Imagine living in a cool, green city filled with parks and sidewalks, bike lanes, and buses that transport people to shops, schools, and service centers in minutes.
That airy dream epitomizes urban planning, embracing the idea of a 15-minute city, where all basic needs and services are within a quarter of an hour’s reach, improving public health and reducing vehicle emissions.
Artificial intelligence can help urban planners realize that vision faster, a new study by researchers at Tsinghua University in China shows how machine learning can create more efficient spatial layouts than humans, and in less time.
Automation scientist Yu Zheng and colleagues wanted to find new solutions to improve our cities, which are rapidly becoming congested and compacted.
They developed an AI system to handle the most tedious, computational tasks in urban planning—and found that it produced urban plans that outperformed human design by about 50 percent on three metrics: access to services and green space, and traffic levels.
Starting small, Zheng and colleagues tasked their model with designing urban areas of just a few square kilometers in size (about 3×3 blocks).
After two days of training, and using multiple neural networks, the AI system found the ideal street layout and land use to match the 15-minute city concept and local planning policies and needs.
While Zheng and colleagues’ AI model has some features to expand its use in planning large urban areas, designing an entire city would be more complicated. Drafting a neighborhood with 4×4 blocks involves twice as many planning decisions as 3×3 blocks, researchers estimate.
But automating some steps in the planning process can save a lot of time: an AI model calculated some tasks in seconds that would have taken human planners 50 to 100 minutes.
Automating the most time-consuming tasks of urban planning will free up planners to focus on more challenging or human-intensive tasks, such as public engagement and aesthetics, researchers say.
Instead of AI replacing people, Zheng and colleagues envision their AI system to act as an ‘assistant’ for urban planners, creating concept designs optimized by algorithms and reviewed, adjusted and evaluated by human experts based on community feedback.
This last step is central to good design, writes Paolo Santi, a research scientist at the Massachusetts Institute of Technology (MIT), commenting on the study.
Urban planning is “not just allocating space for buildings, parks, and functions, but designing the spaces in which urban communities live, work, interact, and, hopefully, thrive over the long term,” he writes.
Comparing their human-AI workflow to a human-only design, Zheng and colleagues found that the collaborative process could increase basic services and parks by 12 and 5 percent, respectively.
The researchers also surveyed 100 urban designers, who did not know whether the plans they were asked to choose from were created by human planners or AI. AI garnered more votes for some of its spatial designs, but for other plans, there was no clear preference among survey participants.
The real test, of course, will be in the communities created by those plans, measured by the reductions in noise, heat and pollution and improvements in public health that better urban planning promises to bring.
This study has been published in Nature Computational Science.