Francesco Carota, Gustavo Amaral
https://doi.org/10.60152/1v0upfcf
Abstract: This paper explores the application of Artificial Intelligence (AI), particularly Natural Language Processing (NLP) models to the evaluation and comparative analysis of design scenarios developed by students in an advanced architectural studio on urban dwelling, taught in fall 2024 by the authors at the University of Kansas. In the face of increasingly complex and uncertain urban conditions, the capacity to envision and communicate future scenarios has become a critical component of architectural and urban design education. Scenarios, understood as structured narratives about possible futures, serve both to articulate design intentions and to reflect on the broader implications of spatial and planning decisions. Building on previous research on the topic, our methodology explores the application of a computational framework that integrates geospatial analytics, semantic parsing of design options, and quantitative metrics derived from urban performance indicators. Using materials of student-generated projects—including plans, diagrams, geotagged program data and textual narratives—the NLP engine extracts key themes, spatial metrics and design intents from textual and drawing descriptions to assess formal and functional design characteristics according to different parameters such as activity intensity, walkability, proximity to services, and spatial diversity. By combining qualitative design thinking with data-informed interpretation, the research aims to support both instructors and students in identifying latent patterns and evaluating the effectiveness of design strategies in shaping livable, inclusive urban environments. The paper contributes to the growing field of AI-assisted architectural education by offering a replicable, scalable, and critically reflective methodology. It demonstrates how AI tools, when used with pedagogical intent, can enhance transparency, foster design awareness, and reinforce the relevance of evidence-based decision-making in architecture.
Keywords: Design Scenarios, Urban Design, Artificial Intelligence, Pedagogy, Design Studio
How to cite this Paper (Harvard referencing style):
Carota, F. and Amaral, G. (2025) ‘Intelligent Pedagogies: Assessing Student Urban Design Scenarios through AI and Semantic Spatial Analysis‘, in R. Bogdanović (ed.) On Architecture — Crosscutting and Fusion of Disciplines, Proceedings. Belgrade, Serbia: STRAND, pp. 29–40.
See publication On Architecture (2025) Conference Proceedings