Robotics Portfolio

✨ AI-Generated Project Ideas

🔌 Smart Power Home Automation System (CAPE)

A comprehensive end-to-end IoT solution for real-time energy monitoring, automated billing, and AI-driven predictive maintenance, designed for homes, industries, and smart city infrastructure.

🏆 WD RoboSoccer Bot

Designed and built a high-torque 4-wheel drive robot for a RoboSoccer competition. Integrated four JGB motors with precision control using a Sabertooth dual-channel motor driver. Enabled smooth remote operation through an FS-iA10B receiver paired with a multi-channel transmitter. Optimized mechanical design for speed, stability, and agility on the soccer arena.

🏆 Awarded Best Technical Design for innovative chassis layout and robust motor control strategy.

Vice Technical Head – IEEE Sensors Council, RIT-B

Driving Innovation, Events, and Real-World Sensor Applications

Spearheaded the flagship RoboSoccer event at TechFest 2025, RIT-B — designed and built four custom 4WD RoboSoccer bots from scratch, controlled via handcrafted wireless remotes using nRF modules. The bots demonstrated agility, precision, and responsiveness on the field, making this the most technically acclaimed and crowd-favorite event of the entire fest.
Conceptualized and executed “Breakpoint”, a unique circuit debugging challenge. Designed two custom sets of PCBs — one with perfect functionality and the other embedded with smart faults. Participants were given the faulty set, testing their real-world electronics and fault-finding skills. The event was a massive hit, blending creativity, hands-on electronics, and critical thinking.
Organized and developed the “SensoryBot” challenge, a beginner-to-advanced robotics event where participants built sensor-based bots under guided problem statements. Personally designed and demonstrated four fully working bots — line follower, light follower, edge detector, and maze solver — as reference builds. Supplied kits and support, turning it into a super engaging, fun, and educational experience.
Currently leading an advanced drone development project under IEEE-RITB, focused on thermal imaging for industrial and defense applications. The drone uses infrared sensors to detect human presence behind obstacles and identify thermal anomalies in machines, aiming to enhance rescue ops, surveillance, and predictive maintenance using real-time sensor fusion and embedded intelligence.

Electrical Subsystem Member – Team Volante

EV Go-Kart and Off-Road Buggy Development | Power Systems | Safety Electronics

Core member of Team Volante, RIT-B’s premier automotive student club focused on building electric go-karts and off-road buggies for national-level EV design competitions.
Contributed to the design and implementation of low-voltage (LV) safety circuits for the team’s electric buggy, including:
- TSAL (Tractive System Active Light): Designed a robust and compliant indicator circuit to signal the high-voltage system's active state, using LED arrays and status feedback from the shutdown loop.
- Shutdown Circuit: Helped integrate emergency kill-switches, brake plausibility checks, and interlocks into a centralized low-voltage loop to ensure rapid and safe deactivation of the powertrain in case of faults or driver-triggered events.
Worked on power distribution and protection planning, ensuring clean and regulated 12V/5V supply lines for sensors, dashboard, telemetry, and auxiliary systems, while adhering to competition rules and safety protocols.
Collaborated with mechanical and drivetrain teams to route wiring safely through the chassis, enabling modular testing, quick diagnostics, and efficient debugging during race day prep.

🤖 Reinforcement Learning Self-Balancing Robot

Objective

To design and develop a real-time, self-balancing two-wheeled robot that integrates traditional PID control with a reinforcement learning (RL) model (PPO) for adaptive balancing and dynamic control. The project aims to demonstrate how real-world robotics can benefit from hybrid control systems, improving stability and responsiveness in uncertain environments.

Key Highlights

Developed a real-time balancing robot using an MPU6050 IMU, BTS7960 motor drivers, Arduino Mega, and a Raspberry Pi for computation and PPO integration.
Implemented a dual control loop: Primary: PID controller for high-frequency, deterministic corrections. Secondary: PPO (Proximal Policy Optimization) model trained using real robot sensor data for adaptive angle correction.
Collected real-time data over serial (angle, error, zone, PID output) and built a custom dataset using Python & CSV logging.
Designed and trained a custom RL environment (OpenAI Gym compatible) based on real-world error-action data using Stable-Baselines3 with PPO.

