From Vision to Reality: Artificial Intelligence Design Assistant (AIDA) to facilitate Collaborative Design Processes

Research Project, 2024 – 2026

The failure to identify building design flaws in the design stages of hospital projects has been shown to be a significant cause of costly remodelling and non-functional operations and buildings.
The building design has a direct impact on the professional operation and their ability to effectively perform their daily tasks. It is therefore important to capture knowledge and take into account the professional operation´s feedback and involvement in connection with the design of their future premises. This is often done today with the help of abstract descriptions, 2D drawings and images which can be difficult to interpret and understand and can lead to different interpretations and lack of communication and decision-making. Which results in the professional operation´s understanding of and requirements for the future premises becoming deficient and thereby not being communicated to consultants and contractors effectively. Additional ways to involve the professional operation are by building expensive ful-scale demo rooms or using visualization tools such as Virtual Reality (VR) as well as the use of standardized guidelines and functional requirements.In this R&D project, the focus will therefore be on implementing, studying and investigating how VR and the use of guidelines, functional requirements can be implemented in the design process. The project originates in the R&D project "Collaborative information structure", which studied how co-virtual environments together with information structures, guidelines and functional requirements linked to the building design can support co-creation and more efficient and better processes. During the R&D project, a co-design process and a tool based on the Virtual Collaborative Design Environment (ViCoDE) was developed. ViCoDE is a co-design and planning tool based on building information models (BIM) and VR in combination with a touch screen and projector screen. Users can interact and modify the virtual model of the construction project together during collaborative workshops. The result is that scenario-based co-design and review could be achieved, thereby allowing the end-users to become active and co-determining in decision-making, but also that a more efficient and better design process can be created with shorter lead times in the design review and planning.However, the results from the case studies showed that the participating participants did not come prepared and also expressed the difficulty of adopting the latest "best practices" and guidelines related to the design of buildings. Nevertheless, they expressed that VR and scenario-based co-design and review together provided a good and efficient process where knowledge exchange and many building design errors were discovered together. Participants also expressed the difficulty of ensuring whether the virtual environment is being reviewed based on the right questions and scenarios.This R&D project has that as a starting point and wants to investigate and further develop whether AI can be used as a user interface to support the users, by helping the participants with scenarios and questions linked to guidelines and "best practices". The project will further develop ViCoDE and implement AI-driven projects and personal assistant, adapted to revolutionize the process of how future building design is carried out with the help of VR, AI and information structures. By integrating a project and personal AI-assistant in the VR environment, the project intends to enable and create a user interface that makes available "best practices" and the knowledge base in how school environments and hospital environments should be designed based on what the user experiences in VR. By using a personalized and interactive user assistance in the design and review process of the premises, this will ultimately lead to a better and more efficient design process, which improved teaching and learning environment and optimized care environment and patient car