"AI discovers paint formula to cool buildings more effectively"

AI-Developed Paint Could Help Cool Cities and Lower Energy Use, Study Finds

Researchers have found that a new type of paint, designed using artificial intelligence, may help reduce excessive urban heat and lower the need for air conditioning in cities.

Experts in material science developed coatings that can keep buildings between 5°C and 20°C cooler than conventional paint when exposed to direct sunlight. These coatings could also be used on vehicles, trains, electrical equipment, and other structures that require cooling as global temperatures rise.

According to a study published in the journal *Nature*, scientists from institutions in the US, China, Singapore, and Sweden used machine learning to create paint formulations optimized to reflect sunlight and emit heat efficiently. This approach is part of a wider trend in which AI is speeding up scientific discovery by surpassing traditional experimental methods.

In a separate case last year, a UK-based company, MatNex, employed AI to develop a new permanent magnet for electric vehicle motors, eliminating the need for rare earth metals, which entail carbon-heavy mining processes. Elsewhere, tech firms have introduced AI tools to assist in designing novel materials, including crystalline structures used in solar panels and medical devices. Similar efforts are underway to improve carbon capture and battery efficiency.

The paint study involved researchers from the University of Texas at Austin, Shanghai Jiao Tong University, the National University of Singapore, and Umeå University in Sweden. Their analysis showed that coating the roof of a four-story building with one of the AI-designed paints could save up to 15,800 kilowatts of electricity annually in hot climates like Rio de Janeiro or Bangkok. Scaling this to 1,000 buildings could provide enough energy savings to power over 10,000 air conditioning units for a year.

Yuebing Zheng, a professor at the University of Texas and co-leader of the research, stated: “Our machine learning system marks a major advancement in thermal material design. By automating the process, we can explore new possibilities and achieve performance levels that were previously out of reach.”

He added that AI allows researchers to develop materials in days rather than the weeks or months required by conventional methods. Previously unforeseen compositions can now be tested based on computational predictions, avoiding lengthy trial-and-error phases.

“With machine learning, we follow its guidance on material selection and structure. This means fewer rounds of physical testing and faster results,” Zheng explained.

Alex Ganose, a chemistry lecturer at Imperial College London who also works on AI-assisted material design, noted the rapid progress in this field: “The pace of innovation is remarkable. Recently, numerous startups have been exploring generative AI for material science applications.”