RA1511004010700 ROHAN SHARMA 5 5 5 5 5 5 5 5 5 4.8 15 19.8
33 ESTVK RA1511004010254 Syed Sharique Ahmad 3 3 3 3 3 3 3 3 3 5 9 14
34 ESTVK RA1511004010360 Anuroop Divakaran Calcolator for the 3 3 3 3 3 3 3 3 3 5 9 14
9 Ms.Srilekha
35 ESTVK RA1511004010338 Jaya Kumar Visually impared 3 3 3 3 3 3 3 3 3 5 9 14
36 ESTVK RA1511004010308 jhabar mal 3 3 3 3 3 3 3 3 3 5 9 14
37 EST-2 RA1511004010692 Visvender Singh 5 4 4 5 4 4 5 4 4 4.4 15 19.4
38 EST-2 RA1511004010684 Rishabh Verma 3 3 3 3 3 3 3 3 3 4.4 9 13.4
Mr. E.Siva
10 Multi Utility Drone
39 EST-2 RA1511004010671 Ritesh Kumar Kumar 4 4 4 4 4 4 4 4 4 4.4 12 16.4
Pawan Kumar
40 EST-2 RA1511004010215 - 2 2 - 2 2 - 2 2 4.4 6 10.4
Saini(eduvirtuoso)
41 EST-2 RA1511004010456 Sakshi awal 4 4 4 4 4 4 4 4 4 4.2 12 16.2
Design and
Goty muhammad implementation of IOT
42 EST -2 RA1511004010469 Mr. K. - 4 - - 4 - - 4 4 4.2 12 16.2
11 ishaq(cognizant) based actutator and
Ramesh
43 EST-2 RA1511004010473 Shaikh mohammad Rahil sensor sytem for 5 5 5 5 5 5 5 5 5 4.2 15 19.2
multistorage building
44 EST-2 RA1511004010459 Shayan ghosh 4 4 4 4 4 4 4 4 4 4.2 12 16.2
45 EST-2 RA1511004010400 Sushobhan Misra 4 4 4 4 4 4 4 4 4 4.8 12 16.8
IOT for Consumer Mr. K.
46 EST-2 12 RA1511004010475 Sangrila Joyesh 4 4 4 5 5 4 4 4 4 4.8 14 18.8
Appliance Application Ramesh
47 EST-2 RA1511004010509 Ashish Deshwal 5 5 4 4 4 5 5 5 5 4.8 15 19.8
N.Prasanth
48 EST-2 RA1511004010071 3 4 4 3 3 3 3 3 3 4.8 12 16.8
Narayan(infosys)
Highly Secured
49 EST-2 13 RA1511004010533 Piyush Gulati Biometric voting Mr. S Nivash 3 4 4 3 3 3 3 3 3 4.8 12 16.8
50 EST-2 RA1511004010022 Gollapudi Raj Bharath system 3 3 3 3 4 3 3 3 3 4.8 11 15.8
51 EST-2 RA1511004010136 Chikkala Surya Pradeep 3 3 3 3 3 3 3 4 4 4.8 11 15.8
 NOVELTY/OBJECTIVE METHODOLOGY
PPT (5 marks)
(5 marks) (5 marks)
Mr. Gudide TOTAL TOTAL
S.No Domain Batch Register.No Student Names Project Title Proj. Guide Dr.CT. Dr.CT.
Mr. AVM. Dr. Brajesh AVM. Dr. Brajesh Mr. AVM. Dr. Brajesh Dr.CT. 5 (15MARKS) (20MARKS)
Manimegal Manimega
Manikandan Kumar Manikand Kumar Manikanda Kumar Manimegalai
ai lai
(5marks) (5 marks) an (5 marks) n (5marks) (5 marks) (5 marks)
(5 marks) (5 marks)
(5marks)

Kariyavula Mounish
52 EST-2 RA1511004010029 4 4 4 4 4 4 4 4 4 4.8 12 16.8
Naidu
Coal Mine safety Dr.
Intelligent System for
53 EST-2 RA1511004010214 Kollipari Avinash monitoring and control Chittaranjan 4 4 4 4 4 4 4 4 4 4.8 12 16.8
Human Computer
14 automation Mr.Sriram
Nayak
Mummadi Ravi Teja Interface using Hand
54 EST-2 RA1511004010236 4 4 4 4 4 4 4 4 4 4.8 12 16.8
Reddy(congnizant) Gesture Recognition

