School Papers

Life on a lifelong quest for the ultimate

Life is a complex case of the shortest path problem. Each
one of us is on a lifelong quest for the ultimate algorithm to reach the end
destination. The freshman year in college marked the source node of my
endeavours in engineering. After being accepted into the student library
association, some of my most delightful days that year were spent tending to
textbooks and STEM journals. Flipping through the June 2014 issue of the MIT
Technology Review, an account on Qualcomm’s latest creation caught my attention. Armed with the revolutionary ‘Zeroth’
neuromorphic chip, the robot ‘Pioneer’ was a
machine near capable of cognition. Spellbound by the profundity of this
single most extraordinary idea, I dug deeper to comprehend how the seemingly
ordinary pieces of silicon could make sense of nearly every existing physical
phenomenon, similar to the human brain. Suddenly, R2-D2 seemed attainable.


My thirst for in-depth knowledge evolved into an obsession
towards finding solutions, which ultimately transformed into my life goal. What
got me up each morning was the desire to bring intelligent machines out from
fiction and into the real world. As a way to augment human life experience, I
aspired to study how the self-learning hardware was engineered. Building an
Artificial Intelligence processor R enterprise became my ambition. Since
it demanded an extensive knowledge base, I exposed myself to various facets of
engineering in my
undergraduate study. Investigating a myriad of technologies through
numerous projects cultivated my research interests and proved to be a window to
understand the theory in the taught curriculum.


Immersed in experimentation, I spent my sophomore and junior
years to thoroughly grasp something utterly intriguing. How was a piece of code breathing life into hardware? To understand
this, I started with the basics. Providing evening tuition to freshman students
to pay for expenses made me self-sufficient and fortified my conceptual
foundations. Most afternoons, I worked in the on-campus research facilities,
investigating both the high & low level design of embedded
hardware-software systems.


Self-help platforms like Coursera
trained me over the concepts left untouched in the classroom. As a capstone
project for the Internet of Things specialization MOOC by the University of
California Irvine, I constructed an ‘Android Controlled Self-balancing Robot’.
Utilizing an extended Kalman filtering algorithm and an aluminum clamp, the
robot was able to climb a 20-degree incline, to gracefully deliver a table-flag
memento to the guest of honor at the Independence Day celebrations in the
college auditorium.


In the e-Yantra Robotics Competition (eYRC) 2015 conducted
by IIT Bombay, I led my team to reach the finals, against 52 other qualifying
groups from across India. We designed an ‘Autonomous Puzzle Solver Robot’ on
Firebird V, a custom-built Atmega 2560 microcontroller based kit. I performed
PID tuning for precise movement of the robotic arm and implemented the
K-Nearest Neighbour algorithm for dynamic motion planning of our robot in the
partially observable environment. Traversing 45 nodes in under 4 minutes, the
robot decoded an arithmetic puzzle in record time.


My inherent belief of contributing to the society compelled
me to volunteer at an NGO which undertook the growth and monitoring of
education for underprivileged children. Here I built a ‘Biometric Attendance
Management System’ for a survey of actual classroom presence of middle school
students. Sparse sampling the fingerprint data and handling it through Amazon
Web Services, I discovered how scalable processing, storage and retrieval of
information was achieved.


To the end of my junior year, I got selected for the highly
coveted Research Internship program at the National Institute of Technology
Allahabad. Opting to work under their research expedition on H.265 Ultra-HD
technology, I investigated information processing further and was introduced to
the beauty of chip design. I developed a high-throughput architecture for the
Sample Adaptive Offset Filter which was integrated into a novel GPU microchip
under development at the institute. During the internship, I closely bonded
with the 2 fellow interns that I was associated with. Coming from an entirely
different ethnic background, they had no knowledge of my native language of
Hindi. An enlightening experience; collaborating with individuals from
different cultures is something I look forward to in the future as well.


Recent innovations by the Indian Space Research Organization
in dehazing methods for satellite imagery and the insights gained from my
research throughout the internship encouraged me to take up an image processing
based major project in my senior year. To build upon the developments by ISRO,
I plan on boosting and automating the image feature extraction in present-day
algorithms by employing neural networks. Designing the coursework and
conducting training workshops on autonomous robotics under the departmental
association ‘Quanta’, the expertise that I gained in machine learning will be
pivotal in accomplishing this project.


Throug­hout my years of overwhelming empirical study, I have
successfully maintained a first-class academic record. Exhausting every
accessible resource at my undergraduate institute, I have widened my breadth of
knowledge and directed myself towards the right track. To delve deep into my
specific interests, I have to tread depth-first on the path towards my aim.
Specializing in Computer Engineering from University of California – Los Angeles is the next
most-promising node towards reaching it.


Encompassing VLSI, Information Processing, and Machine
learning; the program here has a pioneering cross-disciplinary approach to
teaching and research, which is an accurate match for my experience, interests
and objectives. Under the guidance and assistantship of Professor Jason Cong, I wish to explore the design and
development of cognitive computing platforms. Utilizing the resources present
at the pioneering VAST
laboratory, Furthermore, I aim at
being inducted into the quest on Customizable Domain-Specific Computing, so as
to serve the ever-expanding needs for energy-efficient computing technology and
advance the Beyond-CMOS research at UCLA as well. Creating a novel neural processing
architecture and fabricating it faultlessly at the C-DEN,
will culminate my graduate study. On completion of the course, I intend on quenching my curiosity
through industrial experience and
simultaneously work towards realizing my dream.


Resolute yet adaptable, creative yet meticulous, I
complement the academic community at UCLA. Favourable consideration of my
application will serve to prove that I have employed an effective heuristic
towards attaining my destination. An academic scholarship will strengthen my
self-esteem and enhance my self-sufficiency. Joining the university shall be
the next milestone in my journey as a person of science. Thank you.