It is known that the planet is warming, increasing the need for cooling buildings – even in latitudes where air conditioning has been the exception rather than the norm. At the same time, energy in the form of fossil fuels or electricity is becoming more and more expensive, all the more so since humanity should consume less of it anyway in order to reduce greenhouse gas emissions. So what can you do to keep your home and workplace pleasantly fresh? Even simple window insulation plays a central role here – and this is where artificial intelligence and machine learning should help.
Better cooling and a clear view
A US-South Korean research group at the University of Notre Dame in Indiana and Kyung Hee University near Seoul wants to use this to solve a central problem: so-called transparent radiant coolers can be used as window materials to reduce the energy demands of cooling buildings and even reduce the number of vehicles . “However, it is difficult to achieve high visible transparency and high radiative cooling performance at the same time.” This means that if you want to use the coating to keep enough heat inside your house or vehicle, it gets dark too quickly. And you don’t want that either, because then energy would have to be used for lighting.
The answer of scientists around Eungkyu Lee and Tengfei Luo and Co. is the so-called TRC, transparent radiation cooler. It is a thin layer of alternating layers of different materials such as aluminum oxide, silicon dioxide, silicon nitride or titanium dioxide that are applied to the glass. The last layer is polydimethylsiloxane. Difficulty: In order to cool equally well, i.e. reflect invisible infrared and ultraviolet sunlight, and radiate heat while transmitting visible light, the correct composition must be determined – including the amount and thickness of the layers. And that can only be calculated in the experiment with great effort.
30 percent less cooling energy
The appearance of a fellow computer: Lee and colleagues created software that uses machine learning to calculate the sequence and combinations of layers needed. Since it took too long for conventional systems, algorithms from the field of quantum computing were also used – and this is a relatively new development in this work. The result was the TRC, which according to the researchers significantly outperformed commonly produced coatings and could compete with the best commercial glasses.
Compared to conventional windows, the need for cooling energy is reduced by about 30 percent in the summer. The next step is to explore how mass production could be increased. The optimized structures “can potentially be scaled up for practical applications because they can be fabricated using state-of-the-art deposition techniques,” the team says.