DOI: 10.11118/978-80-7701-087-0-0025

AI-MEDIATED SCENARIO MODELLING FOR EXPERIMENTAL INTERVENTIONS IN URBAN VOIDS

Miroslav Čibik$¹

¹ Institute of Landscape Architecture, Faculty of Horticulture and Landscape Engineering, Slovak University of Agriculture in Nitra, 949 76 Nitra, Slovakia

Urban voids, often emerging as residual gaps within compact urban structures, represent both spatial challenges and opportunities for innovative design interventions. This research explores the potential of artificial intelligence (AI) as a tool for generating experimental design scenarios for the regeneration of such spaces. The study focuses on a vacant lot located within a strictly urbanized environment, characterized by fragmented spatial conditions and limited development potential. The research applies AI-based modelling to generate a series of possible spatial interventions that reinterpret the role of the site within the surrounding urban fabric. Rather than producing a single design solution, the AI system is used to explore multiple speculative scenarios reflecting different spatial, ecological, and programmatic strategies. These generated proposals are subsequently analyzed and categorized according to their spatial logic, degree of intervention, and relationship to the existing urban structure. The resulting scenarios form the basis for a structured set of recommendations that contribute to a working database of potential regeneration strategies for urban voids. The findings suggest that AI-assisted modelling can serve as a supportive exploratory tool in early design stages, expanding the range of possible spatial strategies and supporting decision-making processes in the regeneration of small-scale urban gaps within dense city environments.

Keywords: AI-mediated design, generative modelling, spatial scenarios, performative urbanism, experimental urbanism

pages: 25-28, Published: 2026, online: 2026