55 EST-2 RA1511004010178 Chinthala Vineeth Reddy 4 4 4 4 4 4 4 4 4 4.8 12 16.8

56 EST-2 RA1511004010429 Abhishek Mukherjee 4 4 4 4 4 4 4 4 4 4.8 12 16.8
57 EST -2 RA1511004010379 Hetal Sahu Eye Blink Controlled 4 4 4 4 4 4 4 4 4 4.8 12 16.8
15 virtual keyboard using Mr.Bashyam
58 EST-2 RA1511004010377 Sudhanshu Kanth(377) Raspberry Pi 5 5 5 5 5 5 5 5 5 4.8 15 19.8
59 EST-2 RA1511004010455 Surendra 3 4 4 3 4 4 3 4 4 4.8 12 16.8
60 EST-2 RA1511004010527 Nikhil Datta(HPE) 4 4 4 4 4 4 4 4 4 4.2 12 16.2
61 EST-2 RA1511004010623 Gyaneshwar Pal Agriculture Land 4 4 4 4 4 4 4 4 4 4.2 12 16.2
Mrs. D.
16 Monitoring with Dry
62 Vijayalaksmi 4 4 4 4 4 4 4 4 4 4.2 12
EST-2 RA1511004010569 Sidharth Thakur Leaf Detection 16.2
63 EST-2 RA1511004010613 Shravan.S 3 3 3 3 3 3 3 3 3 4.2 9 13.2
64 EST-2 RA1511004010658 Samriddhi Shanker 5 5 5 5 5 5 5 5 5 5 15 20
65 EST-2 RA1511004010668 Garima Sharma 4 4 4 4 4 4 4 4 4 5 12 17
Autonomous way point Mrs. Anilet
17
66 EST-2 RA1511004010670 Aditya Kumar navigation with UGV Bala 4 4 4 4 4 4 4 4 4 5 12 17
67 EST-2 RA1511004010688 Vipul Notani(Continental) 5 5 5 5 5 5 5 5 5 5 15 20

68 EST-2 RA1511004010211 Prerana Roy(High radius) 4 4 4 4 4 4 4 4 4 4 12 16

Automated Detection
69 EST-2 RA1511004010383 Raveena Daawat(amazon) of white blood cell 4 4 4 4 4 4 4 4 4 4 12 16
18 Mrs. E. Chitra
cancer diseases
70 EST -2 RA1511004010189 Aswin kuamr s updated in IOT 4 4 4 4 4 4 4 4 4 4 12 16
71 EST-2 RA1511004010219 Kondagari Sai Nikhila 4 4 4 4 4 4 4 4 4 4 12 16

HOD/ECE Project Coordinator

 Review-III

presentation (50) Pres Poster (150) Proje

entat ct
P1 P2 PC G ion P1 P2 PC G repor Proje
Fina
Proj. (25) Poster(3 t (50) ct Total
S.No Domain Batch Register.No Student Names Project Title l
Guide 0) repor (80)
Mr.
Dr. Dr.CT.
Mr.
Dr. Dr.CT.
(20)
AVM. AVM. t (25)
Brajesh Manimeg Guide(5 Brajesh Manime Guide(1 Guide(5
Manika Manika
Kumar alai 0 Kumar galai 50 0
ndan ndan
(50 (50 marks) (150 (150 marks) marks)
(50 (150
marks) marks) marks) marks)
marks) marks)

ESTVK RA1511004010810 Nikhil Agarwal 40 37 37 40 19.25 127 131 131 107 24.8 40 20 64.05 16

2 ESTVK RA1511004010037 Vishwas Anandani Coal Mine safety Dr. 40 37 37 40 19.25 127 131 131 107 24.8 40 20 64.05 16
1 monitoring and control Chittaranja
3 ESTVK RA1511004010812 Julian Jojo automation n Nayak 40 37 37 40 19.25 127 131 131 107 24.8 40 20 64.05 16

4 ESTVK RA1511004010776 Piyush kumar Singh 40 37 37 40 19.25 127 131 131 107 24.8 40 20 64.05 16

5 ESTVK RA1511004010610 Abhijeet Kumar 33 34 40 48 19.38 140 140 140 148 28.4 50 25 72.78 18.2

6 ESTVK RA1511004010572 Ashmita Sarkar 33 34 40 48 19.38 140 140 140 148 28.4 50 25 72.78 18.2
Landslide detection and Dr. K
2
prediction system Kalimuthu
7 ESTVK RA1511004010357 Sneha Dubey 33 34 40 48 19.38 140 140 140 148 28.4 50 25 72.78 18.2

