Keivalya Pandya

CS5180/4180 Reinforcement Learning and Sequential Decision Making, taught by Prof Chris Amato and Prof Robert Platt. - Develop and maintain RL concept visualization platform (http://cs5180-rl.keivalya.com/) to simulate RL agents in controlled environment - Help students learn the concepts outside the classroom. - Encouraged asking questions, focusing on fundamentals, and learning from the first principle method. - Support hands-on coding from scratch in exercises, and real-world problem solving by employing RL algorithms.

Design, implement, and test learning algorithms to improve autonomous robotic behaviors in real-world environments. Collaborate closely with Al, Hardware, and Software teams to experiment with policy optimization, simulation-to-reality transfer, and performance benchmarking.

Deployed algorithms @ PARCS (People and Robot Collaborative Systems) Lab, that help navigate robots in challenging setting with humans and other robots. "If in a room, what is not only the safest but also the most socially appropriate/acceptable pose for the robot to join a group conversation?" For instance in a party setting, will the robot join a human conversation the same way as in an industry, or pavement? This research becomes more and more relevant as we see robots being deployed rapidly everywhere!

As a part of fully-funded CS7675 (Master's Research) pursuing Research Apprenticeship, guided by Prof Zhi Tan @ People and Robot Collaborative Systems (PARCS) Lab, to work on benchmarking robot learning algorithms (Diffusion Policy, VQ-BeT, OpenVLA, Deep Perception, etc) and figure out what are their limitations, understand why an algorithm performs better than the other and what environmental factors affect learning. Checkout: keivalya.com/cs7675 NEWS feature: https://www.khoury.northeastern.edu/25-in-25-record-number-of-khoury-research-apprentices-showcase-work-in-boston/#keivalya

CS5180 Reinforcement Learning and Sequential Decision Making, for Spring 2025 taught by Prof Robert Platt. - Help students learn the concepts outside the classroom. - Encouraged asking questions, focusing on fundamentals, and learning from the first principle method. - Support hands-on coding from scratch in exercises, and real-world problem solving by employing RL algorithms.

• Development of Autonomous Mobile Robot (AMR) using ROS2. • Mentor Robotics (ROS) and Machine Learning Interns in their internship. • Develop a unified software framework that connects Smart Manufacturing operations over OPC-UA Protocol, using Tkinter. • Designed and developed an E-commerce website for Center of Excellence in Digital Manufacturing Laboratory, using Django.

Roles and Responsibilities • Establish and maintain the Foundation for Smart Manufacturing (FSM) Laboratory to facilitate industry and academia collaboration. • Develop a practical research-oriented ecosystem within the institute for novel research outcomes. • Maintain local server data, utilize the resources & and funds, and oversee students’ projects.

My first fulltime job in India. Balancing my academics while teaching robotics to students across the USA, Canada, UAE etc.

Concluded breakthrough research oriented projects, reviewed (and to be published) in reputed International and National Journals for my work on "Digital Twin of Thermovac Chamber and Quantum Payload" and "Classification of Residential and Non-Residential Buildings from Satellite Images using AI" The exposure at ISRO was much more than just project centric. It was my first connection with the corporate world, how groundbreaking research takes place, why do we need to work in the way that we do, and so much more! I'm forever grateful for this opportunity to intern, while I remember literally working from 6 AM to 8 PM some days, single focused over the project consistently for weeks! I am thankful for the freedom to fail. The organisation gave me freedom to explore multi-modal and linguistic approaches of ML/DL. I learnt so much more than I came here for. Being recognised by the Director sir, Heads or various departments, and several other scientists, made me realise the importance of my work. To all the colleagues I met here, all of them full of ambition and passion towards science, they inspired me to push my limits. I am glad, I chose ISRO as my first industry exposure, final year's internship, and an amazing learning companion.

I have used YOLOv5 in order to detect and classify defects in any given image. The model has been trained over 1500 annotated defective PCB dataset. Project Demos: https://youtu.be/G82gywRR4o8 https://youtu.be/zP__x9c0EuQ

50 days of intensive learning & effectuation. This has been a dream for me! Being able to create a beautiful amalgamation of Mechanical Manufacturing and Maintainance & implementing Machine Learning. I worked on making #DigitalTwin of a conventional Lathe Machine. The journey has been full of ups and downs, but the mentorship provided by the company, made hard work, worth the effort! The concept was entirely new to me. I attended the orientation and got fascinated by that ONE topic, I had no clue whatsoever.

Robotics with concentration in Computer Science. Coursers: - CS7180 Mechanistic Interpretability - CS7150 Deep Learning - CS7675 Master's Research - CS5180 Reinforcement Learning and Sequential Decision Making - ME5250 Robot Mechanics and Control - EECE5550 Mobile Robotics

With the motto of "Work is worship," BVM gave me exposure to the field of Robotics, which grew my interest in Autonomous Robots, then AI and ML domain. Later, I could link my multi-disciplinary understanding of these subjects to building several state and national-level award-winning projects.

Qubit by Qubit's Introduction to Quantum Computing course sponsored by IBM Quantum. The course provides an insight into the basics of quantum computing, learning about many subtopics including Grover’s algorithm, quantum mechanics, quantum key distribution, quantum chemistry, and programming quantum circuits using Qiskit. It consisted of weekly lectures and lab sessions over Zoom and weekly homework assignments.

- Attended Summer University 2023 program by Ural Federal University, fully Government funded. - International Winner of BEST PROJECT AWARD over-all the other projects across multiple domains. - Developed AI text-to-furniture project — Altern-O-Vation, for Alterna. - Indian representative for the program. Made friends for life, from around the globe.