Smart Park

2025

Arduino - C++ - Computer Vision - Flutter - MQTT

This work presents the development of a model representing a smart parking system. The parking lot is equipped with various sensors and actuators to ensure safety, comfort, and customization. The structure is based on a network of interconnected devices capable of communicating and cooperating with each other. Parking space availability is verified through proximity sensors or a Computer Vision model. Using a dedicated app, users can monitor parking availability and environmental conditions in real time, and remotely send commands, thus improving service efficiency and usability.

STRUCTURE

The parking facility will be structured across two levels to demonstrate how two interconnected systems can work collaboratively toward a common goal. Each level will be able to accommodate up to three vehicles. The first level will be equipped with proximity sensors to detect the presence of parked cars, while the second level will feature a computer vision system that analyzes footage from the surveillance cameras.

Both levels will include a ventilation system designed to ensure proper air circulation, which will activate automatically if poor air quality is detected. An RGB lighting system will provide dynamic illumination, which can be adjusted automatically based on environmental conditions. Additionally, manual control will be available through a dedicated mobile application. The facility will also be equipped with an air conditioning system to maintain a comfortable temperature. As with the lighting system, the HVAC system can be managed remotely via the app. Finally, an integrated fire safety system will be in place, capable of detecting fires and responding with both audible and visual alarms, followed by automatic fire suppression to ensure safety.

COMMUNICATION

All sensors and actuators communicate with an Arduino board, which autonomously manages their operation. Arduino, in turn, uses serial communication to interface with the local server acting as a gateway. This server both forwards sensor data externally and receives control commands from outside sources. The video surveillance system also communicates directly with the local server via USB. The server captures the video feed and processes it using the computer vision model.

At the core of the communication infrastructure is a cloud-based MQTT broker, which enables seamless interaction between the local server and the user’s mobile application. This allows the user to receive real-time updates about the parking facility and issue remote commands through the app.

USECASES

This gallery shows some usecases, to illustrate how the system works.

MORE

If you want to learn more, checkout the project report.

Project Report
Go back