Identifying and Evaluating Urban Needs Through Relational Reconstruction

Urban design relies on judgments about existing and anticipated needs, while the available evidence ranges from lived experience and historical records to geospatial observation, statistical estimation, and computational simulation. Contemporary developments in geospatial artificial intelligence, agent-based modeling, urban digital twins, and computational search expand the means of inquiry while raising questions about inferential validity and methodological compatibility. This paper develops a question-guided framework for identifying and evaluating candidate urban needs through the reconstruction of spatial, temporal, social, ecological, and infrastructural relations. Two complementary inquiry pathways are distinguished: hypothesis-driven assessment of specified candidate needs and exploratory search for candidates emerging from observed or simulated relational patterns. Both pathways are connected to explicit procedures of evidential appraisal, causal examination where identifiable, participatory scrutiny, and post-intervention evaluation. A provisional formal interface is proposed in which generative–relational (GR) subjects have question-dependent spacetime-supported representations, heterogeneous evidence is linked by category-theoretic constructions, and selected event structures are encoded through algebraic-topological invariants of labeled two-complexes. Evolutionary updates of these representations and spin-foam-inspired cellular initializations are considered as exploratory modeling options, with the physical meaning of quantum-gravity spin foams kept separate from urban research. The framework is intended to guide methodological comparison and further model development; its empirical adequacy requires case-specific operationalization and validation.