VB

Vansh Bhardwaj

B.Tech ECE student · Robotics and Embedded Systems Enthusiast

Delhi, India

@vansh_bhardwaj

0 followers

B.Tech in Electronics & Communication Engineering (VLSI) at Maharaja Agrasen Institute of Technology · Graduating 2028

About

B.Tech student in Electronics & Communication Engineering with hands-on experience in robotics, embedded systems, reinforcement learning, and computer vision. Built autonomous robots, manipulation systems, and fault-tolerant control software using ROS2, STM32, and PyTorch.

Experience

  • Core Member – Robotics Development Team
    Atom Robotics Lab
    Oct 2025 – Present

    Worked on autonomous mobile robot development, Gazebo digital twin simulation, ROS2 Nav2 and SLAM-based navigation, and learning-based robotic manipulation with MuJoCo, LeRobot, and ACT policy training.

  • Intern – Fault-Tolerant SDR Systems
    SAG – DRDO
    Jun 2026 – Jul 2026

    Built a fault-tolerant SDR control framework with warm worker pool supervision, Unix domain sockets, shared-memory checkpointing, and rapid failover/state restoration mechanisms.

Education

Skills

1 to 2 years 1 to 2 years A year or two of steady use.
<3 months <3 months Just getting started - under three months. This is also what shows when a level has not been set.

Tools / apps / platforms

1 to 2 years 1 to 2 years A year or two of steady use.
Nav2 ROS2 Gazebo
<3 months <3 months Just getting started - under three months. This is also what shows when a level has not been set.
Act! KiCad Linux Docker Raspberry Pi

Projects

  • Real-Time Person Following Robot with Tracking & Trajectory Prediction
    YOLOv8, Raspberry Pi Zero 2W, UDP, ESP32, ROS2, Gazebo

    Built a distributed person-following robot with detection, tracking, trajectory prediction, and PID-based visual servoing, validated in ROS2/Gazebo and deployed on hardware.

  • CNN-Based Double DQN for Procedurally Generated GridWorld Navigation
    PyTorch, CNN, Double DQN, BFS

    Built a 6×6 GridWorld and trained a CNN-based Double DQN agent with solvability validation, reward shaping, experience replay, and target networks.

  • STM32 Micromouse – Autonomous Maze-Solving Robot
    STM32, IMU, Bluetooth, HAL, MPU6500

    Built an autonomous micromouse around STM32 with encoder and IMU-based localization, Flood Fill maze navigation, custom drivers, interrupt-driven processing, and Bluetooth telemetry.

🏆 Achievements & awards

  • Smart India Hackathon 2025 – Hardware Edition Finalist · 2025

Explore the Broxer community

Online · instant AI help
Broxer