A multi-layer tool for analysing, parameterising and generatively shaping urban fabric: computer vision reads façades and streets, a data model orders proportions, rhythms, typologies and continuity, and a generative layer varies solutions within the morphology of a place. The prototype façade module already segments elements and returns close to 120 morphological parameters.
NCBR LIDER UP, phase I, 2026A generator for what-if experiments on façade rhythm, perforation, module width, height, ground floors and street-canyon proportion, from a single façade to a fragment of the city. Research through practice, with expert validation.
ASP Gdańsk research grant, 2026Experiments with aesthetics, structure and algorithm: can generative techniques produce a new way of picturing public space that stays analytical, functional and artistic at once?
ASP Gdańsk research grant, 2026Analysis of the façade code and design guidelines for the Dolne Miasto district in Gdańsk, for the Blaszanka project, with A2P2.
Design guidelines, 2025A large language model as co-author in interpreting the city, built on the triads of Ungers’s City Metaphors. With the UAA Ungers Archiv für Architekturwissenschaft; shown at SIGNATURES 2025 and Art.’em all! 2025.
Own research, 2024 to nowReimagining Ungers’s City Metaphors in the age of AI.
Leonardo, MIT Press, in review, 2026Human and machine creativity in architecture: an AI-assisted redesign of the Szpiegowo estate in Warsaw within an NCN project led by dr Jacek Złoczewski.
Participant, 2025Clustering façade parameters into a numerical taxonomy of street walls.
Article in preparationAI-generated imagery, morphological mimicry and the identity of place.
Article in preparationPublic transport accessibility from GTFS data, set against morphology and the general plan.
Research in progressPostgraduate studies in AI and machine learning (WSB Merito) and in AI and process automation (Gdańsk University of Technology); IAAC Global Summer School programmes in computational design, urban data analytics and generative AI.
2025 to 2026