git@github.com:mehtashaunak/mehtashaunak.github.io.git
Hey There!

I'm Shaunak Mehta

I am a PhD student at the Department of Mechanical Engineering, Virginia Tech, USA. Presently, I am working in the domain of Human-Robot Interaction at Collaborative Robotics Lab (Collab) at Virginia Tech under the guidance of Dr. Dylan Losey
.


I completed my Bachelors in Mechanical Engineering with minor in Robotics and Mobility Systems at Indian Institute of Technology (IIT) Jodhpur in 2021. During my Bachelors, I was working under the supervision of Dr. Suril V. Shah
. I am grateful that I have had the opportunity to work with Dr. Atul Thakur
during my internship at Mechatronics, Instrumentation and Control (MICL) Lab at IIT Patna.


My research interests include Human-Robot Interaction and Perception based Manipulation and Control, and am forever looking for opportunities to expand my knowledge in these ever-evolving and fascinating fields.


Besides research, I love travelling, exploring new cultures and meeting new people.

RESUME
Projects

Projects

Completed Projects
Learning Latent Actions Without Human Demonstrations
August 2021 – September 2021 | Guide - Dr. Dylan Losey | Virginia Tech , USA

• Aimed to remove human from the loop for training a robot to perform assistive tasks for disabled people.
• Trained the robot in a simulated environment by defining a reward function to accomplish a certain task while maximizing the Shanon entropy.
• Combined the learned actions and transferred these complex actions to simple 2 degree of freedom joystick inputs for ease of completing daily tasks.
• Conducted a user study to find the ease of use of our method compared to industry standards and to understand what an inexperienced user prefers to use.

Vision Based Control and Manipulation of Half Humanoid Robot
Indian Space Research Organization (ISRO) | Respond Project | March 2021 - May 2021
Image Based Visual Servoing for Tumbling Objects
November 2020 – February 2021 | Guide - Dr. Suril V. Shah | IIT Jodhpur , India

• Extracted feature points from an uncooperative tumbling object to create an elliptical track in the image plane.
• A proportional controller was utilozed in order to minimize the error between the current elliptical track and the desired features.
• This algorithm was successfully implemented on a 6 DoF UR-5 robot to servo towards a tumblong object.

Student Mixture Model Based Visual Servoing
August 2019 – October 2020 | Guide - Dr. Suril Shah | IIT Jodhpur , India

• Developed a novel approach for visual servoing based on student t-distribution mixture model (SMM).
• Successfully achieved image modelling using SMM, analytically formulated interaction matrix and a SMM based framework for Image Based Visual Servoing (IBVS) without any requirement of feature tracking and correspondence.
• Systematic experimental validation was carried out using various test scenarios, test objects (shapes and sizes), object textures, obstacles etc. showing promising results.

Design and Development of Laparoscopic Simulator with 3 DoF Haptic Feedback
(Summer Research Internship) May 2019 – July 2019 | Guide - Dr. Atul Thakur | Indian Institute of Technology Patna, India

• This project aimed to develop of a novel kinesthetic haptic device for laparoscopic simulation providing a 3 DoF force feedback.
• Development was focussed on user interaction with virtual organs instead of cadavers for realistic experience, for training and improving the skills of surgeons in surgical process of laparoscopy.
• Achieved force feedback in x and y direction using a cable-driven parallel mechanism, and in z-direction using a ferrofluid based damper and a capstan drive mechanism.

Vision Based Control and Grasping with UR-5 Manipulator
January 2019 – April 2019 | Guide - Dr. Suril Shah | IIT Jodhpur , India

• Studied the Image Based Visual Servoing algorithm and operation of UR-5 Robotic Manipulator.
• Successfully controlled the position, velocity and acceleration of 6 Degree of Freedom UR-5 manipulator using ROS.
• Multiple feature detection and tracking was achieved with the implementation of visual servoing algorithm.

Dsign and Development of a Smart Rail Guided Vehicle (RGV)
(Summer Internship) May 2018 – July 2018 |Godrej Interio, Mumbai, India

• The focus of this pojects was to design and prototype an automated RGV for interplant material transfer in an industrial setting.
• A RGV was designed on Fusion 360 taking into onsideration the payload, slope, operating speeds, etc. and a detailed analysis was performed for the same.
• A LIDAR sensor was integrated to get a real-time feedback of the open environment to avoid the obstacles/personnel in its path for the purpose of avoiding hazards.

Bluetooth Integrated Automotive with Obstacle Avoidance
January 2018 – April 2018 | IIT Jodhpur , India

• Interfaced a battery operated vehicle with smart phone with the help of a HC-05 bluetooth module using an Arduino platform.
• Developed an application to connect to and control the car using the smart phone.
• A SONAR interfaced with an Arduino platform was used to scan the surrounding areas for obstacles and accordingly govern the car to safety surpassing the input given by the smartphone.

Publications

Publications

2020
[1] Mithun. P, Shaunak A. Mehta, Suril V. Shah, Gaurav Bhatnagar, K.Madhava Krishna. (2020). Student Mixture Model Based Visual Servoing. Preprint arXiv:2006.11347.
.
Contact Me

Contact

Contact Info
Phone

+1 (267) 466-7174

Social Media
LinkedIn

Connect

Google Scholar

Citations