I love to solve problems on a computer. Be it in any field-from general Operating system's errors to a specialized domain in Robotics. I found that my passion lies in Autonomous Systems. On this journey, I have acquired skills pertaining to machine learning, embedded hardware, robotics tools, programming languages, and optimization softwares.

My work entails programming in C++, Python and Matlab/Simulink and extensive use of ROS/ROS2 framework.

My field of Interest lies in providing solutions to problems in Perception, Localization, Planning, and Control of Dynamical Systems.

I am working as a Graduate Service Assistant at BELIV lab at Arizona State University. I am currently spearheading the Digital Twin and Virtual Reality Project here.

I am implementing my own software stack for autonomous vehicle based on ROS2 and inspired by Autoware.

I am Implementing a Deep Reinforcement Learning Autonomous Driving agent that takes help from human demostrations to learn.

I used stable-baselines3 to train an agent to follow a trajectory autonomously using RGB/Semantic sensor as input.

I trained a CNN using transfer learning then deployed it on Arduino Nano BLE and Beaglebone Black.

An implementation of navigation algorithm using two approaches- Map-less and Map to move a mobile robot to a destination.

An implementation of Linear Matrix Inequality based on Model Predictive Control for an autonomous vehicle to follow a trajectory.

A Digital Twin of the BELIV lab's autonomous vehicle and virtual reality in the sense that the software stack on real vehicle can respond to the virtual objects in the simulator.

I created a pick and place environment in Gazebo harmonic with ROS 2. I also created a Gazebo plugin to attach objects to the gripper on contact.

Scripts to get groundtruth CARLA town map information (drivable area, sidewalks, crosswalks, etc)

I created a bridge to run simulations with Carla and Autoware Universe.

I ran simulations for crash tests for class 8 trucks according to NHTSA's crash data. Then created ADAS algorithms so that the trucks can avoid collisions.