Optimizing urban nature-based solutions through artificial intelligence: Prospects for the city of Niš
Abstract
Modern smart city concepts are evolving from techno-centric models focused on raw data collection toward ecologically sustainable urban ecosystems. The primary challenge of contemporary urbanism lies in mitigating climate change impacts, particularly the urban heat island (UHI) effect and air quality degradation. This paper explores the potentials of optimizing urban Nature-based Solutions (NbS) and green infrastructure through Artificial Intelligence (AI) and Urban Digital Twins (UDT) as advanced analytical tools. A theoretical framework is established by integrating concepts of intelligent Nature-based Solutions, IoT-based environmental monitoring, digital twin technologies, and generative urban design. Through a comparative analysis of global best practices - specifically Barcelona, Helsinki, and Vienna, the research identifies how AI-supported decision-making can improve environmental monitoring, optimize green infrastructure placement, enhance mobility management, and increase the efficiency of municipal services. The second part of the paper focuses on highlighting the prospective opportunities and open possibilities for the city of Niš, exploring how these global methodologies could address critical local issues such as summer thermal stress and winter air pollution. Rather than proposing an immediately deployable technological solution, the paper develops a conceptual framework for the gradual introduction of AI-assisted ecological planning adapted to local institutional and infrastructural conditions. While a systematic institutional framework for AI urban implementation currently does not exist in the local context, this study highlights the importance of early-stage exploratory research. Such academic inquiry is essential for bridging theoretical data science and local urban planning practice, positioning AI strictly as an analytical assistant for ecological optimization rather than a replacement for human expertise.