8 ESTVK RA1511004010343 Chiranshi Baxi 33 34 40 48 19.38 140 140 140 148 28.4 50 25 72.78 18.2
Shubhendu Ojas
9 ESTVK RA1511004010672 47 47 47 50 23.88 145 138 138 149 28.5 50 25 77.38 19.3
Tewary

10 ESTVK RA1511004010752 Aishwarya AR Dr. 47 47 47 50 23.88 145 138 138 149 28.5 50 25 77.38 19.3
AGROBOT:Bot for
3 Neelaveni
agriculture assistance
11 ESTVK RA1511004010567 Anoushka Shukla Ammal 47 47 47 50 23.88 145 138 138 149 28.5 50 25 77.38 19.3

12 ESTVK RA1511004010701 Ashish Sharma 43 43 43 50 22.38 145 138 138 149 28.5 50 25 75.88 19

13 ESTVK RA1511004010553 kedar prasad karpe 48 48 48 50 24.25 143 146 146 150 29.25 50 25 78.5 19.6

14 ESTVK RA1511004010511 Dhruv pant 48 48 48 50 24.25 143 146 146 150 29.25 50 25 78.5 19.6
cooperative transport DR. R
4
using Multi-robot system .Kumar
15 ESTVK RA1511004010712 Nimish pastaria 48 48 48 50 24.25 143 146 146 150 29.25 50 25 78.5 19.6

16 ESTVK RA1511004010654 jayati singh 48 48 48 50 24.25 143 146 146 150 29.25 50 25 78.5 19.6

17 ESTVK RA1511004010322 Ayushi Kandoi 45 45 45 50 23.13 138 140 140 150 28.4 50 25 76.53 19.1

18 ESTVK RA1511004010352 Pradyumna Hegade 45 45 45 50 23.13 138 140 140 150 28.4 50 25 76.53 19.1
Dr.Dhanala
5 Surveillance Robot
kshmi
19 ESTVK RA1511004010320 Utkarsh Goel 45 45 45 45 22.5 138 140 140 150 28.4 50 25 75.9 19
Venkata
20 ESTVK RA1511004010306 Harshavardhan 45 45 45 50 23.13 138 140 140 150 28.4 50 25 76.53 19.1
Reddy Allu
21 ESTVK RA1511004010788 Botta venkata akhil 40 40 40 45 20.63 141 141 141 144 28.35 46 23 71.98 18

22 ESTVK RA1511004010774 Akhil rayana 40 40 40 32 19 141 141 141 144 28.35 46 23 70.35 17.6
Health Monitoring Dr.Diwaka
6
System Using IOT r R Marur
 Health Monitoring Dr.Diwaka
6
Bolugoddu sai System Using IOT r R Marur
23 ESTVK RA1511004010698 40 40 40 32 19 141 141 141 144 28.35 46 23 70.35 17.6
dattathreya
Manchikanti tarun
24 ESTVK RA1511004010805 40 40 40 32 19 141 141 141 144 28.35 46 23 70.35 17.6
krishna

25 ESTVK RA1511004010593 MOHIT ANAND 43 43 43 49 22.25 137 138 138 142 27.75 44 22 72 18

26 ESTVK RA1511004010594 AKASH KUMAR 43 43 43 49 22.25 137 138 138 142 27.75 44 22 72 18
IOT Based Vehicle Dr.J.MANJ
7
License Tracking System ULA
27 ESTVK RA1511004010590 CHIRAG BHATLA 43 43 43 49 22.25 137 138 138 142 27.75 44 22 72 18

28 ESTVK RA1511004010588 ANCHAL SOHAL 43 43 43 49 22.25 137 138 138 142 27.75 44 22 72 18

29 ESTVK RA1511004010683 ANJIMA S NAIR 38 42 43 46 21.13 137 140 140 144 28.05 50 25 74.18 18.5
A Human Robot
30 ESTVK RA1511004010686 KRISHNA MANOJ 38 42 43 45 21 137 140 140 144 28.05 50 25 74.05 18.5
Environment Interactive Dr.K.VIJA
8
Reasoning Mechanism For YAN
31 ESTVK RA1511004010696 D.SHRUTHI 38 42 43 47 21.25 137 140 140 144 28.05 50 25 74.3 18.6
Object Sorting Robot

32 ESTVK RA1511004010700 ROHAN SHARMA 38 42 43 46 21.13 137 140 140 144 28.05 50 25 74.18 18.5

33 ESTVK RA1511004010254 Syed Sharique Ahmad 37 38 40 48 20.38 111 134 134 144 26.15 42 21 67.53 16.9

34 ESTVK RA1511004010360 Anuroop Divakaran 37 38 40 48 20.38 111 134 134 144 26.15 42 21 67.53 16.9
Calcolator for the Visually
9 Ms.Srilekha
impared
35 ESTVK RA1511004010338 Jaya Kumar 37 38 40 48 20.38 111 134 134 144 26.15 42 21 67.53 16.9

