Graduation project: Architecture Intelligence

STM Group - formerly known as Staalmeesters - has been a established name since 1996. Due to our rapid growth and (inter)national projects, we now operate under the name STM Group. Behind this new name is the same down-to-earth, ambitious team from Twente.

‍

Within the STM Group ecosystem, construction projects are realized from start to finish. In recent years, the architectural aspect of these projects has received increasing attention. Although STM Group is not a traditional architectural firm, the company actively considers design and architectural choices during the engineering process. The goal is to design, engineer, and build projects at a high pace, with the creation of a fully developed 3D model in BIM software (Revit) typically completed within a matter of weeks after the initial client meetings — ensuring an efficient and transparent process.

‍

With the rapid rise of Artificial Intelligence, STM Group sees opportunities to further speed up and optimize this design process. Specifically, the company is interested in exploring how AI — such as ChatGPT or similar AI models — can use predefined parameters and client requirements to automatically generate concept 3D models in Revit. By strategically implementing AI, concept designs can be produced more quickly, allowing design decisions to be visualized, evaluated, and discussed earlier in the project workflow.

‍

STM Group would like to investigate how Artificial Intelligence can be applied to accelerate and improve the early design phase of construction projects. The goal of this assignment is to explore how client preferences and project requirements — collected through a smart (digital) intake process — can be translated into parameters that AI (such as ChatGPT) can process to automatically generate a 3D model in Revit.

‍

This project will focus on developing a conceptual workflow in which:

• client data and project parameters are gathered through a structured digital intake,

• this data is automatically converted into AI prompts,

• and AI uses these prompts to generate a concept 3D model in Revit, which STM Group can further refine and build upon.

‍

Throughout the assignment, a prototype of this intake and design workflow will be developed, including a proof of concept for the AI-generated output in Revit. This will enable STM Group and its clients to visualize multiple design options in the early stages of a project and make better-informed, faster decisions.

‍

Possible research questions:

• How can AI tools like ChatGPT be used to automatically generate design concepts in Revit based on client preferences and project data?

• Which existing AI technologies and tools are best suited for automatically generating 3D models in Revit?

• What would an effective, smart digital intake process look like to efficiently capture and structure client wishes and project requirements?

• How can the results from this intake be translated into AI prompts and parameters for generating a Revit model?

• In what ways can AI-generated concept models be integrated into STM Group’s existing engineering software and workflows?

‍

‍

Want to apply or have questions? Feel free to reach out to Dunya via phone, WhatsApp, or email.

• Phone: +31 (06) 51 68 23 66
• Email: dw@staalmeesters.com