Published Date: 29/07/2024
Municipalities in Quebec's capital region are pioneering the use of artificial intelligence to monitor various urban features, from tree cover to cars and even backyard pools. The Communauté métropolitaine de Québec, a grouping of Quebec City and its suburbs, aims to utilize this technology to achieve environmental targets, assess parking availability, and track urban development.
The organization has trained a deep learning model on high-definition aerial photos of Quebec City and the surrounding region taken in 2021. This AI model can identify and highlight various features, including buildings, trees, vehicles, swimming pools, backyard trampolines, and jungle gyms. The performance of this model is equivalent to that of a human, but with much greater speed, enabling it to process a large amount of work in a short time.
The data obtained can be used in different ways. For instance, Quebec City and surrounding municipalities can measure progress towards their urban greening and tree cover targets. Conversely, it can also track how much green space has been converted to asphalt over time. Municipalities can also utilize the images to determine parking availability in different areas. Moreover, tracking backyard pools can help cities decide where to send inspectors.
A new set of aerial photos is being taken this summer, and Frédérick Lafrance, geomatics development manager at CMQuébec, hopes to analyze them this winter. He also plans to apply the technology to photos taken as far back as the mid-20th century to assess the region's development over time.
Renee Sieber, an associate professor of geography at McGill University, notes that municipalities across Canada are already using AI in various ways. Edmonton, for example, has used AI to monitor wildlife entering the city, while Montreal and Toronto have experimented with AI to reduce traffic congestion.
However, Sieber cautions that some applications of AI raise more concerns than others. She emphasizes the importance of clearly defined objectives and transparency to avoid 'mission creep.' In the case of Quebec City, the aerial images already exist, and the AI is not revealing new information that couldn't be tracked manually.
The possibilities for municipalities using AI are vast. Google has used a combination of aerial images and AI to estimate tree cover in hundreds of cities worldwide. Richard Khoury, a computer science professor at Université Laval, suggests that the Quebec data could be used to assess property values and target less-developed areas for urban development projects.
However, AI has the potential to reveal more than just urban features. A 2017 U.S. study used deep learning to identify the make and model of cars in 50 million images from Google Street View. The researchers found that cities with more sedans than pickup trucks tended to vote Democrat, while cities with more pickups than sedans tended to vote Republican.
Municipalities need to consider how the public will respond to their use of artificial intelligence. As Sieber notes, 'No tool is neutral,' and social acceptance is crucial. Communauté métropolitaine de Québec (CMQuébec) is a metropolitan community that groups Quebec City and its suburbs. McGill University is a public research university in Montreal, Quebec. Université Laval is a French-language public research university in Quebec City, Quebec.
Q: What is the Communauté métropolitaine de Québec?
A: The Communauté métropolitaine de Québec is a metropolitan community that groups Quebec City and its suburbs.
Q: How does the AI model track urban features?
A: The AI model is trained on high-definition aerial photos and can identify and highlight various features, including buildings, trees, vehicles, swimming pools, backyard trampolines, and jungle gyms.
Q: What are the potential applications of AI in urban planning?
A: AI can be used to track tree cover, assess parking availability, monitor urban development, and estimate property values.
Q: What are the concerns surrounding the use of AI in urban planning?
A: Experts caution about the need for clearly defined objectives, transparency, and social acceptance to avoid 'mission creep' and ensure that the technology is used responsibly.
Q: How can AI be used to analyze historical data?
A: AI can be applied to photos taken as far back as the mid-20th century to assess the region's development over time.