36 ESTVK RA1511004010308 jhabar mal 37 38 40 48 20.38 111 134 134 144 26.15 42 21 67.53 16.9

37 EST-2 RA1511004010692 Visvender Singh 45 46 45 50 23.25 141 141 141 146 28.45 47 23.5 75.2 18.8

38 EST-2 RA1511004010684 Rishabh Verma 45 46 45 50 23.25 141 141 141 146 28.45 47 23.5 75.2 18.8
Mr. E.Siva
10 Multi Utility Drone
Kumar
39 EST-2 RA1511004010671 Ritesh Kumar 45 46 45 50 23.25 141 141 141 146 28.45 47 23.5 75.2 18.8
Pawan Kumar
40 EST-2 RA1511004010215 45 46 45 50 23.25 141 141 141 146 28.45 47 23.5 75.2 18.8
Saini(eduvirtuoso)

41 EST-2 RA1511004010456 Sakshi awal 42 42 42 48 21.75 128 137 137 133 26.75 42 21 69.5 17.4
Design and
Goty muhammad
42 EST -2 RA1511004010469 implementation of IOT 42 42 42 48 21.75 128 137 137 133 26.75 42 21 69.5 17.4
ishaq(cognizant) Mr. K.
11 based actutator and sensor
Shaikh mohammad Ramesh
43 EST-2 RA1511004010473 system for multistorage 42 42 42 48 21.75 128 137 137 133 26.75 42 21 69.5 17.4
Rahil building
44 EST-2 RA1511004010459 Shayan ghosh 42 42 42 48 21.75 128 137 137 133 26.75 42 21 69.5 17.4

45 EST-2 RA1511004010400 Sushobhan Misra 35 40 45 45 20.63 125 135 135 135 26.5 47 23.5 70.63 17.7
IOT for Interactive
Mr. K.
46 EST-2 12 RA1511004010475 Sangrila Joyesh Refrigerator with smart 35 40 45 45 20.63 125 135 135 135 26.5 47 23.5 70.63 17.7
Ramesh
response
47 EST-2 RA1511004010509 Ashish Deshwal 35 40 45 45 20.63 125 135 135 135 26.5 47 23.5 70.63 17.7
N.Prasanth
48 EST-2 RA1511004010071 42 44 46 49 22.63 128 137 137 150 27.6 50 25 75.23 18.8
Narayan(infosys)

49 EST-2 RA1511004010533 Piyush Gulati 42 44 46 49 22.63 128 137 137 150 27.6 50 25 75.23 18.8
Highly Secured Biometric Mr. S
13
Gollapudi Raj voting system Nivash
50 EST-2 RA1511004010022 42 44 46 49 22.63 128 137 137 150 27.6 50 25 75.23 18.8
Bharath
Chikkala Surya
51 EST-2 RA1511004010136 42 44 46 49 22.63 128 137 137 150 27.6 50 25 75.23 18.8
Pradeep
Kariyavula Mounish
52 EST-2 RA1511004010029 42 42 45 48 22.13 139 141 141 145 28.3 48 24 74.43 18.6
Naidu
Intelligent System for
53 EST-2 RA1511004010214 Kollipari Avinash 41 49 45 48 22.88 139 141 141 145 28.3 48 24 75.18 18.8
Human Computer
14 Mr.Sriram
Interface using Hand
Gesture Recognition
 Intelligent System for
Human Computer
14 Mr.Sriram
Mummadi Ravi Teja Interface using Hand
54 EST-2 RA1511004010236 40 40 45 48 21.63 139 141 141 145 28.3 48 24 73.93 18.5
Reddy(congnizant) Gesture Recognition
Chinthala Vineeth
55 EST-2 RA1511004010178 40 40 45 48 21.63 139 141 141 145 28.3 48 24 73.93 18.5
Reddy

56 EST-2 RA1511004010429 Abhishek Mukherjee 35 46 45 48 21.75 140 142 142 142 28.3 43 21.5 71.55 17.9

57 EST -2 RA1511004010379 Hetal Sahu Eye Blink Controlled 35 46 45 48 21.75 140 142 142 142 28.3 43 21.5 71.55 17.9
Mr.Bashya
15 virtual keyboard using
Sudhanshu m
58 EST-2 RA1511004010377 Raspberry Pi 35 46 45 48 21.75 140 142 142 142 28.3 43 21.5 71.55 17.9
Kanth(377)

