Hi, I'm Sahil Rajpurkar

Robotics Engineer | AI Enthusiast | Embedded Systems Developer

Building intelligent robotic systems that integrate software, hardware, and AI to solve real-world problems.

Sahil Rajpurkar

About Me

I'm a passionate robotics engineer with expertise in autonomous systems, computer vision, and embedded development. My work focuses on creating intelligent robotic solutions that bridge the gap between software and hardware.

I have extensive experience with ROS/ROS2, SLAM algorithms, and real-time perception systems. My technical skills span the entire robotics stack, from low-level embedded programming to AI-driven decision making.

Here are some technologies I've been working with recently:

ROS/ROS2
Isaac Sim
Gazebo
Python/C++
SLAM
OpenCV
Working on robotics project

Technical Skills

Robotics Frameworks

ROS/ROS2, Isaac Sim, Gazebo, MoveIt, Nav2, RTAB-Map

Programming

Python, C++, HTML, CSS, JavaScript, YAML, JSON, Bash, MATLAB

AI & Computer Vision

OpenCV, TensorFlow, SLAM, Object Detection

Embedded Systems

Arduino, Raspberry Pi, STM32, AVR (Atmel), Sensor Integration

Navigation & Control

Path Planning, PID, Kalman Filters

Tools & Platforms

Git, Docker, Linux, AWS IoT

Professional Experience

Research Assistant – Robotics and Communication

Apr 2023 – Present

Communication Networks Institute, TU Dortmund

Dortmund, Germany

Click to view details

  • Autonomous Navigation: Designed systems using ROS2 and SLAM algorithms
  • 3D Object Detection: Developed LiDAR-based system using NVIDIA Isaac Sim, ROS2, and Python
  • Robotic Air Hockey: Created system with user-controlled arm and AI-trained autonomous arm
  • Drone Interference Analysis: Collected spectrum data and trained ML models to detect signals (Bluetooth, FHSS, OFDM)

Research Assistant – Embedded and Testing

Jul 2020 – Sep 2022

Biomedical Engineering and Technology Innovation Centre (BETiC), IIT Bombay

Mumbai, India

Click to view details

  • Medical Devices: Led development of ventilators and oxygen generators, ensuring performance and ISO compliance
  • Sensor Integration: Worked with UART, SPI, and I2C protocols and programmed STM32/ATMEL microcontrollers
  • Real-time Algorithms: Developed diagnosis algorithms in Embedded C using RTOS with Git version control
  • Testing: Conducted software- and device-level testing of medical devices in accordance with ISO 13485 standards

Featured Projects

Robotics
LineFollowing PID vs RL

LineFollowing PID vs RL

A virtual robotics racing simulation inspired by the Robocon competition, built with Isaac Sim and ROS2. Two Robotnik Summit robots race by following a white line path while avoiding obstacles: Robot 1 uses PID control, and Robot 2 improves its line-following with Reinforcement Learning. Run your own control code on the robots using ready ROS2 topics and sensors!

  • Isaac Sim
  • ROS2
  • PID Control
  • Reinforcement Learning
  • Python
Robotics
UR5-BarAlign-RL-Isaac-sim

UR5-BarAlign-RL-Isaac-sim

Developed a reinforcement learning pipeline for bar alignment using the UR5 robotic arm in NVIDIA Isaac Sim and ROS2. The workflow includes data collection with manual joystick control, structured episode formatting for LeRobot, training with diffusion policy, and model evaluation in simulation.

  • Isaac Sim
  • ROS2
  • LeRobot
  • Reinforcement Learning
  • Python
Robotics
LiDAR-based 3D Object Detection System

LiDAR-based 3D Object Detection System

Developed a LiDAR-based 3D object detection system for a mobile robot using NVIDIA Isaac Sim, ROS2, and Python. Built 3D SLAM using LiDAR, colorized the point cloud with an RGB camera, and streamed the fused 3D data to MMDetection for real-time object detection.

  • Isaac Sim
  • ROS2
  • LiDAR
  • MMDetection
  • 3D SLAM
  • Python
  • TensorFlow
Robotics
Robotic Air Hockey System

Air Hockey System using xArm Robotic Arm

Designed a robotic air hockey system using uFactory xArm 6-axis robotic arms. Simulated a human vs machine game where one arm was user-controlled and the other trained using machine learning for autonomous gameplay. Enabled real-time interaction, trajectory tracking, and strategy learning to mimic competitive behavior.

