In a recent study, researchers outline a proposal to minimize urban planners’ workloads using deep learning systems to perform some of their basic tasks.
The study noted that traditional urban planning can be both laborious and time-consuming; it’s possible AI can be used to automatically improve existing urban planning solutions.
The team created an adversarial learning framework known as LUCGAN to provide effective land-use configurations by focusing on human mobility, urban geography, and socioeconomic data.
From the study:
Effective urban planning can help to mitigate the operational and social vulnerability of a urban system, such as high tax, crimes, traffic congestion and accidents, pollution, depression, and anxiety. Due to the high complexity of urban systems, such tasks are mostly completed by professional planners. But, human planners take longer time. The recent advance of deep learning motivates us to ask: can machines learn at a human capability to automatically and quickly calculate land-use configuration, so human planners can finally adjust machine-generated plans for specific needs?
In order to generate a suitable and excellent land-use configuration solution objectively and reduce the heavy burden of urban planning specialists, we proposed an automatic land-use configuration planner framework. This framework generates the land-use solution based on the context embedding of a virgin area. Specifically, we obtained the residential community and its context based on the latitude and longitude of residential areas firstly. we then extracted the explicit features of the context from three aspects: (1) value added space; (2) poi distribution; (3) traffic condition. Afterward, we mapped the explicit feature vectors to the geographical spatial graph as the attributes of the corresponding node. Next, we utilized the graph embedding technique to fuse all explicit features and spatial relations in the context together to obtain the context embedding. Then we distinguished the excellent and terrible land use configuration plans based on expert knowledge. Finally, the context embedding, excellent and terrible plans were input into our LUCGAN to learn the distribution of excellent plans. The LUCGAN can generate a suitable and excellent land-use configuration solution based on the context embedding when the model converges. Ultimately, we conduct extensive experiments to exhibit the effectiveness of our automatic planner.