59 EST-2 RA1511004010455 Surendra 35 46 45 48 21.75 140 142 142 142 28.3 43 21.5 71.55 17.9

60 EST-2 RA1511004010527 Nikhil Datta(HPE) 38 40 40 42 20 135 135 135 140 27.25 45 22.5 69.75 17.4
61 EST-2 RA1511004010623 Gyaneshwar Pal Agriculture Land Mrs. D. 38 40 40 42 20 135 135 135 140 27.25 45 22.5 69.75 17.4
16 Monitoring with Dry Leaf Vijayalaks
62 EST-2 RA1511004010569 Sidharth Thakur Detection mi 38 40 40 42 20 135 135 135 140 27.25 45 22.5 69.75 17.4
63 EST-2 RA1511004010613 Shravan.S 38 40 40 42 20 135 135 135 140 27.25 45 22.5 69.75 17.4

64 EST-2 RA1511004010658 Samriddhi Shanker 48 48 48 48 24 139 140 140 144 28.15 48 24 76.15 19

65 EST-2 RA1511004010668 Garima Sharma 48 48 48 48 24 139 140 140 144 28.15 48 24 76.15 19
Autonomous way point Mrs. Anilet
17
navigation with UGV Bala
66 EST-2 RA1511004010670 Aditya Kumar 48 48 48 48 24 139 140 140 144 28.15 48 24 76.15 19
Vipul
67 EST-2 RA1511004010688 48 48 48 48 24 139 140 140 144 28.15 48 24 76.15 19
Notani(Continental)
Prerana Roy(High
68 EST-2 RA1511004010211 38 38 42 48 20.75 127 135 135 141 26.9 40 20 67.65 16.9
radius)
Raveena
69 EST-2 RA1511004010383 Automated Detection of 38 38 42 48 20.75 127 135 135 141 26.9 40 20 67.65 16.9
Daawat(amazon) Mrs. E.
18 white blood cell cancer
Chitra
70 EST -2 RA1511004010189 Aswin kuamr s diseases updated in IOT 38 41 42 48 21.13 127 135 135 141 26.9 40 20 68.03 17
Kondagari Sai
71 EST-2 RA1511004010219 38 41 42 48 21.13 127 135 135 141 26.9 40 20 68.03 17
Nikhila

HOD/ECE Project Coordinator

 SRM Institute of Science and Technology
College of Engineering and Technology
Department of ECE

AY 2018-2019
15EC496L -Major Project Details ( CO & PO Mapping)
Review 2 Review 3
Review 1 (10)
(15) (20)
Sl No Register No Students Name(s) Project Supervisor Project Title CO1 & CO2 CO3 & CO4 CO5
PO1, PO4, PO2, PO3, PO8, PO10,
PO6, PO7 PO5, PO9 PO11, PO12
RA1511004010107 Pooja Anand
9.65 14.75 18.8
Low Cost Digitalization (Industry 4.0) Solution for Siemens Sinumerik
1 Dr. P. Eswaran
RA1511004010059 Vinitha Lea Philip CNC System to Increase the Transparency and Utilization of the Machine.
9.3 14.75 18.7

RA1511004010553 kedar prasad karpe 10 12.125 19.6

RA1511004010511 Dhruv pant 10 12.25 19.6
2 DR. R .Kumar cooperative transport using Multi-robot system
RA1511004010712 Nimish pastaria 10 12.25 19.6
RA1511004010654 jayati singh 10 11.5 19.6

Course Outcomes: Program Outcomes

CO 1: To provide learners with the opportunity to apply the knowledge and skills acquired in their
courses to a specific problem or issue. PO 1: Engineering knowledge:
CO 2: To allow learners to extend their academic experience into areas of personal interest, working PO 2: Problem analysis
with new ideas, issues, organizations, and individuals. PO 3: Design/development of solutions:
CO 3: To encourage learners to think critically and creatively about academic, professional, or social PO 4: Conduct investigations of complex problems
issues and to further develop their analytical and ethical leadership skills. PO 5: Modern tool usage
CO 4: To provide learners with the opportunity to refine research skills and demonstrate their PO 6: The engineer and society
proficiency in written & oral communication skill. PO 7: Environment and sustainability
CO 5: To take on the challenges of teamwork, prepare a presentation in a professional manner, and PO 8: Ethics
document all aspects of design work. PO 9: Individual and team work
PO 10: Communication
PO 11: Project management and finance
PO 12: Life-long learning

