Researchers from Tsinghua University have found that AI-based urban planning is significantly more effective than human-created ones. Yu Zheng and his team tested whether a machine learning model could create a space adhering to the concept of a 15-minute city. As a result, the AI program developed a plan in seconds that took human planners 50 to 100 minutes to complete.
Our public spaces determine our individual lifestyles, such as our access to health care and groceries. However, designing cities to provide a high-quality community becomes increasingly difficult as our population grows. Fortunately, artificial intelligence could speed up and improve urban planning to provide better living spaces in the future.
This article will discuss the latest AI urban planning study from Tsinghua University. Next, I will explain other benefits of using artificial intelligence to shape living spaces.
How did an AI plan a city?
Tsinghua University researchers were looking for new solutions to improve cities that are becoming more congested and less green. Therefore, they created an AI program that could take care of the more tedious and computational aspects of urban planning. This is how the system worked:
- Yu Zheng and his colleagues tasked their model with designing urban spaces a few square kilometers wide, or about 3 × 3 city blocks.
- They trained the AI system using numerous neural networks after two days of training.
- Next, the AI planner searched for ideal road layouts and land uses that fit local planning policies and the 15-minute city concept.
- The latter consists of facilitating access on foot to essential places in the city such as offices and supermarkets.
Yu Zheng and his colleagues found that their model only needed a few seconds to complete tasks that took humans between 50 and 100 minutes. Additionally, their AI-assisted workflow increased access to core services by 12% and fleets by 5%.
However, they admitted that their program had flaws. For example, the model could plan large urban areas but could not design an entire city. Drafting the latter with 4×4 blocks requires twice as many planning decisions as 3×3.
You may also like: Air quality is a problem for cities around the world
However, automating certain planning steps could save a lot of time. Zheng said automating time-consuming tasks could allow planners to focus on more human-centered tasks.
ScienceAlert says they could focus on public engagement, aesthetics and other difficult tasks. This is why the team wants its AI system to serve as an assistant to urban planners.
The program could generate designs and optimize them with algorithms. Then, human experts could review, adjust, and evaluate them based on community feedback. The website states that the last step is an integral part of proper design.
How can artificial intelligence improve smart cities?
Chris Chiancone, Carrollton City’s Chief Information Officer, discussed how AI tools could change the way we plan for smart cities. For example, it could simulate multiple versions of a city to meet multiple requirements.
Your AI urban planner could create a version that prioritizes natural parks or improves accessibility to public amenities like schools and hospitals. Additionally, the program can predict how a city will grow and transform over time.
For example, it will show areas that will see an increase in population and additional buildings. As a result, these forecasts can help urban planners make more informed decisions regarding infrastructure design and resource allocation.
You May Also Like: Japan Announces Dogen Floating City
Artificial intelligence could also help improve existing cities. For example, an AI tool could check various data to suggest where to build new roads, expand public transportation, or determine the location of new public spaces.
Chiancone believes that generative AI, the artificial intelligence behind ChatGPT, could detect correlations and patterns that human planners may miss. For example, it could determine a link between the location of public parks and the health of nearby residents.
Believe it or not, we’ve done it before concrete examples of AI urban planning. For example, Amsterdam and Singapore have used generative AI to simulate new urban scenarios.
Researchers at Tsinghua University tested whether an AI program could plan cities. They discovered that this technology had potential but needed more research and development.
Given the rapid advancements in artificial intelligence, AI urban planning could become a global standard sooner than you think. Therefore, future urban areas could become more livable than existing ones.
Learn more about this Tsinghua University study on Nature Computational Science website. Also, discover more digital tips and trends at Inquirer Tech.
To subscribe to APPLICANT MORE to access The Philippine Daily Inquirer and over 70 other titles, share up to 5 gadgets, listen to news, download as early as 4 a.m. and share articles on social media. Call 896 6000.
For comments, complaints or inquiries, Contact us.