Find a project you can actually build
Working IoT, AI/ML and embedded builds with abstracts, component lists, documentation, and developer support — ready for your final-year submission.
Atm face regonistion transcation
Ensuring secure and contactless ATM transations has become a major concern in today’s digital banking landscape. This project proposes a multi-factor authentication system titled “ATM Face Recognition Transaction System”, which integrates facial recognition using ESP32-CAM with RFID-based user identification and a Flask-based machine learning backend for real-time verification and secure transaction authorization. The system initiates when a user presents an RFID card to an RFID reader, which reads the unique card ID and verifies it as the first layer of authentication. Upon successful card detection, the ESP32-CAM module captures a live image of the user's face and transmits it to a Python Flask server, where a trained machine learning face recognition model compares it against a pre-registered dataset. If the face matches the card owner, the transaction is authorized, and the user is granted access to the ATM functions. If the face does not match or is unrecognized, the transaction is denied, ensuring that stolen or cloned RFID cards cannot be used fraudulently. The system operates over Wi-Fi, enabling seamless integration between the ESP32-CAM, Flask web server, and database of registered users. This dual-factor authentication approach—RFID card + facial recognition—provides an enhanced layer of security, reduces physical contact, and paves the way for future touchless ATM interfaces. This project demonstrates a cost-effective, scalable, and secure ATM authentication system suitable for smart banking applications, combining IoT, embedded systems, and AI-powered face recognition.
Alcohol Detection-based Vehicle Control & Automatic Driver Drowsiness Alert System using Python Flask with GPS (Helmet-based)
Road safety has become a critical issue due to the increasing number of accidents caused by drunk driving, driver fatigue, and lack of protective gear. To address these challenges, this project presents an integrated system titled “Alcohol Detection-Based Vehicle Control and Automatic Driver Drowsiness Alert System” utilizing the capabilities of IoT, embedded systems, and sensor fusion. The proposed system employs an ESP32 microcontroller as the core unit, interfaced with an alcohol sensor (MQ-3/MQ-7) to detect intoxication, a SpO₂ sensor (MAX30102) to monitor blood oxygen and heart rate, and an eyeblink sensor to determine signs of driver fatigue. Additionally, a helmet detection module ensures the rider's compliance with safety regulations by preventing ignition in the absence of a helmet. For location tracking and emergency alerts, a GPS module is integrated, enabling real-time positioning. The system communicates with a Python Flask-based web server that logs sensor data, displays driver vitals, and generates alerts when thresholds are breached. All hardware components are mounted on a custom-designed PCB board, ensuring compactness and durability. In operation, if alcohol is detected, the ignition is disabled, preventing the vehicle from starting. Similarly, in cases of drowsiness or abnormal SpO₂ readings, the system issues audio-visual alerts and notifies connected monitoring systems via the Flask server. This multi-sensor, real-time safety system demonstrates an effective approach toward reducing road accidents by enforcing driver fitness checks, promoting helmet usage, and enabling live health and location tracking.
A Navigation and Reservation Based Smart Parking Platform Using Genetic Optimization for Smart Cities
Urbanization has led to a surge in the number of vehicles, resulting in significant challenges in parking management, especially in smart cities. Conventional parking systems often suffer from congestion, time delays, and inefficient space utilization. To address these issues, this project proposes a smart parking platform that integrates real-time navigation and reservation with Genetic Optimization algorithms for space allocation. The system is built using an Arduino Uno microcontroller interfaced with IR sensors to detect the availability of parking slots, a 20x4 LCD display to provide user instructions and slot availability, and an RFID card reader to authenticate and identify users. When a vehicle approaches, the user is authenticated via the RFID card, and available slots are checked using IR sensors. The system reserves the most optimal parking slot based on predefined conditions such as proximity to entry, user preferences, or slot rotation logic. To enhance the allocation process, a Genetic Optimization Algorithm is used at the backend (on a connected server or simulated in software) to determine the most efficient slot assignment, minimizing time and space conflict while maximizing user convenience and parking space usage. This integration of embedded hardware and intelligent algorithmic decision-making paves the way for a highly scalable, automated, and user-friendly smart parking solution. The system contributes to reduced traffic congestion, efficient urban space management, and enhanced user experience, making it ideal for deployment in smart cities.