Smart Parking System Using Artificial Intelligence and IoT with Convolutional Neural Network Algorithm

Rapid urbanization and the exponential rise in vehicle ownership have intensified parking-related challenges in metropolitan areas, resulting in traffic congestion, fuel wastage, and environmental degradation. This paper presents the design and development of an intelligent parking management framework that integrates Artificial Intelligence (AI), Internet of Things (IoT), and Convolutional Neural Networks (CNN). Deployed IoT sensors and surveillance cameras continuously capture occupancy data from parking slots, which is transmitted through an edge gateway to a cloud server for storage and analysis. A trained CNN model processes camera imagery to accurately determine slot availability even under adverse lighting or weather conditions. Users interact with the system via a mobile or web application that provides real-time slot status, advance reservation, turn-by-turn navigation, and automated digital payments. An administrative dashboard offers comprehensive monitoring, analytics, and revenue management capabilities. System evaluation encompassed unit, integration, white-box, black-box, and acceptance testing, all demonstrating high accuracy and reliability. The proposed architecture is modular, scalable, and extensible to smart city ecosystems, contributing to reduced urban congestion, lower carbon emissions, and enhanced driver convenience.