Tools & Frameworks

🧠 Reinforcement Learning: PPO via Stable-Baselines3 | 🧰 Embedded Systems: Arduino C++ (MPU6050, motor control, PWM) | 🐍 Python: Serial communication, data logging, RL model training

Achievements & Results

Achieved robust self-balancing with < ±2° steady-state error under most conditions.
Trained PPO agent successfully learned corrections that improved recovery time after large disturbances by up to 25%.

View Code on GitHub

🏆 Smart Power Home Automation System (CAPE)

An advanced, full-stack IoT solution developed to tackle critical challenges in energy consumption, safety, and efficiency — CAPE (Control and Automation of Power Efficiency) is a hardware-software integrated system that enables users to monitor, control, and optimize their electrical usage in real time.

🥇 Awarded “Best Hardware Project” for its innovative design and real-world impact.

🎯 Problem It Solves

Unmonitored energy consumption leading to high electricity bills.
No real-time visibility into per-device power usage.
Risk of fire or equipment damage due to unnoticed overcurrent/voltage.
Manual and delayed electricity billing process.
Lack of predictive maintenance or preventive alerts in home electrical systems.

🌟 Features & Functionality

Real-Time Monitoring: Voltage and current readings captured per device via sensors connected to an Arduino.
Web Dashboard (Flask-based): Real-time data visualization and device control, accessible via local network.
Automated Billing System: Bills are generated every 30 days and can also be triggered on-demand, using CSV-based consumption logs.
AI-Powered Predictive Maintenance: Anomaly detection engine (built with LightGBM and scikit-learn) predicts potential faults based on historical energy usage patterns.

🔧 Technology Stack

Hardware: Arduino UNO, Current & Voltage Sensors, Relays, Keypad, GSM module, Raspberry Pi | Software: Python, Flask, CSV logging, pandas, NumPy, scikit-learn, LightGBM | Frontend: HTML, CSS, Chart.js

🧠 Why This Project Stands Out

Combines real-time monitoring, automated control, AI, and IoT into a fully functional and scalable prototype.
Goes beyond traditional automation by providing proactive safety and personalized energy-saving advice.

🚀 Future Goals

We are actively working on transforming this prototype into a market-ready product. The next stage includes designing a single integrated PCB, enabling plug-and-play compatibility, and adding mobile app integration.

🛠️ AI-Powered Corrosion Monitoring System

This project presents a scalable, low-cost, and intelligent solution to one of the most critical challenges in civil, marine, and industrial infrastructure: corrosion monitoring and prediction. Using the AD5933 impedance analyzer IC and a Raspberry Pi 3B+, I built a real-time system that monitors the corrosion state of metal samples submerged in corrosive media.

🔍 Problem Statement

Traditional corrosion monitoring techniques are often invasive, expensive, and incapable of real-time analysis or prediction. This project aims to solve that by providing a portable, scalable, and intelligent corrosion detection system using embedded hardware and software.

⚙️ Technologies & Tools Used

Hardware: AD5933 Impedance Analyzer, Raspberry Pi 3B+ | Software: Python (smbus2, pandas, NumPy), matplotlib, TensorFlow Lite (planned) | IoT: ThingSpeak & Node-RED (optional)

📊 Key Features

Real-time impedance measurement using AD5933.
Visualization of impedance vs. time to observe corrosion trends.
Data logging with timestamped records for ML training.

🚀 Future Advancements

Train and deploy advanced ML models to predict corrosion rate and remaining useful life (RUL), add environmental sensors, and build solar-powered LoRa-based sensor nodes for remote deployment.

🌟 AURORA – Autonomous Mobile Robot (AMR)

AURORA is a state-of-the-art Autonomous Mobile Robot (AMR) developed by the Robotics and Automation Society (RAS) to address real-world logistics, industrial transport, and indoor mobility challenges. Built with modularity, intelligence, and adaptability in mind, AURORA integrates high-performance hardware with an intelligent software stack capable of localization, navigation, environment sensing, and fail-safe autonomous operations.

