Speaker
Description
The development and maintenance of built environments require continuous decision-making processes aimed at improving current conditions. By analyzing on- and off-site contexts, we have delivered outcomes of where and to what extent to develop, redevelop and refurbish built environments. With the continuous improvement of computational technologies such as geographic information systems and artificial intelligence, we have been able to efficiently and effectively identify relevant characteristics of built environments.
This presentation focuses on two recent studies conducted in my lab, the KAIST Urban Design Lab. The first study involved the development of new methods for predicting disaster risks in declining small urban areas. By analyzing present disaster risks and using climate change scenarios, we predicted changes in disaster risks in declining small urban areas. We collected risk analysis factors such as exposure, vulnerability, and mitigation capacity for eight hazards and calculated the risk of these hazards at present. By using climate change scenarios, such as RCP 4.5 and 8.5, and analyzing relationships between hazards, we predicted the future risk of these hazards. With the developed spatial database, we then developed a decision support system that visualizes disaster risks in small declining urban areas.
The second study aimed to visualize conceptual land use in a given urban block by developing a generative AI method. We used an AI model trained to comprehend forms, land uses, and density in urban blocks using a pix2pix approach, with data such land uses and densities collected from a small urban block in Seoul, South Korea. We established image datasets that included spatial information in urban blocks and developed an AI advisor for conceptual urban land use planning. This advisor generated and visualized land uses and densities, with colors and heights representing given urban contexts. This step was a significant advancement in the adoption of AI approaches in urban planning.
By utilizing advanced technologies, we have endeavored to help people make efficient and effective decisions to make cities better for humans.
References
Park, C., NO, W., Choi, J., Kim Y. (in press) Development of an AI Advisor for conceptual land usen planning, Cities, forthcoming.
Byun, G., NO, W., Park, C., Lee H.-K., Kim Y. (2022) Predicting the Potential of Rainfall Disaster Risk using Cellular Automata in Small Urban Declining Areas, KIEAE Journal, 22(6): 29-35. https://doi.org/10.12813/kieae.2022.22.6.029
Byun, G., NO, W., Park, C., Lee H.-K., Jang K.M., Kim Y. (2021) Analysis of the Disaster Risk Assessment for Urban Declining Areas by Applying Weights of Risk Indicators: Focused on Heavy Rainfall and Snow, KIEAE Journal, 21(6): 23-30. https://doi.org/10.12813/kieae.2021.21.6.023
Park, C., Byun, G., Lee, H.-K., Jang K.M., NO, W., Kim Y. (2020) Development of a Risk Assessment Model of Rainfall for Small Area in Declining Urban Areas, KIEAE Journal, 20(6): 7-12. https://doi.org/10.12813/kieae.2020.20.6.007
| Keywords | Urban regeneration, spatial analysis, urban analytics, smart cities, GIS, AI |
|---|
