Tejas Rao | Roboticist

Hi, I'm Tejas Rao.

A

Self-driven, quick starter, passionate roboticist with a curious mind who enjoys solving a complex and challenging real-world problems.

About

I'm an alumnus of IIT Madras, with a bachelor's degree in mechanical engineering. I'm passionate about programming, robotics and machine learning. I'm have been working at Stochlab, Indian Institue of Science for the past year. My work there has primarily focussed on implementing control frameworks for robots. Through my project work, I have gained experience in Python, ROS and Computer Vision frameworks. I love to solve problems and develop solutions that can make a tangible impact.

Languages: Python, C++, MATLAB, HTML, Bash
Databases: MySQL
Libraries: NumPy, Pandas, OpenCV
Frameworks: ROS, PyTorch
Tools & Technologies: Git, Docker

Looking for opportunities that would challenge me and allow me to use my skills to make meaningful contributions.

Experience

Stochlab, IISc

Project Associate

Developed a perception stack for obstacle pose and velocity estimation using depth cameras.
Implemented a non-linear controller (Collision-Cone Control Barrier Functions) on bicycle and unicycle model robots for obstacle avoidance.
Performed hardware testing of the C3BF controller using motion capture frameworks.
Submitted results to the Transactions on Automatic Control (TAC).
Implemented an orientation aware model predictive controller (OA-MPC) for a bipedal robot.
Tools: Python, OpenCV, Pytorch, ROS

(view project section for details)

June 2023 - May 2024 | Bangalore, India

Jaguar Land Rover

Mechatronics Intern

Performed Automotive Benchmarking on Panoramic Roof Structure using A2MAC1 software.
Analyzed frontal crash kinematics of competitor vehicles (SUV) with A2MAC1.
Placed 1 st in the JLR Hackathon for the Conditioning Monitoring of sensors.
Tools: A2MAC1, Excel, MATLAB

May 2022 - July 2022 | Bangalore, India

Eigen Dynamix

Software Intern

Worked on remote transmission of sensor data using MQTT Bridges and Websockets.
Remotely operated and transmitted sensor data from a catvehicle model.
Tools: Python, ROS

June 2021 - July 2021 | Remote, India

Projects

                
Collision-Cone Control Barrier Functions
                  An obstacle avoidance controller for wheeled mobile robots.
                
AccomplishmentsCode base for an obstacle avoidance controller for wheeled mobile robots.
Package can be used for both gazebo simulations as well as physical robots.
Tools: Python, ROS

Mobile Robot Localization
                  B.Tech Project - IIT Madras
                
AccomplishmentsCode base for a localization algorithm for indoor mobile robots.
Uses an EKF to fuse 2-D lidar and odometry data. 
Implements a seeded-growing line extraction algorthm to detect walls.
Tools: Python, ROS, OpenCV, Pytorch

Motion Planning for Surgical Robot
                  Implementing path planning algorithms for a Surgical Robot with RCM constraint.
                
AccomplishmentsCode base for simulating and Kuka iiwa surgical arm.
Implements several variants of the RRT algorithm for path planning with an RCM constraint.
Tools: ROS, Gazebo, Moveit!, Python

Velocity Estimation using Depth Cams
                  Framework to estimate the pose and velocity of objects using RGB-D images and deep learning.
                
AccomplishmentsTrained a yolov5 model on a custom dataset for bounding box detection.
Obtained object pose using the bounding box coordinates and depth information.
Estimated velocity by computing differences between successive frames.
Tools: OpenCV, Pytorch, Python, matplotlib

Facial Attractiveness Classifier
                  A CNN model that rates the Attractiveness of a person.
                
AccomplishmentsTrained a CNN model on the SCUT-fbp5500 dataset to rate facial attractiveness.
Preprocessed the dataset by cropping out the face using pytorch-facenet.
Tools: OpenCV, Pytorch, Python, matplotlib, scikit-learn