Team AGV is a research group based in IIT Kharagpur. Our vision is to develop the first self-driving car of India, made for Indian conditions.I was a part of the Control Systems & Mechanical Team. My work revolved around designing and programming control systems for autonomous vehicles, designing and fabrication of mechanical chassis and sensor mounts and developing the electrical architecture for many robots and cars in our lab.We are also building a self-driving car for Mahindra Rise Prize Driverless Car Challenge, where our team was selected among top 13 teams in India.

• Developed time-optimal algorithms for spatial sampling problem for the forest fire boundary search problem as Markov Decision process• Implemented the Stochastic Shortest Path search and Monte Carlo Tree Search algorithm for the MDP considering a uniform prior, and a non-uniform prior on the localized change point of a one-dimensional threshold classifier for optimizing the time of travel and sampling• Worked with the autonomous sensing drone DJI Matrice 100 with the responsibility of deploying the sampling algorithms onboard along with existing algorithms like Q-learning using Python and C++ programming language using ROS interface on Nvidia TX2 board

• Implemented a novel Kino-dynamic motion planning strategy based on the Hybrid A* algorithm along with successful testing of two other algorithms: - 1. TP-RRT [Rapidly exploring random trees] 2. TEB [Timed elastic band] on the golf cart using odometry sensors and 3D LIDAR• Worked and implemented the high level and low level controller for the velocity and steering angle measurements by establishing a two-way communication between them using ROS and microcontroller Arduino• Contributed in experiments for stochastic resonance stimulation and vibration on muscle contraction and study of kinematic analysis of gait motion of knee range motion and ankle joint motion using Motion Capture System and MATLAB