An evaluation framework for AI powered generative concept design in the real estate site planning process

An evaluation framework for AI powered generative concept design in the real estate site planning process

Smit, Menno (TU Delft Civil Engineering & Geosciences)

van Bueren, Ellen (mentor)
Vrijhoef, R. (graduation committee)
Wang, T. (graduation committee)

Delft University of Technology

Civil Engineering | Construction Management and Engineering

2022-11-25

Due to the stricter sustainability requirements and associated governmental regulations, the real estate development process in the Netherlands is becoming increasingly complex, narrowing the real estate developer’s opportunity space and increasing the uncertainty regarding the feasibility of the real estate development project. Recently developed AI powered generative design tooling could potentially be used to decrease this uncertainty during the concept design process. However, there is currently no scientific method that helps the (concept) developer evaluate the concept design with the use of this AI powered generative design tooling and make decisions based on more objective design criteria instead of estimated guesses. The goal of this research is to develop an evaluation framework for the concept design process that can be used by real estate development professionals in the Netherlands, during the concept design stage of inner-city apartment buildings, to make more informed choices (based on additional objective information) earlier on in the design process to decrease the uncertainty with regards to the feasibility of the real estate development project. The framework will be developed considering the current visualisation and simulation abilities of AI powered generative design tooling.

Using the concept design evaluation matrix, the real estate developer and other relevant stakeholders are encouraged to 1) carefully determine what they consider important for the specific real estate development project, taking into account the site-specific characteristics, prior to the evaluation and decision-making process. 2) By subsequently focusing on these priorities during the location/market investigation and concept design process (for example by focusing on identifying any uncertainties and/or risks related to environmental factors such as noise pollution), the expert panel expects that the developer 3) can better steer the process because the different interests are clear and the developer can better evaluate the results with stakeholders by using the analysis results and ranking scores of the different parameters in negotiations/discussions. The expectation is that the evaluation matrix combined with the use of AI powered generative design tooling will decrease the uncertainty regarding the feasibility of the real estate development project, and will provide the developer with a handle to ultimately achieve an even better real estate development result.

generative design
real estate development
urban planning
concept design development
decision-making

http://resolver.tudelft.nl/uuid:863142af-edee-400c-8171-06e26725b8e9

Student theses

master thesis

© 2022 Menno Smit