AI for Greener Cities: Tackling the Low-Hanging Fruit of Climate Change

Artificial intelligence (AI) is revolutionizing a diversity of sectors before our eyes. From more efficient healthcare to helping researchers unravel some of physics’ greatest mysteries, AI is used to create digital tools for solving complex problems. But how will AI help society to tackle one of our most existential challenges yet — climate change?

Though AI applications for combating climate change have so far received less funding and focus than more commercially-oriented applications in marketing, finance, and healthcare, the tide is turning (or, more precisely, the sea levels are rising). Governments, tech companies and investors are turning their sights toward the growing climate change AI sector.

One of the biggest areas of impact for climate change AI is the building and construction industry. AI tools can play a role in designing efficiently and for efficiency, and optimizing energy output and consumption. AI-powered approaches to green building and construction can be used both to improve existing infrastructure and design tomorrow’s cities, making a powerful case for priority investment.

Building Future Cities

Curbing greenhouse gas (GHG) emissions from the building sector is a major challenge to meeting the 2030 climate targets. Buildings account for about 40% of all U.S. energy consumption and a similar proportion of greenhouse gas emissions. With global populations continuing to make the move from rural to urban areas, the construction of new buildings is not expected to slow any time soon.

According to an expansive 2019 paper from industry and academic researchers, Tackling Climate Change with Machine Learning, buildings are “low hanging fruit” when it comes to reducing GHG emissions. The authors state:

“While the energy consumed in buildings is responsible for a quarter of global energy-related emissions, a combination of easy-to-implement fixes and state-of-the-art strategies could reduce emissions for existing buildings by up to 90%.”

AI is expected to play a critical role in designing new materials to build future cities in a more sustainable way and making smart buildings even smarter.

Quitting Carbon-Intensive Materials

The most fundamental piece of low-carbon construction is building material. The cement and steel industries are two of the most polluting industries in the world, yet we rely on these products to create nearly all built environments today. Taken together, the production of cement and steel account for over 10% of all global GHG emissions. The problem is the high levels of heat — known as industrial heat — required to forge these materials.

To put this impact into perspective, the global emissions of industrial heat are greater than the CO2 emissions of all the world’s cars and planes combined, and the cement industry alone emits more GHGs than every country apart from the US and China.

Emissions aside, the extraction required to create critical materials like steel and rubber has led to destruction of the natural environment, displacement of human and animal populations, and even slave labor. The discovery of new materials has undoubtedly shaped human society and our relationship to the natural world; the products we’ve created with them have played a major role in our civilization’s progress. But what if new materials could be simulated and created in a laboratory setting rather than extracted from the ground?

Luckily, alongside other advances in material science and technology such as 3D printing, AI holds promise for transforming these high-carbon industries in a more ethical way.

One potent way that AI can help to reduce the need for carbon-intensive materials is by assisting researchers in the development of novel structural materials. AI can be used to predict chemical reactions in order to simulate molecular synthesis and determine the suitability of a potential candidate as a construction material. Databases like the Materials Project can support the development of this machine learning application, helping researchers to discover and create low-carbon materials in place of steel and concrete.

Another way AI can help is in limiting the excess of carbon-intensive materials used to create building components. Generative design is a technique in engineering that uses AI to optimize building designs according to the goals and parameters set forth by an engineer. Particularly when paired with 3D printing, generative design can facilitate the creation of components that can be printed to the exact sizes and shapes required for a project. Structural products can thus be developed with a reduced amount of cement and steel to the same, or even better, effect — making for more efficient building materials.

Making Buildings and Cities Smarter

It’s not only what you see on the outside of buildings that makes them unsustainable, but also the systems within them. The rarely-considered systems that control things from the temperature to the lighting within buildings — including HVAC systems, water chillers, air compressors — can be incredibly energy inefficient. They often use far more energy than they need to. Historically, this was due to our inability to tailor the output of these systems to the changing needs of building’s inhabitants; specific data that could inform more efficient use of resources was not tracked and control systems were not equipped with the capacity to respond to changing indicators.

Today, enabling technologies like the IoT (made up of smart devices and sensors) can provide the granularity of data required to make energy-saving adjustments. But the data is only so helpful without a system that can collect and analyze it, make decisions and then communicate them to control systems. AI is a critical piece of the intelligent building puzzle.

Intelligent control systems can not only decrease the carbon footprint of a building by reducing the energy consumed but can also provide the means to incorporate low or zero carbon energy alternatives. As described in our last post, The Energy Transition Needs AI, integrating renewables into the grid is another important use case for AI.

In order for systems, from a building-wide to a city-wide scale, to make the green transition they will need AI.

Nowhere has the government taken such a strong initiative in intelligent urban planning as in China. According to Reuters, as of two years ago, over 500 smart cities were being built in China. These modern urban centers are “equipped with sensors, cameras, and other gadgets that can crunch data on everything from traffic and pollution, to public health and security.” A major benefit of smart cities is environmental; sustainability can be a bottom-up design approach.

Limitations and a Long-View

Intelligent buildings and cities also carry a host of thorny ethical dilemmas. Surveillance, data capture and data privacy are serious concerns. Likewise, AI programs that can run things like molecular simulation for discovering novel building materials require a huge amount of energy. Data processing and storage remain serious hurdles for scalable, AI-powered solutions: data processing centers consume around 2 percent of the global electricity supply. As more data and tools to use it are created, that number will continue to rise.

AI-powered technology is a powerful tool in the fight against climate change, but it is not a complete solution. This fight against time will require innovative solutions as much as a breaking with old ways. Ultimately, it will require a shift in human perspective that radically repositions our relationship to the natural environment. That’s a challenge that machines — no matter how intelligent — won’t be able to help us with.

Fighting climate change will be a battle on diverse fronts, and AI is one advantage we have in our global toolkit. Understanding both the advantages and limitations of every tool is fundamental to using it effectively. With AI, strategic investment in impactful areas of climate change will help us to get to where we need to be. The grass is greener in tomorrow’s cities, but only if we plant the seeds today.