Methods, Sectors, and Ways: Artificial Intelligence for Decision-Making in Urban Circular Economy Transitions (A Narrative Literature Review)

Abstract

The transition toward an urban circular economy (CE) is fundamentally a complex decision-making challenge characterized by highly heterogeneous data streams, fragmented policy landscapes, and conflicting stakeholder objectives. While Artificial Intelligence (AI) has emerged as a vital mechanism for data processing, existing research remains siloed within individual technical applications or localized sectors. To address this gap, this study conducts a narrative literature review with thematic synthesis to examine how AI comprehensively supports decision-making in urban CE transitions across three core dimensions: applied methodologies, operational sectors, and decision-making quality. The synthesis reveals a complementary division of labor where analytical methods (machine learning and predictive analytics) serve as cognitive engines, while physical systems (computer vision and AI-driven robotics) act as the operational executioners. Infrastructurally, AI interventions primarily stabilize the heavy-emission flows of the energy and waste management sectors, while emerging applications in the built environment, urban mobility, and bio-metabolic loops (water and agriculture) reshape spatial and resource circularity. Furthermore, this study establishes that AI elevates the caliber of governance from reactive crisis containment to an evidence-based, proactive paradigm. This is achieved by anchoring decisions in multi-criteria decision analysis (MCDA), predictive lifecycle design, and cyber-physical Digital Twins that automate regulatory compliance and simulate real-time scenario testing. Ultimately, this review offers a consolidated framework illustrating that successful urban circularity depends on integrating regional and urban planning with adaptive digital ecosystems, positioning intelligent decision support at the core of sustainable metropolitan governance.