Projects

!! Project addition still in progress !!

eYSIP Projects

• 2026

  • DADSI Bot: Data Driven System Identification (2 interns)

    • fitted pitch and encoder models for a balancing bot

    • Extracted robot dynamics from input–output data via discrete transfer function fitting, validating separate pitch and encoder models.

    • Enabled controller design robust to noise and parameter drift without analytical modelling

  • VLA (Vision Language Action) grasp (1 intern)

  • AMCAF: Ackermann-Based Robot Collision Avoidance Framework (3 Interns)

  • Autonomous Navigation & on field scheduled task execution on 4WD (3 interns)

• 2025

  • Remote access of turtlebot (1 intern)

  • Industrial arm automation (2 interns)

    • low-latency remote UR5 control with a multi-layer safety framework.

    • Built low-latency remote UR5 control with Cartesian servoing, RRT planning and TCP force feedback.

    • Integrated a safety framework covering singularity, self-collision and boundary checks, validated on hardware.

  • Origami in robotics (3 interns)

• 2024

  • ROS-based SLAM implementation and hardware optimization for small skid-steer bot (4 interns)

  • Hardware automation for remote hardware (3 interns)

  • Software automation for remote hardware ( 2 interns)

  • Multimodal Generative AI for robotics (4 interns)

  • Uniwheel balancing Robot (2 interns)

• 2023

  • Air-Ground Robotic Collaboration (2 interns)

  • An autonomous outdoor navigation system stack for rover (2 interns)

  • Exploring algorithms for grasping unknown objects using two-finger gripper (1 intern)

  • 3D reconstruction using deep learning (2 interns)

eYRC

• Krishi Bot (2022-23): trained 373 teams

  • Tech stack: ROS, Gazebo, Control, Image Processing, Moveit, python

  • greenhouse cobot with vision-based artificial pepper detection, navigation and collaborative arm manipulation

• Cosmo Logistic (2023-24): trained 350 teams

  • Tech stack: ROS2, Gazebo, Nav2, SLAM, Moveit2, Image Processing, python

  • MoveIt2 manipulation pipeline for autonomous package sorting with remote hardware access; 350 teams trained.

• Logistic coBot (2024-25): trained 237 teams

  • Tech stack: ROS2, Gazebo, Nav2, Image processing, Nav2, Manipulation kinematics, python

  • navigation plus manipulation kinematics stack; 237 teams trained.

• Niti Vahan (2026-27) [Theme Lead of 2 member team]: Ongoing

  • Tech Stack: Python, Coppeliasim, Image Processing, 3D design

  • scaled-down basic self-driving car doing lane-keeping, lane-change and parking from image processing cues (ongoing).

Planning and Control Framework for Laneless Driving: Building an HIL Pipeline                                                 

M.Tech Thesis, IIT Bombay; Guide: Asst. Prof. Parthib Khound

•Evaluating global planner and control framework combinations for unstructured laneless traffic on a 1:10 scaleF1TENTH vehicle.

•Layering a Control Barrier Function safety filter over the planner, solving a CBF-QP each step to keep the vehicle inside the collision-free set under actuation limits.

•Building a hardware-in-the-loop pipeline that validates controllers against physical parameters, with pose from camera and OptiTrack localisation.

Energy Efficient and Low Cost Retrofit Smart Analog Water Meter

Aug 2021 – Mar 2022

B.E. Final-Year Project, SPCRC Lab, IIIT Hyderabad

•Designed CAD retrofit models that mount onto existing analog meters without replacing installed hardware

•Read the dial via on-device edge image processing and adaptive sampling, trading capture rate against solar energy budget.

Learning-Based Locomotion in a Snake Robot 

Aug 2025 – Nov 2025 

M.Tech Seminar, IIT Bombay; Guide: Prof. Debraj Chakraborty (Mumbai, India) 

•Worked on multi-joint snake robot with alternate-axis joint motion in Gazebo simulator. 

•Learned locomotion via gradient ascent and reinforcement learning: value iteration and Q-learning techniques.