Coordinator
HOD/ECE
 PROJECT REPORT – 4
6. TLP 5 for Review 1, 2, 3
11/29/2019 Zoho Creator - TLP5 2018-19 EVEN Report

FACULTY
FACULTY OF
OF ENGINEERING
ENGINEERING AND
AND TECHNOLOGY
TECHNOLOGY
SRM Institute of Science and Technology, Kattankulathur
(ACADEMIC YEAR 2018 - 2019 - EVEN)

FORMAT TLP5

Test Name : Review I

Component Max. Mark: 10.00 Marks

15EC496L(Major Project) handled by Dr. C.T. Manimegalai(100224)

S.No. Reg. No Name Dept Obtained Mark %

1 RA1511004010215 Pawan Kumar Saini ECE 7.00 70.00
2 RA1511004010254 Syed Sharique Ahmad ECE 7.00 70.00
3 RA1511004010308 Jhabar Mal ECE 7.00 70.00
4 RA1511004010338 Jaya Kumar ECE 7.00 70.00
5 RA1511004010343 Chiranshi Baxi ECE 8.50 85.00
6 RA1511004010357 Sneha Dubey ECE 8.50 85.00
7 RA1511004010360 Anuroop Divakaran ECE 7.00 70.00
8 RA1511004010511 Dhruv Pant ECE 8.30 83.00
9 RA1511004010527 Nikhil Datta ECE 8.10 81.00
10 RA1511004010553 Kedar Prasad Karpe ECE 9.80 98.00
11 RA1511004010567 Anoushka Shukla ECE 8.40 84.00
12 RA1511004010569 Sidharth Kumar Thakur ECE 8.10 81.00
13 RA1511004010572 Ashmita Sarkar ECE 8.50 85.00
14 RA1511004010588 Anchal Sohal ECE 8.20 82.00
15 RA1511004010590 Chirag Bhatla ECE 8.20 82.00
16 RA1511004010593 Mohit Anand ECE 8.20 82.00
17 RA1511004010594 Akash Kumar ECE 8.20 82.00
18 RA1511004010610 Abhijeet Kumar ECE 8.50 85.00
19 RA1511004010613 Shravan.S ECE 6.60 66.00
20 RA1511004010623 Gyaneshwar Pal ECE 8.10 81.00
21 RA1511004010654 Jayati Singh ECE 8.30 83.00
22 RA1511004010658 Samriddhi Shanker ECE 10.00 100.00
23 RA1511004010668 Garima Sharma ECE 10.00 100.00
24 RA1511004010670 Aditya Kumar ECE 8.50 85.00
25 RA1511004010671 Ritesh Kumar ECE 8.20 82.00
26 RA1511004010672 Shubhendu Ojas Tewary ECE 9.40 94.00
27 RA1511004010683 Anjima S Nair ECE 9.40 94.00
28 RA1511004010684 Rishabh Verma ECE 7.00 70.00
29 RA1511004010686 Krishna Manoj ECE 9.90 99.00
30 RA1511004010688 Vipul Notani ECE 10.00 100.00
31 RA1511004010692 Visvender Singh ECE 9.70 97.00
32 RA1511004010696 D. Shruthi ECE 9.90 99.00
33 RA1511004010698 Bolugoddu Sai Dattathreya ECE 7.80 78.00
34 RA1511004010700 Rohan Sharma ECE 9.90 99.00
35 RA1511004010701 Ashish Sharma ECE 8.40 84.00
36 RA1511004010712 Nimish Pastaria ECE 9.80 98.00
37 RA1511004010752 Aishwarya A R ECE 9.40 94.00
38 RA1511004010774 Rayana Akhil ECE 8.30 83.00
39 RA1511004010788 Botta Venkata Akhil ECE 8.30 83.00
40 RA1511004010805 Manchikanti Tarun Krishna ECE 7.80 78.00

Total strength 40 Range of marks No.of students

Total absentees 0 0-49 0

https://creatorexport.zoho.com/exportPermaViewHeader.do?sharedBy=srm_university&appLinkName=academia-academic-services&viewLinkNa… 1/2
11/29/2019 Zoho Creator - TLP5 2018-19 EVEN Report
Total no. of failures 0 50-59 0
Pass MARK 50% 60-69 1
Pass percentage 100.00 70-79 8
80-89 19
90-100 12

