From Data to Policy: AI-Based Predictive Technologies for Urban Mobility and Air Quality Governance

Air pollution in cities has been a major problem, and road transport is a major contributor to nitrogen dioxide (NO₂) and particulate matter (PM). LEZs have become a mainstream regulatory intervention, especially in Europe, but the effectiveness of these strategies is limited by their fixed geographical boundaries, slow feedback and their poor sensitivity to evolving urban dynamics. New developments in Digital Twin (DT) technologies provide a platform to approach adaptive, real-time governance of urban air quality through continuous sensing, predictive modelling and data-driven decision support. The present paper has provided a systematic analysis of the technological and Governance underpinnings needed to facilitate DT-driven LEZs. It investigates the purpose of sensing infrastructures and data structures, and machine learning, computer vision and algorithmic traffic models such as microscopic car-following models that convert raw urban data into useful environmental intelligence. In the review, these technologies are also placed in the larger framework of urban data governance, smart-city preparedness, and adherence to the European General Data Protection Regulation (GDPR). The main strength of this paper is the formulation of an integrated framework that connects perception, mobility modelling, emissions estimation, and policy enforcement in a consistent set of DT-supported LEZ ecosystem. The synthesis of technological innovations and institutional and legal aspects guides the paper to offer advice on the implementation of equitable, scalable, and empirically supported interventions to address air quality in European cities and outline important directions of future cross-city benchmarking and standardisation in the research field.