Smart City Digital Twin Platform Architecture for Mobility and Transport Decision Support Systems
Pierfrancesco Bellini, Enrico Collini, Marco Fanfani, Luciano Alessandro Ipsaro Palesi, Paolo Nesi
Abstract
Addressing mobility and transport problems is nowadays of paramount importance for any city due to the increasing urbanization. Traffic congestion, pollutant emissions, energy consumption are some of the problems related to urban mobility. Therefore, there is the need of tools able to support decision-makers in studying, evaluating, and planning sustainable urban evolutions. A few open-source and proprietary solutions are available requiring on-premises installations, large effort, and providing limited capabilities to actually handle real-time data (from data spaces, and standards). Moreover, they are limited in terms of analytic integration and do not offer automatic generation of suggestions. In practice they do not manage the explosion of complexity regarding computational and storage/models aspects. For these reasons, this paper presents a comprehensive architecture for a Smart City Digital Twin platform, specifically designed to support mobility and transportation decision-making through advanced what-if analysis and optimization. The platform, integrated within the Snap4City system, enables real-time data processing and complex analytics to create virtual urban environments for evaluating potential infrastructure changes. Through microservice architecture, the platform supports massive data ingestion, scenario creation, and predictive modelling, facilitating both short-term and long-term planning. The solution leverages artificial intelligence (AI), machine learning (ML), and reinforcement learning (RL) to optimize city operations and suggest actionable insights, aiding city planners in strategic and tactical decisions. This architecture has been validated through implementations in Italian cities, demonstrating scalability and flexibility to accommodate diverse urban needs and improve traffic flow, energy efficiency, and environmental impact. This work has been performed in the context of OPTIFaaS Flagship of CN MOST, the National Centre for Sustainable Mobility in Italy, and for CN HPC Big Data and Quantum Computing, ICSC.