  • xArm 6
  • Robotics Simulation
  • Machine Learning
  • Python
  • Real-time Control
AI/ML
Prototypical Chatbot for SI/PI/EMC

Prototypical Chatbot for SI/PI/EMC | TU Dortmund

Created an AI-based chatbot using RASA NLU to assist PCB designers with Signal Integrity (SI), Power Integrity (PI), and Electromagnetic Compatibility (EMC) issues, providing real-time guidance and optimization within EDA workflows. Integrated simulation software into the chatbot, enabling on-platform electrical design simulations.

  • RASA NLU
  • AI/ML
  • EDA Tools
  • Python
AI/ML
Drone Signal Interference Detection

Drone Signal Interference Detection Using ML

Developed an AI-based signal classification pipeline to detect and identify wireless interference in drone communication systems. Collected real-time spectrum data using SDR tools and trained models to classify interference from Bluetooth, FHSS, and OFDM signals.

  • Software Defined Radio (SDR)
  • Spectrum Analysis
  • Signal Classification
  • Python (Scikit-learn, NumPy)
  • Bluetooth / FHSS / OFDM
  • Machine Learning
Robotics
Rescue Robotics Demonstration

Rescue Robotics Demonstration

Integrated mm-Wave health monitoring sensors onto a Boston Dynamics quadruped robot in a Linux-based environment, enabling real-time physiological data acquisition over 5G. Conducted rescue operation simulations at the Deutsches Rettungsrobotik Zentrum (DRZ) to evaluate system performance in search-and-rescue scenarios.

  • Boston Dynamics
  • mm-Wave Sensors
  • Linux
  • 5G Communication
  • Rescue Robotics
  • Sensor Integration
Robotics
Palletizer-Cell-Control-Security-System

Palletizer Cell Control Security System

Built a complete simulation of a ROS 2-driven robotic palletizing cell integrated with a REST API and real-time web-based HMI. The system processes palletizing requests from a warehouse management system (WMS), monitors safety-critical parameters, and allows human interaction through a live dashboard.

  • ROS 2 Humble
  • REST API (Python FastAPI)
  • WebSocket + JavaScript HMI
  • Real-Time Simulation
  • Robotic System Control
  • Linux (Ubuntu 22.04)
Embedded
ESP32-S3 Fire Redundant System

ESP32-S3 Fire Redundant System using IR Thermal Imaging

Built a fire detection and response system using ESP32-S3, MLX90640 IR thermal camera, and a relay-controlled actuator. The system detects abnormal heat zones and automatically triggers a fire extinguisher while sending real-time WiFi alerts to a server. Features a configurable temperature threshold and modular Arduino codebase.

  • ESP32-S3
  • MLX90640 Thermal Camera
  • Relay Control
  • Arduino
  • WiFi & HTTP
  • Embedded Systems
Embedded
Fire Bird Robot - Line-Following & Obstacle Avoidance

Line-Following with Obstacle Avoidance: Fire Bird Robot

Designed and implemented a line-following robot using the Fire Bird V platform (ATmega2560) developed under IIT Bombay’s MHRD e-Yantra initiative. The robot follows a line and avoids obstacles in real time using IR and proximity sensors, and is fully programmed at the register level via Atmel Studio.

  • ATmega2560 & ATmega8
  • IR Proximity Sensors
  • Line Sensors
  • Motor Control (PWM)
  • Register-Level Programming
  • Atmel Studio
  • e-Yantra IIT Bombay
Embedded
Robot Drive Systems

Robot Drive Systems: Multi-Drive Robot Control with AVR

Designed and implemented drive control systems for differential, 3-wheel omni, and 4-wheel omni robots using AVR ATmega2560 microcontroller and Bluetooth-based PS2-style controller input. Features include PWM motor control, drive-specific kinematics, and wireless joystick integration for advanced robotic motion.

  • AVR ATmega2560
  • 2W / 3W / 4W Omni Drive
  • PWM Motor Control
  • Kinematic Mapping
  • Bluetooth (HC-05/06)
  • PS2 Joystick Controller
  • Atmel Studio
  • UART Communication
Embedded
line-following-three-wheel-omni-drive

Line-Following Three Wheel Omni Drive

Designed and developed a three‑wheel omni‑directional robot for the Robocon competition. The robot autonomously follows a white line using an OPT101 sensor array. A custom line-following algorithm was implemented on an AVR ATmega2560 microcontroller. The system leverages PWM motor control and kinematic mapping to achieve smooth and responsive navigation.

