One evening, after a long and tiring day, I decided to pre-order my food so I could save time on my way home. I thought the pickup would be quick and effortless. However, when I reached the drive-through, was meet with a never ending queue, which I waited for a very long time. I still had to stop, search for my order confirmation, tell the staff my order number, and wait while they verified my details. As more cars lined up behind me, I realised that even though I had already ordered and paid, I was still spending unnecessary time waiting.
At that moment, I asked myself, "Why can't the system recognise me automatically when I arrive?"
That simple question became the inspiration for this project. I wanted to create a smarter drive-through experience where the customer's vehicle acts as their identity. By using AI to recognise the vehicle's number plate and instantly retrieve the pre-order, the system can reduce waiting time, minimise manual verification, and make the pickup process seamless.
This project is my attempt to use Artificial Intelligence, Computer Vision, and IoT to solve a real-world problem that millions of people experience every day. My goal is not just to build a smart system, but to create technology that makes everyday life faster, easier, and more convenient.
Problem StatementCustomers who pre-order food at drive-through restaurants still need to provide their order number and wait while staff verify their order. This manual process causes delays, long queues, and reduces the convenience of pre-ordering. A smart and automated system is needed to identify customers instantly and retrieve their orders, making the pickup process faster, easier, and more efficient.
SolutionThe proposed solution is an AI-powered Smart Drive-Through Order Recognition System that uses the vehicle's number plate as a customer identifier. A camera captures the arriving vehicle's number plate, and AI-based OCR recognises it. The system then matches the number plate with the pre-order database and instantly displays the customer's order to the restaurant staff. This reduces waiting time, eliminates manual verification, minimises human error, and provides a faster and smoother drive-through experience.
System Workflow- Customer places a food order and registers their vehicle number.
- The order details and vehicle number are stored in the database.
- When the customer arrives at the drive-through, the XIAO ESP32S3 Sense camera captures the vehicle's number plate.
- AI-based OCR reads the number plate.
- The system matches the number plate with the pre-order database.
- If a match is found, the customer's order details are displayed to the restaurant staff.
- The kitchen is notified, and the customer is directed to the pickup point.
- The order is handed over quickly without requiring the customer to provide an order number.
The proposed system consists of four main modules:
1. Order Registration Module- The customer places a food order through the restaurant's website or mobile application.
- The customer enters their vehicle number.
- The order details and vehicle number are stored in the database.
- The XIAO ESP32S3 Sense camera captures an image of the arriving vehicle's number plate.
- The image is sent to the backend over Wi-Fi.
- AI-based OCR extracts the vehicle registration number.
- The recognised number plate is compared with the restaurant's pre-order database.
- If a match is found, the corresponding order details are retrieved.
- The customer's order is displayed to the restaurant staff.
- The kitchen is notified that the customer has arrived.
- An RGB LED or buzzer indicates a successful match.
- The customer is directed to the pickup counter or gate to collect the order.
Hardware: Seeed XIAO ESP32S3, Camera module, OLED Display.
Development Tools:Platform.io, Visual Studio Code
OCR (Number Plate Recognition): EasyOCR
Database: Firebase Realtime Database
Communication: Wi-Fi
Cloud Services: Firebase
Version Control: Git & GitHub
Key Features- Automatic Number Plate Recognition – Identifies the customer's vehicle using AI-powered OCR.
- Instant Order Retrieval – Fetches the customer's pre-order automatically from the database.
- Contactless Pickup – Eliminates the need to provide order numbers or confirmation messages.
- Real-Time Notifications – Alerts restaurant staff and the kitchen as soon as the customer arrives.
- Reduced Waiting Time – Speeds up the pickup process and minimizes queues.
- AI-Based Automation – Reduces manual verification and human errors.
- Wi-Fi Connectivity – Enables seamless communication between the ESP32 camera and the backend.
- Smart Status Indicators – Uses an OLED display, RGB LED, or buzzer to indicate successful order matching.
- Scalable Design – Can be integrated into restaurants of different sizes and chains.
- Secure Data Management – Stores customer and order details securely in the database.
