Analytical Computing for Spatial Design: An Artificial Intelligence Perspective

Next-generation people-centred design systems, frameworks, assistive tools, educational aids, and design policies necessitate foundational abstraction and computational building blocks where the modalities of human perception, action, environmental experience, and design conception and semantics are central. Our research in this context addresses the following questions: – Contemporary CAAD tools provide robust geometric modeling methods; how can the future evolution of design computing bring notions of design semantics, structure, function, and peoplecentred design to the fore at an ontological, representational and computational level? – What is the role of specialized forms of visuo-spatial perception, abstraction, and commonsense spatial reasoning, within the broader realm of design computing, spatial design assistance, and tools for design learning and education? – What is the nature and form of the assistive design feedback that designers and planners expect during the early design conception and iterative refinement phase? What are the implications of this from the viewpoint of the usability, interface, and interaction design aspects of spatial design (assistance) systems? This article presents an overview of the above stated aspects in the backdrop of relevant examples; we present abstraction, representation, and reasoning problems involving the formal modelling of structural form with respect to a desired / anticipated artefactual (mal)function. The discussion is grounded in the domain of assistive decision-support for computer-aided architecture design. Our methods are essentially AI-centric, i.e., we relate most directly with the articulation of the Science of Design by Herbert Simon and the paradigmatic relevance of Artificial Intelligence in that context.