Pre-Final Year · BTech Robotics & Automation

Chaitanya
Belekar

|
FK_SOLVE · LIVE
SCROLL
01 · About

Building systems that
move intelligently.

I got into robotics through a simple question: how do you make a machine understand its environment and act on it? That question has driven everything since — from writing my first forward kinematics solver to integrating vision pipelines with physical hardware.

My focus is on the software layer of robotics — perception, motion, and the systems that tie them together. I work in ROS2, Python, and C++, and I care about writing code that's clean, modular, and actually runs on hardware.

Right now I'm working toward becoming an industry-ready robotics software engineer — someone who can take a problem from specification to a deployed, reliable system.

Robotics Software
ROS2 ecosystems, URDF modeling, motion planning, real-time control loops.
Perception & Vision
Camera pipelines, object detection, ML integration for autonomous decision-making.
Simulation
Gazebo environments, kinematics validation, testing before hardware deployment.
Systems Thinking
Designing modular, maintainable architectures that scale from prototype to production.
02 · Skills
Robotics
ROS2KinematicsGazeboURDFMotion Planning
Programming
PythonC++OOPData StructuresAlgorithms
Perception & ML
OpenCVModel TrainingModel EvaluationCNNFeature Extraction
Tools
GitLinuxStreamlitVS CodeCMake
03 · Projects

Real-Time Face Recognition Attendance System

ML
Problem

Manual attendance systems are prone to proxy fraud and administrative overhead.

What I Built

A custom-trained face recognition pipeline that identifies individuals in real-time from a webcam feed and logs attendance automatically, eliminating proxy entries.

PythonOpenCVface_recognitionStreamlitSQLite

Achieves >95% recognition accuracy in varied lighting. Fully automated — zero manual input post-setup.

View on GitHub

ArmSim – Robot Arm Kinematics Playground

Simulation
Problem

Understanding forward kinematics is abstract without real-time visual feedback.

What I Built

A GUI-based 2D/3D robot arm simulator that lets users manipulate joint angles interactively, with live end-effector position updates derived from DH parameter computations.

PythonPyQt5NumPyMatplotlibSympy

Educational tool used to validate custom kinematics code before hardware deployment. Cuts debug time significantly.

View on GitHub

Vision-Based Color Sorting Robotic Arm

Robotics
Problem

Industrial sorting tasks require precise object discrimination and coordinated arm control.

What I Built

A perception-to-action pipeline: OpenCV detects and classifies object colors in a live feed, computes pick coordinates, and sends joint commands to a servo-driven physical arm.

PythonOpenCVArduinoPySerialNumPy

Demonstrated closed-loop vision-to-hardware control on real hardware. Sorting accuracy: ~92% under controlled lighting.

View on GitHub

OpenKinematics – Modular Robotics Library

Robotics
Problem

Kinematics utilities are scattered across frameworks with inconsistent APIs.

What I Built

A clean, modular Python library providing DH-parameter forward kinematics, Jacobian computation, and workspace visualization — structured for extensibility and reuse across projects.

PythonNumPyMatplotlibpytest

Reused across personal projects for kinematics validation. Designed with packaging conventions for future PyPI release.

View on GitHub
04 · GitHub Activity

Pinned repositories and active work. All projects are open-source.

face-attendance-system12

Real-time ML-based attendance via face recognition.

Python
armsim-kinematics8

Interactive robot arm forward kinematics simulator.

Python
vision-sorting-arm5

Color-based object sorting with OpenCV + hardware.

Python
openkinematics7

Modular Python library for robotic kinematics.

Python
View all repositories on GitHub →
05 · Contact

Open to opportunities.

Internships, research collaborations, or interesting robotics projects. Feel free to reach out.

Emailhello@chaitanya.devLinkedInin/chaitanya-belekarGitHubgithub.com/chaitanya-belekar