As one of the most widely used construction components in the world, cement is a source of around 8% of carbon emissions.
In an effort to help cut down on that burden, researchers at Facebook parent company Meta recently set out to develop a less energy-intensive formula for the concrete it uses in its data centers, tapping machine learning to optimize for sturdiness and sustainability.
The company claims the project, a joint effort with a team at the University of Illinois Urbana-Champaign, has now yielded a concrete-making method that’s 40% less carbon-intensive than the regional average of the Chicago-area data center where it’s been tested.
“Using the input data on concrete formulas along with their corresponding compressive strength and carbon footprint, the Al model was able to generate a number of promising new concrete mixes that could meet our stated data center requirements with a lower embodied carbon impact than the industry standard,” the study authors said in a statement.
The researchers trained an artificial intelligence model on a database of different concrete formulas and their corresponding carbon impact and strength. The algorithm was able to create a list of formulas, from which the team chose five to test and refine in their lab.
The study authors were eventually able to narrow down the list to one final recipe, which made use of cement substitutes like fly ash and slag to minimize carbon usage. Concrete is typically made of a mix of ingredients like cement, gravel and sand.
The researchers next tested the formula on a few noncritical structures at one of Meta’s data centers in DeKalb, Illinois. The material was able to meet strength requirements and cut construction energy consumption by around 40% of the regional average, the company claimed.
The experiment comes as big tech companies are facing more scrutiny over the energy consumption of their physical infrastructure, namely the vast tracts of servers they maintain in far-flung places around the world. Meta has set a goal to reach net zero emissions across its supply chain by 2030.
“Climate change is one of the biggest challenges we face,” Meta chief technology officer Mike Schroepfer said in a statement. “Delivering essential technologies and reliable climate information to billions of people is at the heart of how Meta can help address the crisis. And we believe we can do it with a net zero carbon footprint.”
Meta is now tweaking the formula so the resulting concrete can achieve its full strength in a shorter time frame to account for its particular construction needs. It’s also testing the concrete’s resilience against certain extreme weather conditions. The authors wrote in the research paper accompanying the project that they hope the results will eventually help others in the construction industry better adjust their ingredient ratios to maximize sustainability.
The company also hopes to bring the same AI approach to other aspects of infrastructure design and construction in the future.
“The resulting concrete mixes from our model can be used outside of data center construction and there is an opportunity to further develop this model to address other use cases,” the authors wrote. “Our exploration of innovative solutions to reduce data center construction emissions is not limited to concrete. There are opportunities to reduce the emissions of other materials. We are also exploring innovative data center designs as another way to improve sustainability.”
