EXPLORATION OF POTENTIAL USE OF ARTIFICIAL INTELLIGENCE IN URBAN DESIGN
Machine Learning in Urban Design
Urban design is a complex and time-consuming work. Nowadays, the artificial intelligence (AI) is developing rapidly, especially machine learning. Deep generative model, an approach of machine learning has shown the ability to learn from and interpret data to generate 2D and 3D designs (Creswell et al., 2018). Some researchers have explored the application of deep generative model in urban form generation. However, most urban form generation approaches assisted by deep generative model of machine learn rely on pixel, ignoring the overall topological information among streets, blocks and buildings.
This research aims to develop a framework of machine learning aided urban form generation based on topology to improve the creativity and efficiency of urban design. There are six parts in this research. Firstly, this research reviews urban form generation, AI and machine learning in generative urban design. Secondly, a survey is conducted to understand the user acceptance of professionals, academics and students in the field of architecture and urban design towards deep generative model. Thirdly, A framework of machine learning aided urban form generation based on topology is proposed. Fourthly, machine learning aided street network generation based on topology is explored. Fifthly, machine learning aided building configuration generation based on topology is explored. Sixthly, a case study is conducted to validate the developed machine learning aided urban form generation framework based on topology.
He is an architect and a lecturer in Wenzhou University, China. Currently, he is a PhD candidate at the Welsh School of Architecture. He holds a Bachelor degree in Landscape Architecture from Huazhong Agricultural University, China; a Bachelor degree in English from Huazhong University of Science and Technology, China; a Master degree of Landscape Architecture from Huazhong Agricultural University, China; and a MSc degree in Planning Design and Development from University College London, UK. He is interested in the application of artificial intelligence in architecture and urban design.