⚙️ Core Objectives

Enable intelligent autonomous navigation in cluttered indoor spaces.
Transport dynamic payloads up to 100 kg.
React to unpredictable obstacles with sub-second delay.
Learn and optimize routes over time using onboard edge computing
Serve as a scalable platform for warehouse, research, and industrial logistics

🧩 System Architecture

Mechanical: Chassis: Aerospace-grade aluminum frame, Drive: 4WD skid-steer, 48V 30Ah Li-ion pack. | Compute: Raspberry Pi 5 (SLAM, Path Planning), Arduino Mega 2560 (Low-level control). | Sensors: 2D LIDAR, ToF, IMU, Gas/Temp/Humidity, Load Cells.

🧠 Software Stack

OS: Raspberry Pi OS | Middleware: ROS2 Humble | Navigation: RTAB-Map/Cartographer + Nav2 | UI: Flask

🚀 Future Advancements

Auto-docking/charging, fleet/swarm logistics, mission profiles, and edge-based reinforcement learning for path optimization.

Analog PID-Based Line Following Robot

This project is a line follower robot built using analog IR sensors and a PID (Proportional-Derivative) control algorithm to follow a black line on a white surface with high precision and smooth movement. The robot uses three analog IR sensors to detect the line position and adjusts its motor speeds accordingly using a PID algorithm implemented on an Arduino-based microcontroller. The motor control is achieved using the DRV8833 dual H-bridge motor driver.

⚙️ Hardware Components

Arduino (e.g., Nano/Uno), 3x TCRT5000 analog IR sensors, DRV8833 motor driver, 2x N20 DC gear motors, Wheels and chassis, Battery pack (7.4V or similar).

🧠 How It Works

Three analog IR sensors detect the line, giving values based on surface reflectivity.
A weighted average method computes the error in the robot's position.
The Proportional (Kp) and Derivative (Kd) components are used to calculate the correction factor.
The left and right motor speeds are adjusted based on the correction value to steer the robot.

✅ Project Highlights

Implements PID control for smoother and accurate line following.
Works well on curved tracks due to proportional correction.
Ideal as a foundation for advanced robotics: maze solvers, obstacle avoiders, or autonomous path planners.

View Code on GitHub

🔍 ESP32-CAM Smart Vision System

🧠 Project Objective

The goal of this project is to develop an affordable, compact, and efficient image processing system using the ESP32-CAM module to assist people with disabilities—particularly those with visual impairments or mobility challenges. The system is designed to provide visual input recognition, process real-time images, and convert them into actionable feedback, such as voice alerts or control signals.

🔧 Hardware Used

ESP32-CAM Module (OV2640 camera), FTDI Programmer, 5V regulated power source, MicroSD card, Peripheral output (buzzer, speaker, or vibration motor).

🔬 Mathematics & Algorithms Involved

Image Representation: Images are matrices of pixels (RGB image → 3D matrix).
Color Thresholding & Grayscale Conversion.
Edge Detection (Sobel Operator).
Face/Object Detection (Haar Cascades or TinyML).

🤝 Social Impact

This project is a step toward accessible assistive technology. This low-cost ESP32-CAM implementation can guide visually impaired individuals, help individuals with limited mobility recognize surroundings, and be expanded into smart gloves, wearable visors, or home assistant devices.

View Code on GitHub

Tinkercad Circuit Simulations

🅿️ Autonomous Parking Robot

🎯 Project Objective

To design and implement a fully autonomous parking robot capable of randomly assigning itself to one of twelve designated slots and performing precise parking and exit maneuvers — all while following a line-based navigation path and providing real-time visual feedback using a 4-digit 7-segment display.

🧰 Hardware Used

🧠 Arduino Uno | 🚗 2x DC Motors & L298N Driver | 🟢 3x HW201 IR Sensors | 🔲 TM1637 4-digit display | 🔋 7.4V/12V Li-ion PSU

⚙️ Functional Workflow

Random Slot Assignment: Generates a random slot (1-12) and displays it.
Line-Following Navigation: Autonomously follows a path using IR sensors.
Parking Logic: Reverses into the assigned slot and displays a 10-second countdown.
Exit Phase: Exits the slot after the countdown and continues on the path.

🌟 Key Highlights

Robust state-machine-based logic handles each stage of operation cleanly.
Real-time slot display and countdown improve clarity and challenge compliance.