  • Line-Following
  • AVR ATmega2560
  • 3W Omni Drive
  • PWM Motor Control
  • Kinematic Mapping
  • Bluetooth (HC-05/06)
  • PS2 Joystick Controller
  • Atmel Studio
  • UART Communication
Embedded
Swerve Drive System

Swerve Drive System

A custom swerve drive implementation using ATmega2560 and AVR C. It controls four drive motors and two power window motors via a belt-gear mechanism for real-time wheel orientation. The system uses encoder feedback for precise control and is operated wirelessly with a PS2 Bluetooth joystick.

  • AVR ATmega2560
  • Swerve Drive
  • PWM Motor Control
  • Kinematic Mapping
  • Bluetooth (HC-05/06)
  • PS2 Joystick Controller
  • Atmel Studio
  • UART Communication
Embedded
Prototype Color Sorter Machine

Prototype Color Sorter Machine

This project presents a working prototype for automatically sorting objects based on their color. The system uses an OPT101 photodiode sensor for color detection and is controlled by an STM32F103 microcontroller. Color is determined via ADC values, and two servo motors handle object positioning and sorting into designated bins.

  • STM32F103x
  • OPT101 Color Sensor
  • ADC Signal Processing
  • Servo Motor Control
  • Embedded C
  • Hardware Prototyping
Embedded
IR Health Monitor on Shoulder Brace

IR Health Monitor: Wearable Vital Monitoring System

Developed a wearable health monitoring system integrated into a shoulder brace for post-surgery recovery. Powered by an ESP32, the device reads SpO2, breath rate, and skin temperature using the SparkFun MAX30102 and MLX90640 sensors, and transmits the data wirelessly to a remote server via HTTP.

  • ESP32
  • MAX30102 (SpO2 & Respiration)
  • MLX90640 Thermal Sensor
  • Arduino
  • HTTP & JSON
  • IoT Health Monitoring

Education

Master's in Automation and Robotics

Oct 2022 – Feb 2025

Technical University Dortmund

Dortmund, Germany

Click to view details

  • Specialization: Robotics and Cognitive Systems
  • CGPA: 2.3/4.0
  • Thesis: "Development of Machine Learning-based Radar Object Detection for Logistics Entities"
  • Key Courses: Mobile Robots, Machine Learning, Robotic Manipulators, Computer Vision

Bachelor's in Electronics and Telecommunication

Aug 2016– Jul 2020

Mumbai University – K.J. Somaiya College of Engineering

Mumbai, India

Click to view details

  • Major: Embedded Systems
  • CGPA: 2.0/4.0
  • Thesis: "Controlled Motion of a Pneumatic Quadruped Robot"
  • Key Courses: Control Systems, Embedded Systems, Digital Signal Processing

Thesis

Radar-Based Detection Preview Radar Setup
🎓 Master's Thesis

Towards 6G-Driven Sensing: Radar-Based Object
Detection for Logistics Entities

Technical University Dortmund

Oct 2023 – Jun 2024

  • Developed ML-based detection pipeline using TI’s IWR6843ISK mmWave radar for logistics automation.
  • Trained YOLOv7, Detectron2, and OpenPCDet models on radar-derived 2D/3D data for entity detection (e.g., forklifts, robots).
  • Enabled real-time operation in cluttered indoor environments, demonstrating radar’s resilience and privacy-preserving capabilities over LiDAR and cameras.
Thesis Preview Thesis Diagram
🎓 Bachelor's Thesis

Controlled Motion of a Pneumatic
Quadruped Robot

K J Somaiya College of Engineering

Sep 2019 – Apr 2020

  • Collaborated with Mercury Pneumatic Pvt. Ltd. to design a pneumatic quadruped robot with two active degrees of freedom.
  • Developed and tested C++ control algorithms for walking and pronking gaits to evaluate robot dynamics and controllability.
  • Customized the PCB to optimize space and ensure seamless integration within the robot’s structure.

Publications

Gaits Stability Analysis for a Pneumatic Quadruped Robot Using Reinforcement Learning

Soofiyan Atar, Adil Shaikh, Sahil Rajpurkar, Pragnesh Bhalala, Aniket Desai, Irfan Siddavatam

International Journal of Industrial and Manufacturing Engineering, Vol. 15, No. 9, WASET, 2021

This paper presents a stability analysis of various gaits for a pneumatic quadruped robot, leveraging reinforcement learning techniques to enhance locomotion performance and adaptivity.

