Linking Early Design Stages with Physical Simulations using Machine Learning

Abstract

State-of-the-art workflows within Architecture, Engineering, and Construction (AEC) industry are still caught in sequential planning processes. Digital design tools founded in this domain often lack proper communication between different stages of design and relevant domain knowledge. Furthermore, decisions made in the early stages of design, where sketching is used to initiate, develop, and communicate ideas, heavily impact later stages, offering the need for fast feedback to the architectural designer to proceed with adequate knowledge regarding design implications. Accordingly, this paper presents research on a novel AEC workflow based on a 4D sketching application targeted for architectural design as a form-finding tool coupled with two modules: (1) Shape Inference module, which is aided by machine learning enabling automatic surface mesh modelling from sketches, and (2) Structural Analysis module which provides fast feedback with respect to the mechanical performance of the model. The proposed workflow is a step towards a platform integrating implicit and explicit criteria in the early stages of design, allowing a more informed design leading to increased design quality.

Advanced Computational Design – digitale Methoden für die frühe Entwurfsphase

“…Somit wird durch die Algorithmen ein rekonstruiertes Netzflächenmodell generiert, wodurch die Übertragung sowie die Bearbeitung der reinen Geometrie in weiteren analytischen Werkzeugen ermöglicht werden, wie z. B. die Rückkopplung von tragwerksanalytischen Berechnungen aus architektonischen Entwurfsskizzen [26].…”