Gaby
AI can help urban planners turn land-development visions into reality far faster than people can, according to Chinese researchers. Their study shows how machine learning can produce more efficient spatial designs than human planners while using far less time.
AI Plans Developments More Efficiently and Quickly Than Humans
AI can quickly lay out the 15-minute city ideal: green spaces, pedestrian and bike paths, and buses that connect people to shops, schools, and service centers. The work took mere minutes to sketch this idea, arranging everything so essential needs and services are within a 15-minute walk, which promotes public health and cuts vehicle emissions.
Automation expert Yu Zheng and his colleagues set out to find new solutions for rapidly growing, densely built cities. They developed an AI system to handle the most exhausting computational tasks in urban planning. The system produced urban plans that outperformed human designs by about 50% across three metrics: access to services, access to green space, and traffic levels.
AI Saves Significant Time in City Planning
Starting small, Zheng and his team tasked their AI model with designing urban areas a few city blocks in size (roughly a 3-by-3 block area). After two days of training and using several neural networks, the AI system found ideal layouts for roads and land use that aligned with the 15-minute city concept and local planning goals.
The model the researchers built includes features that could scale up for larger planning tasks, but designing an entire city would be much more complex. The team estimates that planning a neighborhood of 4-by-4 blocks requires about twice as many decisions as a 3-by-3 block area.
Still, automating even a few steps in the planning process can save a huge amount of time. The AI model completes certain tasks in seconds that would take human planners 50 to 100 minutes. In an interview with ScienceAlert, the researchers said automating the most labor-intensive tasks will free planners to focus on more complex, human-centered work like community engagement and aesthetics.
Not Full Automation, But Collaboration Between AI and Humans
Zheng’s team says AI won’t replace humans but will act as an assistant for urban planners. The system can generate concept designs optimized by algorithms, and human experts will then review, adjust, and evaluate those designs based on community feedback.
As Paolo Santi, a research fellow, points out, urban planning isn’t just about allocating space for buildings, parks, and functions; it’s about designing places where communities will live, work, interact, and thrive over the long term.
By comparing workflows, Zheng and his colleagues found that a collaborative process could increase access to essential services by 12% and access to parks by 5%. The researchers also surveyed 100 urban designers who didn’t know whether the plans they reviewed were made by humans or by AI.
In that survey, AI received significantly more votes for some spatial designs, while for other plans designers showed no clear preference. The researchers say the ultimate test will be the neighborhoods built from these plans, judged by reductions in noise, heat, and pollution and by improvements in public health that better urban planning aims to deliver.
