Overview
The construction industry is one of the least digitized sectors of the economy (cf. McKinsey Global Institute Industry Digitization Index 2015). In order to meet the requirements of an increasingly complex construction process and to ensure a continuous flow of information throughout the entire life cycle of a structure, specific technology solutions are required.
There are several AI-powered platforms and tools that support the design, planning and fabrication process in the construction industry. As part of the state of the art, various offerings have been elicited and analyzed (e.g., Midjourney, DALL-E, ChatGPT, Rhino/ Grasshopper) that generate images, text, and 3D models (Fig. 1). What is missing so far is a methodology that shows the possible uses of AI applications and describes the integration possibilities into existing systems. For example, this involves the workflow from prompt engineering in ChatGPT to the Phyton programming language as an interface for integrating scripts to the Rhino + Grasshopper geometric applications.
In the application the methodology serves the users (among others architects and engineers) e.g. for the development of design variants to achieve a better integration of sustainability aspects. The focus here is on the development of innovative and creative solutions that bring about a reduction in CO2 emissions in the construction industry, e.g. through AI-optimized designs and AI-supported recycling. Furthermore, with an AI-supported information flow, from design to factory planning to machine fabrication, the construction process is optimized.
The motivation of the project is to demonstrate the potential of AI-driven design and fabrication processes in creating user-centered, cost-effective, and sustainable buildings. The preliminary study will create a methodology that can be used in both teaching and practice.
Through the methodology, PAZ-Parametric Academy Zürich GmbH will be supported on the one hand in the development of course offerings that demonstrate the application possibilities of AI-supported platforms and tools in the construction industry and teach the corresponding workflow. The first offers in this area are shown in Fig. 2. On the other hand, the methodology supports PAZ in construction projects that require complex geometric planning (Fig. 3).
At a workshop with potential implementation partners, the findings of the designed methodology will be critically reflected and questions identified with the aim of gaining further partners for the envisaged Innosuisse project.