SIGNATURE OF STAFF
SIGNATURE OF HOD
Report Date:29-Nov-19

https://creatorexport.zoho.com/exportPermaViewHeader.do?sharedBy=srm_university&appLinkName=academia-academic-services&viewLinkNa… 2/2
11/29/2019 Zoho Creator - TLP5 2018-19 EVEN Report

FACULTY
FACULTY OF
OF ENGINEERING
ENGINEERING AND
AND TECHNOLOGY
TECHNOLOGY
SRM Institute of Science and Technology, Kattankulathur
(ACADEMIC YEAR 2018 - 2019 - EVEN)

FORMAT TLP5

Test Name : Review II

Component Max. Mark: 15.00 Marks

15EC496L(Major Project) handled by Dr. C.T. Manimegalai(100224)

S.No. Reg. No Name Dept Obtained Mark %

1 RA1511004010215 Pawan Kumar Saini ECE 13.00 86.67
2 RA1511004010254 Syed Sharique Ahmad ECE 12.50 83.33
3 RA1511004010308 Jhabar Mal ECE 12.50 83.33
4 RA1511004010338 Jaya Kumar ECE 12.50 83.33
5 RA1511004010343 Chiranshi Baxi ECE 13.50 90.00
6 RA1511004010357 Sneha Dubey ECE 13.50 90.00
7 RA1511004010360 Anuroop Divakaran ECE 12.50 83.33
8 RA1511004010511 Dhruv Pant ECE 14.00 93.33
9 RA1511004010527 Nikhil Datta ECE 12.50 83.33
10 RA1511004010553 Kedar Prasad Karpe ECE 15.00 100.00
11 RA1511004010567 Anoushka Shukla ECE 15.00 100.00
12 RA1511004010569 Sidharth Kumar Thakur ECE 12.50 83.33
13 RA1511004010572 Ashmita Sarkar ECE 14.00 93.33
14 RA1511004010588 Anchal Sohal ECE 13.50 90.00
15 RA1511004010590 Chirag Bhatla ECE 15.00 100.00
16 RA1511004010593 Mohit Anand ECE 14.50 96.67
17 RA1511004010594 Akash Kumar ECE 15.00 100.00
18 RA1511004010610 Abhijeet Kumar ECE 14.00 93.33
19 RA1511004010613 Shravan.S ECE 13.00 86.67
20 RA1511004010623 Gyaneshwar Pal ECE 12.50 83.33
21 RA1511004010654 Jayati Singh ECE 14.00 93.33
22 RA1511004010658 Samriddhi Shanker ECE 14.50 96.67
23 RA1511004010668 Garima Sharma ECE 14.50 96.67
24 RA1511004010670 Aditya Kumar ECE 14.50 96.67
25 RA1511004010671 Ritesh Kumar ECE 13.00 86.67
26 RA1511004010672 Shubhendu Ojas Tewary ECE 15.00 100.00
27 RA1511004010683 Anjima S Nair ECE 13.50 90.00
28 RA1511004010684 Rishabh Verma ECE 13.00 86.67
29 RA1511004010686 Krishna Manoj ECE 13.50 90.00
30 RA1511004010688 Vipul Notani ECE 14.50 96.67
31 RA1511004010692 Visvender Singh ECE 13.00 86.67
32 RA1511004010696 D. Shruthi ECE 13.50 90.00
33 RA1511004010698 Bolugoddu Sai Dattathreya ECE 14.50 96.67
34 RA1511004010700 Rohan Sharma ECE 13.50 90.00
35 RA1511004010701 Ashish Sharma ECE 15.00 100.00
36 RA1511004010712 Nimish Pastaria ECE 14.00 93.33
37 RA1511004010752 Aishwarya A R ECE 15.00 100.00
38 RA1511004010774 Rayana Akhil ECE 14.50 96.67
39 RA1511004010788 Botta Venkata Akhil ECE 14.50 96.67
40 RA1511004010805 Manchikanti Tarun Krishna ECE 14.50 96.67

Total strength 40 Range of marks No.of students

Total absentees 0 0-49 0

https://creatorexport.zoho.com/exportPermaViewHeader.do?sharedBy=srm_university&appLinkName=academia-academic-services&viewLinkNa… 1/2
11/29/2019 Zoho Creator - TLP5 2018-19 EVEN Report
Total no. of failures 0 50-59 0
Pass MARK 50% 60-69 0
Pass percentage 100.00 70-79 0
80-89 12
90-100 28

SIGNATURE OF STAFF
SIGNATURE OF HOD
Report Date:29-Nov-19

https://creatorexport.zoho.com/exportPermaViewHeader.do?sharedBy=srm_university&appLinkName=academia-academic-services&viewLinkNa… 2/2
11/29/2019 Zoho Creator - TLP5 2018-19 EVEN Report

