Building at the intersection of embedded hardware, real-time control, and graphical simulation. From servo registers to OpenGL renderers — end-to-end, from scratch.
Electronics & Communication Engineering undergraduate at Arya College of Engineering & IT, focused on robotics, embedded systems, and real-time sensing. I build robotics prototypes that integrate physical hardware with software — kinematics, motion control, computer vision, and graphical simulation working together in one system.
My interest is in the layer where hardware meets intelligence: writing firmware that talks directly to actuator registers, implementing state estimators from first principles, and rendering live simulations that mirror physical robot state.
Currently extending: 2D SLAM toward least-squares optimization and real sensor integration · Exploring Nav2 and deeper ROS 2 middleware.
| Domain | Tools & Skills |
|---|---|
| Programming | C, C++, Python (NumPy, OpenCV) |
| Robotics | Forward & Inverse Kinematics, Motion Control, Swarm Robotics, SLAM, EKF-based State Estimation, Graph SLAM |
| Embedded Systems | ESP32, Arduino, Half-duplex UART, Register-level Serial Control, RF Mesh Networking, PCA9685, nRF24L01, GSM, RTC |
| Computer Vision | OpenCV, YOLOv8, Kinect Depth Sensing, Pose Estimation |
| Graphics & Simulation | OpenGL (GLFW, GLAD, GLM), Real-time 3D Rendering |
| Robotics Middleware | ROS 2, Perception Pipelines, cv_bridge |
| Tools | Blender, Fritzing, Arduino IDE, Git |
|
Real-Time IK/FK Visualization + Physical Control Mouse-driven OpenGL interface that computes FK/IK in real-time to translate cursor position to joint angles — then sends those angles directly to a physical Dynamixel AX-12A servo arm. Built manual half-duplex TTL direction-switching circuit on ESP32 without a dedicated shield.
|
EKF + Graph SLAM from Scratch Full 2D SLAM implementation built from first principles: EKF with simulated noisy odometry and landmark measurements, Graph SLAM with factor graph structure for pose and landmark nodes, animated trajectory visualization. Extending toward least-squares optimization and real sensor integration.
|
|
Autonomous Perception + Motor Control Robot that detects and follows humans using YOLOv8 pose estimation and Kinect depth sensing. Real-time perception pipeline built in ROS 2 with OpenCV tracking and Arduino motor control. Supports both autonomous AI mode and voice command operation.
|
Obstacle-Aware Terrain Robot · Patent Filed Obstacle-aware surveillance robot with fused camera and ultrasonic sensing. Terrain-capable chassis designed for rocky environment mobility. Servo scanning for spatial assessment. Patent application filed (in progress).
|
|
Wireless Power Grid Monitoring · SIH Prototype Wireless system detecting LV AC overhead conductor breakage for power distribution monitoring. Passive inductive sensor using coil + transistor daisy chain for contactless current detection. Mesh + node RF network for distributed real-time monitoring. Working prototype demonstrated at SIH.
|
RF Mesh · IK Locomotion · Autonomous + RC Swarm robot platform where units communicate via RF mesh topology. Inverse kinematics implemented for walking stability and coordinated locomotion. Supports both autonomous behavior and remote-controlled operation.
|
Other Projects: AIER Meter — Energy Monitoring & Fault Detection (nRF24L01 · GSM · RTC) · OA-CAT / OA-KITTO — Autonomous Obstacle Avoidance Robots (3D Printed · Ultrasonic · Servo IK)
Embedded Hardware ████████████████████ Register-level control, half-duplex UART, RF mesh
Kinematics & Control ████████████████░░░░ FK/IK, EKF, SLAM, motion planning
Computer Vision ███████████████░░░░░ OpenCV, YOLOv8, Kinect depth sensing
Graphics & Simulation ████████████░░░░░░░░ OpenGL real-time 3D, trajectory animation
ROS 2 & Middleware ██████████░░░░░░░░░░ Perception pipelines, ROS 2 (growing)
"The most interesting problems live at the boundary between hardware and software — where a register write moves a physical joint, and a sensor reading reshapes a belief state. I build in that space: precise, low-level, and deliberate."