- Customer Places an OrderThe customer places a food order through the restaurant's website or mobile app and enters their vehicle number.
- Order is StoredThe order details and vehicle number are securely stored in the restaurant's Firebase database.
- Vehicle ArrivesWhen the customer reaches the drive-through, the Seeed XIAO ESP32S3 Sense camera captures an image of the vehicle's number plate.
- Number Plate RecognitionThe captured image is sent to the backend, where OpenCV and EasyOCR detect and read the vehicle number.
- Database MatchingThe recognized number plate is compared with the stored vehicle numbers in the database.
- Order RetrievalIf a match is found, the customer's order details are retrieved instantly.
- Staff NotificationThe order details are displayed on the restaurant's dashboard, and the kitchen is notified that the customer has arrived.
- Quick PickupDisplay confirms a successful match, and the customer proceeds directly to collect their order without needing to provide an order number.
- Identify the problem in the traditional drive-through pickup process.
- Design the system architecture and workflow.
- Set up the Seeed XIAO ESP32S3 Sense camera.
- Connect the OLED display, servomotor and other required components.
- Configure the Wi-Fi connection.
- Program the ESP32 using VS Code.
- Develop the backend using Python.
- Create a staff dashboard using HTML, CSS, and JavaScript.
- Create a Firebase database.
- Store customer details, vehicle numbers, and order information.
- Connect the backend with Firebase for real-time data retrieval.
- Integrate OpenCV and EasyOCR for number plate detection and recognition.
- Process captured images and extract the vehicle registration number.
- Connect the ESP32 camera, backend, AI module, and Firebase database.
- Test communication between all modules over Wi-Fi.
- Test the system using sample vehicle number plates.
- Verify order retrieval accuracy and response time.
- Fix errors and optimize the system.
- Build a miniature drive-through lane.
- Use toy cars with printed number plates.
- Demonstrate automatic number plate recognition and instant order retrieval.
The proposed AI-Powered Smart Drive-Through Order Recognition System was successfully implemented as a working prototype. The system accurately recognizes the vehicle's number plate, retrieves the corresponding pre-order from the database, and displays the order details to the restaurant staff in real time.
The AI-Powered Smart Drive-Through Order Recognition System can be further enhanced with advanced features to improve efficiency and user experience. Future developments include:
- Integrating automatic payment verification for a completely contactless pickup process.
- Adding ETA prediction to prepare orders before the customer arrives
- Implementing voice greetings to welcome customers and provide pickup instructions.
- Introducing loyalty and VIP customer recognition for personalized services and rewards.
- Enabling automatic gate control using a servo motor after successful order verification.
- Providing real-time queue monitoring and analytics to help restaurants optimize operations.
- Integrating with restaurant POS systems and popular food delivery platforms.
- Supporting multiple languages for improved accessibility.
- Enhancing the AI model to achieve higher number plate recognition accuracy under different lighting and weather conditions.
- Expanding the system for use in fuel stations, parking facilities, toll plazas, and warehouse pickup centers.
- OpenCV. Open Source Computer Vision Library. https://opencv.org/
- EasyOCR. EasyOCR Documentation. https://www.jaided.ai/easyocr/
- Firebase. Firebase Documentation. https://firebase.google.com/docs
- Seeed Studio. XIAO ESP32S3 Sense Documentation. https://wiki.seeedstudio.com/xiao_esp32s3_getting_started/
- Flask Documentation. Flask Web Framework. https://flask.palletsproj
- Python Software Foundation. Python Documentation. https://docs.python.org/3/
- Bradski, G. (2000). The OpenCV Library. Dr. Dobb's Journal of Software Tools.
- Gonzalez, R. C., & Woods, R. E. (2018). Digital Image Processing (4th ed.). Pearson.
- Szeliski, R. (2022). Computer Vision: Algorithms and Applications (2nd ed.). Springer.
This project was developed through the collaborative efforts of our team under the guidance and support of Mr. Abhinav.
We sincerely thank the IoT club of MACE for their never ending support.
Finally, we extend our gratitude to everyone who supported us with testing, feedback, and motivation during the development of this project.






















Comments