FACULTY
FACULTY OF
OF ENGINEERING
ENGINEERING AND
AND TECHNOLOGY
TECHNOLOGY
SRM Institute of Science and Technology, Kattankulathur
(ACADEMIC YEAR 2018 - 2019 - EVEN)

FORMAT TLP5

Test Name : Review III

Component Max. Mark: 20.00 Marks

15EC496L(Major Project) handled by Dr. C.T. Manimegalai(100224)

S.No. Reg. No Name Dept Obtained Mark %

1 RA1511004010215 Pawan Kumar Saini ECE 18.80 94.00
2 RA1511004010254 Syed Sharique Ahmad ECE 16.90 84.50
3 RA1511004010308 Jhabar Mal ECE 16.90 84.50
4 RA1511004010338 Jaya Kumar ECE 16.90 84.50
5 RA1511004010343 Chiranshi Baxi ECE 18.20 91.00
6 RA1511004010357 Sneha Dubey ECE 18.20 91.00
7 RA1511004010360 Anuroop Divakaran ECE 16.90 84.50
8 RA1511004010511 Dhruv Pant ECE 19.60 98.00
9 RA1511004010527 Nikhil Datta ECE 17.40 87.00
10 RA1511004010553 Kedar Prasad Karpe ECE 19.60 98.00
11 RA1511004010567 Anoushka Shukla ECE 19.30 96.50
12 RA1511004010569 Sidharth Kumar Thakur ECE 17.40 87.00
13 RA1511004010572 Ashmita Sarkar ECE 18.20 91.00
14 RA1511004010588 Anchal Sohal ECE 18.00 90.00
15 RA1511004010590 Chirag Bhatla ECE 18.00 90.00
16 RA1511004010593 Mohit Anand ECE 18.00 90.00
17 RA1511004010594 Akash Kumar ECE 18.00 90.00
18 RA1511004010610 Abhijeet Kumar ECE 18.20 91.00
19 RA1511004010613 Shravan.S ECE 17.40 87.00
20 RA1511004010623 Gyaneshwar Pal ECE 17.40 87.00
21 RA1511004010654 Jayati Singh ECE 19.60 98.00
22 RA1511004010658 Samriddhi Shanker ECE 19.00 95.00
23 RA1511004010668 Garima Sharma ECE 19.00 95.00
24 RA1511004010670 Aditya Kumar ECE 19.00 95.00
25 RA1511004010671 Ritesh Kumar ECE 18.80 94.00
26 RA1511004010672 Shubhendu Ojas Tewary ECE 19.30 96.50
27 RA1511004010683 Anjima S Nair ECE 18.50 92.50
28 RA1511004010684 Rishabh Verma ECE 18.80 94.00
29 RA1511004010686 Krishna Manoj ECE 18.50 92.50
30 RA1511004010688 Vipul Notani ECE 19.00 95.00
31 RA1511004010692 Visvender Singh ECE 18.80 94.00
32 RA1511004010696 D. Shruthi ECE 18.60 93.00
33 RA1511004010698 Bolugoddu Sai Dattathreya ECE 17.60 88.00
34 RA1511004010700 Rohan Sharma ECE 18.50 92.50
35 RA1511004010701 Ashish Sharma ECE 19.00 95.00
36 RA1511004010712 Nimish Pastaria ECE 19.60 98.00
37 RA1511004010752 Aishwarya A R ECE 19.30 96.50
38 RA1511004010774 Rayana Akhil ECE 17.60 88.00
39 RA1511004010788 Botta Venkata Akhil ECE 18.00 90.00
40 RA1511004010805 Manchikanti Tarun Krishna ECE 17.60 88.00

Total strength 40 Range of marks No.of students

Total absentees 0 0-49 0

https://creatorexport.zoho.com/exportPermaViewHeader.do?sharedBy=srm_university&appLinkName=academia-academic-services&viewLinkNa… 1/2
11/29/2019 Zoho Creator - TLP5 2018-19 EVEN Report
Total no. of failures 0 50-59 0
Pass MARK 50% 60-69 0
Pass percentage 100.00 70-79 0
80-89 11
90-100 29

SIGNATURE OF STAFF
SIGNATURE OF HOD
Report Date:29-Nov-19

https://creatorexport.zoho.com/exportPermaViewHeader.do?sharedBy=srm_university&appLinkName=academia-academic-services&viewLinkNa… 2/2
PROJECT REPORT – 4
7. Certificate by HoD