Keywords: Computational Urban Design, Human Perception, Optimization, Built Environment, Wellbeing
In most urban areas there is a need for densification. The densification of existing urban areas comes along with risks and influences many different stakeholders and aspects within the existing urban system. In order to manage the influences of potential (re)development projects in an urban context, insight on the impact of potential (re)development projects is needed in an early stage of the development process. Computational Urban Design enables fast generation of conceptual urban designs in an existing urban context. These designs can be optimized to align to certain design aspects and the influence of the generated urban designs on multiple design aspects can be calculated and visualized easily. However, current computational urban design tools do not prov ide insight or include all relevant design aspects in an existing urban environment. Since cities are built to facilitate the life of humans, the wellbeing of humans can be considered as an important design aspect. Yet, wellbeing of humans is not incorporated in computational urban design. One of the aspects influencing the wellbeing of humans that is not yet incorporated in computational urban design is human perception. This research demonstrates how human perception can be incorporated in computational urban design. Within this research, first the relation between the built environment and human perception is analysed and quantified using a big data approach, including stated choice data and a multinomial logit analysis. From the analysis it was among others found that the presence of trees, the dimensions of building volumes and the urban morphology influences human perception. The quantified relationships between human perception and the built environment have been implemented in computational urban design by creating an extension on an existing parametric urban design methodology. This extension enables parametrically designed urban designs to be analyzed and optimized on human perception.
As a result of this research, a first methodology has been described and tested that enables the incorporation of human perception in computational urban design. The most important considerations for future research should be to increase the accuracy of the quantified relation between the built environment and human perception. In relation to the applicability in current practice, future works could focus on improving the technical capabilities of the computational urban design methodology by increasing the design freedom, the design generation speed and the comprehensiveness by including more design aspects.