A stated preference survey for designing the physical elements of communities

Developing designs for the physical elements of communities involves both complexity and trade-offs. The human computer interface and generative design (GD) potentially address this challenge and present an opportunity for improved efficiency as compared to designing based solely on a human planner. The purpose of this paper is to design and test a repeated Stated Preference (SP) survey, as part of a broader GD decision-support process, that obtains feedback on land use distribution and road network design from stakeholders. The results from the SP survey are quantified into binary logit model parameter estimates that are used as inputs for subsequent rounds of the GD process. This iterative process allows for the improvement of community designs with each round of consultation until a desired design outcome is reached. This use of SP survey methodology within a GD process to produce community designs that reflect the goals of the stakeholders, is novel. This paper provides a demonstration and preliminary application of the method.