Abstract
Innovation districts can spur urban economic growth but present challenges like socio-spatial segregation. To mitigate segregation, we emphasize the importance of “staying activities” such as sitting, playing, and relaxing, over transient activities like walking or cycling in the design of public spaces. Focused on One-north, an innovation district in Singapore, we examine how different spatial qualities influence the “staying activities” of diverse socioeconomic groups. Using a GeoAI workflow that integrates various data-sources with machine learning and AI, we assess thirteen spatial quality indicators and their impact on staying activities. Our deep neural network (DNN) model, which achieves an 83.74% accuracy rate, confirms the effectiveness of these indicators in capturing the co-location of staying activities between diverse income groups in One-north. Our findings reveal that typical markers of well-designed public spaces, such as visual diversity and greenery, do not consistently enhance staying activities. Their effectiveness varies across One-north and is shaped by the unique functional contexts and land-use types of each sector in the district. Our research thus highlights the necessity for context-specific designs that encourage meaningful social interactions and reduce socioeconomic segregation. Our proposed GeoAI workflow can assist in the proactive evaluation of design strategies, supporting decision-making prior to major construction projects.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call