View Code on GitHub

Fatigue Detection System Using Arduino Uno

This project focuses on designing and implementing a low-cost, hardware-based driver fatigue detection system using simple, reliable sensors and embedded control logic. It aims to reduce road accidents caused by drowsy driving, particularly for truck drivers, night-shift taxi drivers, and long-distance commuters.

Key Features:

Real-time monitoring of driver alertness using two simple sensors.
Immediate warning using sound and light signals.
Timeout logic for persistent alert in case of no driver response.
Clean embedded logic using Arduino (no external computing needed).

Technical Overview:

The capacitive touch sensor is used to continuously detect the presence of the driver’s hand on the steering wheel. The tilt sensor identifies the driver’s head orientation. If both sensors report fatigue conditions, the system activates a buzzer and LED and starts a 10-second countdown. If the driver does not respond, the alert continues.

View Code on GitHub

FruitBridge: Amazon Sambhav Finalist

FruitBridge is a project developed for the Amazon Sambhav Hackathon, where it was selected as a Top 20 Finalist from over 50,000 participants. The project aims to bridge the gap between farmers and consumers, providing a platform for fair trade and fresh produce.

View Code on GitHub

🔦 Light Follower Robot – Arduino Based

📌 Overview

The Light Follower Bot is an autonomous two-wheeled mobile robot designed to detect and follow the brightest light source in its environment. This project was developed using an Arduino Uno/Nano, basic electronic components like LDRs (Light Dependent Resistors), and a motor driver module (L298N/L293D). It serves as a beginner-friendly introduction to sensor-based robotics, analog signal processing, and embedded control systems.

🎯 Objective

The goal of this project was to build a light-seeking robot that could detect varying light intensities using LDR sensors and dynamically adjust its direction of movement toward the source. The bot constantly reads analog values from two LDR sensors and makes real-time decisions to turn left, right, or move forward based on the comparative intensity of light received from each side.

🔧 Hardware Used

Arduino Uno / Nano (ATmega328P), L298N / L293D Motor Driver Module, 2× DC Geared Motors with Wheels, 2× LDRs (Light Dependent Resistors), 2× 10kΩ Resistors (Voltage Divider Circuit), Battery Pack (7V–12V recommended), Chassis and caster wheel, Jumper wires and breadboard for prototyping.

View Code on GitHub

🤖 Sign Language Interpreter & Smart Learning Glove

An intelligent wearable system designed to both recognize and teach sign language through a sensor-rich glove integrated with machine learning, servo actuation, and chatbot assistance.

🔍 Overview

This project tackles two problems: 1. Helping the general population understand and recognize sign language. 2. Teaching individuals how to perform correct signs interactively. The solution is a smart glove that can interpret sign language in real-time using flex sensors and motion tracking, while also acting as an assistive learning tool by physically demonstrating signs using servo motors and an integrated AI chatbot.

🧠 Key Features

Real-Time Sign Detection: Equipped with flex sensors and an IMU, the glove captures finger positions and hand orientation.
Machine Learning Integration: Sensor data is sent to a Raspberry Pi 4, which runs a trained ML model to classify the gesture.
Chatbot-Guided Learning Mode: A built-in chatbot interface allows users to ask, “How do I say X in sign language?” The system responds by actuating the glove to physically demonstrate the sign.

MODEL: ATLAS-PRIME

MOHAMMED RAYAN

// CONNECT

// ABOUT_ME

// CORE_SKILLS

// FEATURED_PROJECTS (Click on each Section to view it)

Smart Power Home Automation (CAPE)

Self-Balancing Assistive Walker

AURORA – Autonomous Mobile Robot

Analog PID Line Follower

Autonomous Parking Robot

Fatigue Detection System

Smart Vision System for Disabilities

Light Follower Robot

AI-Powered Corrosion Monitoring

Sign Language Interpreter Glove

Tinkercad Circuit Simulations

FruitBridge

// POSITIONS_OF_RESPONSIBILITY

Vice Technical Head

Electrical Subsystem Member

// ACHIEVEMENTS

Best Hardware Project

Best Technical Design

// RESEARCH

PID vs PPO for Real-Time Self-Balancing Robots

// CERTIFICATIONS

// ✨ AI_PROJECT_IDEATION