Hello,
I am Ruthwik...
I thrive on tackling challenges and dream of a life filled with constant innovation, inventive pursuits, and the creation of inspiring wonders. Armed with skills in robotics and mechanical design, I am not just an eager learner but also a curious explorer, ready to dive into the vast sea of knowledge. This enthusiasm shapes my journey, leading me through a rich tapestry of projects and learning experiences. Currently pursuing a Master's in Robotics at the University of Maryland, I am on a mission to explore the captivating realms of robotics. This portfolio is my canvas, showcasing the brushstrokes of my previous adventures.
LATEST PROJECTS
Graduate Assistant
Prof. Nikhil Chopra, Maryland Robotics Center @ UMD
Working on modeling and control of a hybrid rigid-soft robot, comprising both rigid and soft links serially attached.
Developing a dynamic model for the rigid-soft manipulation system with model-based controllers.
Building a prototype with soft and hard links to perform manipulation, test its performance, and improve.
Image : Source: https://arxiv.org/abs/2402.16114
Feb 2024 Course Project: Robot Learning
Teaching a Humanoid Robot to walk.
Worked with OpenAI Gymnasium and MuJoCo Libraries and implemented SAC and PPO Algorithms to Teach a Humanoid Robot to Walk.
Compared Results from Gymnasium and Brax libraries. Performed Gait with a 30 DOF Humanoid Model
Feb 2024 Course Project: Planning for Autonomous Robots
Designed a novel Path Planning Algorithm - RU-RRT*
An Algorithm developed to use nodes and paths from previously computed trees, on the same Map. This method is innovated to reduce computation time and resources, providing an optimized path with lower costs. This is proven to be effective and best-suitable when repeatedly operating on larger maps.
Autonomous Turtlebot Path Planning and Navigation Competition - Spring 2024
š Won 2nd place among 100+ Robotics Masters Students by finishing the task in 33 Seconds
Performed path planning using A star algorithm, and implemented a proportional controller to navigate through the path using Turtlebot3 Waffle
Skills: Ā
ROS2, Gazebo, TurtleBot Configuring, PathPlanning algorithms Implementation, PID Control System
Feb 2024 Course Project: Perception for Autonomous Robots
Piloting Trajectory of an object from a video
Using Hough Transformation and Harris corner detection worked on detecting edges and corners of objects in a video
Worked on Stitching Images by matching features and Wrapping them. Used Brute force search, RANSAC for feature matching, Homography Techniques
Final Project Task
Implemented Autonomous perception and Navigation in real-time
Turtle Bot Waffle
Horizon line Detection
Paper Detection and Autonomous Navigation
STOP sign detection and stopping the vehicle
Detecting dynamic obstacles and acting to stop or move based on the position relative to the horizon line
Successfully executed all the tasks during the final competition and presented multiple approaches for each task during the presentation.
MY PUBLICATIONS
RESEARCH PAPER - OCEANS 2022 Conference
Co-authored a publication featured at the OCEANS 2022 Conference, hosted by the IEEE Oceanic Engineering Society Publication details the design and fabrication of a mini-robot inspired by manta rays, created for the specific task of swimming into a pipe, capturing a net, and returning with it. This project is showcased in the MATE ROV 2022 Competition. Created 3D CAD models of robot, conducted 3D printing and Silicone Casting process, and conducted simulations using ANSYS Fluid. These efforts contributed to the robot successfully executing its designated task at the competition.
ANSYS Fluid Analysis of Fin UndulationĀ
2 Finger Gripper Mechanism
Robot Fin Movement Mechanism Design
Manta-Ray Softrobot - PLA Casting
PATENT PUBLICATION
Published a patent funded by IIIT Kancheepuram in collaboration with Prof. Jayachandra Bingi for final year thesis. Developed vibration sensing technology inspired by concepts from bio-inspiration and photonics. Modeled and 3D printed the parts, creating multiple prototypes tested to improve performance and reduce cost. Swarms of these devices are strategically buried in patterns across a designated area, comprehensively monitoring by responding to vibrations and accurately identifying the source and direction based on signal patterns received from the devices.