Architectural design variable assessment in multi-objective life cycle optimisation

A stakeholder perspective

AI generated image showing lego stakeholders around an architectural model

Architectural design variables in early design stage have the largest impact on life cycle performance and cost. Defining crucial variables is the first step and a significantly important factor in developing the machine learning based optimisation program. However, most researches do not specify the reason of the selection of initial variables, the important variables in early design stage are still unknown. Therefore, the performance of developed programs’ outcome could not be guaranteed.

Early design stage always involves multiple optimisation objectives and the decision of multiple variables. There is a need for developing a method to quantify the impact of a range of ADVs on project, and to also be able to evaluate and compare the importance of different optimisation goals. There are many existing optimisation methods in architecture academia, however, few of them have been applied in practice.

The main reason of this is the lack of stakeholders’ engagement. The developed tool could be hard for architects to use, and the initial ADVs defined in the tool could not be the ones architects want to play with in the early design stage. This project aims to identify the important architectural design variables in early design stage for life cycle optimisation by interacting with stakeholders who have major influence in early design stage.

Xinyue Wang
Xinyue Wang
PhD student

My research interests include life cycle assessment, life cycle cost, building optimization and machine learning.

Alexander Hollberg
Alexander Hollberg
Associate Professor

Alexander Hollberg is Assistant Professor in the Division of Building Technology, at Chalmers.