Internships

DIY Robotics Intern

Jun 2020 - Jul 2020

Robotbanao

Mumbai, India

Click to view details

  • Arduino-based automation projects with Bluetooth and voice control
  • Simulated circuits in Proteus for robotic arm and home automation
  • Implemented servo, relay, and ultrasonic sensor logic
  • Created documentation with pin diagrams and simulation outputs

Embedded Systems Intern

Jan 2020 - May 2020

Enigma Machines Private Limited

Mumbai, India

Click to view details

  • Developed custom ladder logic for EPE foam lamination machine automation
  • Designed an HMI-based GUI for real-time machine monitoring and control
  • Enhanced operational efficiency with tailored PLC solutions
  • Addressed lack of local automation options with a custom-built control system

Embedded Systems Intern

Jul 2019 - Dec 2019

Kidyfit Care Private Limited

Mumbai, India

Click to view details

  • Worked on a smart health monitoring device for newborns to detect sleep apnea
  • Contributed to PCB design, component selection, and prototype development
  • Handled microcontroller programming, sensor integration, and debugging
  • Delivered a successful first-pass PCB embedded in the working prototype

e-Yantra Summer Internship

Jun 2018 - Jul 2018

e-Yantra Lab, IIT Bombay

Mumbai, India

Click to view details

  • Developed a line-following robot with object (color) detection using Fire Bird V ATMEGA2560
  • Programmed in Embedded C using Atmel Studio with register-level control (PWM, ADC)
  • Interfaced IR sensors, white line sensors, and motor drivers

Competitions

RoboCon Image 1 RoboCon Image 2
🏆 5th Place | 🏅 Winner - MATLAB Competition

RoboCon 2019

Team KSJCE RoboCon

IIT Delhi • Aug 2019

  • Led a 30-member interdisciplinary team (mechanical, software, hardware) for RoboCon and MATLAB competitions
  • Developed and programmed 3-wheel and 4-wheel autonomous robots using laser mapping
  • Implemented line-following and custom SLAM algorithms, improving navigation speed by 40%
  • Competed against 100+ university teams at a national level
RoboCon Image 1 RoboCon Image 2
🥇 1st Place | 🏅 Winner - Safety Award

Shell Eco-marathon 2019

Team Etaη

Shell Technology Centre, Bangalore • Nov 2019

  • Refined a bicycle-based engine to achieve an exceptional fuel efficiency of 268.7 km/l
  • Designed and integrated a thermal camera-based fire-redundant safety system with IoT support
  • Delivered a technical presentation to a panel of industry experts
eMEDHA Image 1 eMEDHA Image 2

eMEDHA 2021

🏅 Best Creativity Award
Team 22

IIT Bombay, Mumbai • May 2021

  • Built a cost-effective, adjustable shoulder brace for post-surgery recovery
  • Defined problem through clinical meetings with doctors
  • Integrated IR sensor to monitor SpO2, skin temperature, and breath rate
  • Implemented real-time health alerts for patient safety

Awards & Honors

PUSH Scholarship

KJ Somaiya College of Engineering

Mumbai, India

5th Place

Robocon 2019

IIT Delhi, India

1st Place

Shell Eco-Marathon 2019

Bangalore, India

Best Creativity Award

eMEDHA

IIT Bombay, Mumbai, India

Volunteering

Technical Volunteer

ROBOCUP 2024


Eindhoven, Netherlands

17th – 21st Jul 2024
  • Assisted participants in using the vSTING module to simulate network degradation effects
  • Helped configure latency and packet loss settings
  • Supported troubleshooting during simulations

Students Volunteer

DEGIS-German Association for International Students

Bochum, Germany

Nov 2022 – May 2023
  • Organized events for international students, ensuring an inclusive and memorable experience
  • Documented events through videos and photos
  • Managed social media posts and engagement

Mentor

Medical Device Innovation Camp (eMEDIC)

Mumbai, India

Oct 2021
  • Helped participants define clear problem statements for the medical hackathon
  • Provided feedback and analysis on pitch presentations (PPTs)
  • Assisted with hardware component planning and sourcing

Coding Instructor

Workshop - Introduction to Arduino


Mumbai, India

June 2018
  • Teach robotics fundamentals to underprivileged students
  • Develop curriculum for beginner Arduino workshops
  • Mentor teams for internal robotics competitions

Get In Touch

Contact Information

sahilrajpurkar1998@gmail.com

+49 176 71678983

Dortmund, Germany