Artificial intelligence, or more commonly known as AI, has become more widely accessible and used amongst the public, but this is not without controversy. AI has gained a very bad reputation these days for the many bad hands it has fallen into. For example, AI art generators take away income from talented artists and ChatGPT is like the black plague for students in academia — something to avoid at all costs. Not to mention, there is the looming fear that AI will take away jobs. On the surface, AI brings a slew of troubles to people, however, UW has found a few ways to use AI to make lives better.
Buildings are at constant risk of airflow leakage. This is a problem because buildings need heat and moisture. If the air is escaping, then buildings will have to use more heat, resulting in energy loss. This makes for expensive maintenance of buildings, and this only gets worse when you take the ever-present issue of climate change into account. This is bad for residents of buildings too because of the higher electricity bills.
What did UW do to fix this problem? Researchers along with Mohamad Araji, director of UW’s architectural engineering program and head of the Symbiosis Lab, made a platform that detects where heat loss is occurring within buildings by measuring heat flow to pinpoint where the heat is escaping. It was made by AI, infrared technology, and a mathematical model. They tested this platform on a multi-unit residential building in the Canadian prairies, marked by its huge issue of heat loss combined with the climate. Using the platform, they found that there were 28 regions of massive heat loss, most of which occurred around wall intersections and windows. If 70 per cent of those regions were fixed, then 25 per cent of energy could be saved.
AI’s help doesn’t stop there as the platform was trained using AI tools to increase the accuracy of recognizing areas of heat loss. The platform was able to find areas of heat loss with 81 per cent accuracy which is almost 10 per cent more than humans. Human error has been reduced in this case and the speed of which this data could be obtained has increased.
Beluga whales, much like buildings, also struggle with climate change. The difference between the two is that beluga whales are also suffering from the effects of oil and gas development. Beluga whales are near threatened according to the World Wildlife Fund (WWF), which means that they are close to being endangered. Even so, they should be protected to prevent further loss of biodiversity. Research scientists like Marianne Marcoux monitor beluga whale health and populations by utilizing aerial surveys. Again, there are issues because these surveys produce detailed images that are susceptible to error because the whales are only a few pixels large and they’re only trackable in summer. They’re also time consuming and labour-intensive.
UW’s Vision and Image Processing Research Lab work with cancer cells and their methods could be applied to looking at beluga whales. They analyze digital images and identify beluga whales in them through the use of an AI-based algorithm developed by David Clausi, Linlin Xu, and Muhammad Patel. The AI was able to reduce the number of images taken without beluga whales by 95 per cent. This helped Marcoux complete image analysis in a summer as opposed to in one year. Xu believes this could be cut down further to eight hours with more training of the AI.
AI can be used to make lives better, whether it be human or animal. It’s a shame that AI is commonly being looked down upon due to improper uses. In a time when climate change is so destructive, I find it interesting that a modern tool like AI is helping mitigate the damage. The development of AI can be something to look forward to if it means that more problems, especially if related to climate change, will be solved.
