AI-Powered Predictive Analytics for Smart Urban Management IoT-Driven Monitoring of Informal Settlements

The rapid growth of informal settlements has become a major urban challenge, particularly in rapidly expanding cities like Mashhad, Iran. Conventional urban planning approaches rely on static datasets and manual assessments, making them ineffective for real-time monitoring and predictive urban expansion analysis. In this study, we propose the Smart Urban AI-IoT Framework, a novel approach that integrates CNN-LSTM deep learning with IoT and GIS for real-time urban monitoring. This method enhances predictive capabilities for smart city management and improves decision-making efficiency. The CNN model processes high-resolution satellite imagery to detect informal settlements with ۹۲.۴% accuracy, while the LSTM model, trained on ۱۰ years of historical urban data, predicts expansion trends with ۸۹.۱% accuracy. A LoRaWAN-enabled IoT sensor network gathers real-time air quality, traffic congestion, and security data, visualized through a GIS-based policymaking dashboard. Results show that the proposed framework outperforms traditional methods by ۲۷%, reduces urban management costs by ۳۶%, and enhances environmental monitoring by ۲۲%. The model’s applicability was further validated using data from São Paulo and Mumbai, confirming its scalability. This research underscores the potential of AI-driven urban governance and highlights future directions, including blockchain-based data security and reinforcement learning for adaptive city planning.

کلیدواژه ها:

Smart urban management ، Informal settlements ، AI-IoT integration ، Deep learning ، GIS ، Urban planning ، Predictive analytics