Join top executives in San Francisco on July 11-12, to hear how leaders are integrating and optimizing AI investments for success. Learn More
As the COP26 United Nations climate change summit wraps up today, it seems unlikely that the world’s governments will take swift action to save the planet. Could technology leaders do so instead? Some are trying.
The Microsoft Cloud for Sustainability, announced in June, aims to help companies understand and improve their climate footprint. Just this week, Nvidia CEO Jensen Huang ended his keynote at the company’s annual tech conference with a commitment to creating an E2 (Earth 2) supercomputer capable of creating a digital twin of the entire Earth. By modeling the planet at unprecedented resolution, E2 is meant to accurately predict the climate decades into the future and guide efforts to mitigate global warming.
“It seems we’re a little late in resolving some of these problems, so it would be great if tech can help us move faster,” said Paige Marie Morse, sustainability lead at AspenTech, one of the companies participating in the COP26 summit.
The concept of a digital twin — a simulation so exact that it can serve as a stand-in for the real thing — holds some of the greatest potential because digital twins make it possible to test large-scale changes in industrial processes before they are implemented in hardware and concrete.
Join us in San Francisco on July 11-12, where top executives will share how they have integrated and optimized AI investments for success and avoided common pitfalls.
While not quite as planetary scale in its ambitions as Nvidia, AspenTech also points to the use of digital twins for chemical process optimization and the accelerating adoption of clean-burning hydrogen fuels. One stand-out example: BASF redesigned its methanol production process to cut CO2 emissions by recycling the carbon-containing off-gases back into the process. AspenTech says a digital twin created with its software modeled planned changes to the production process, so adjustments could be made before they were put into production.
Founded 30 years ago as a spinoff from MIT research, AspenTech creates software for process industries such as chemical manufacturing and fossil fuel refineries, which these days include software for meeting sustainability goals.
AspenTech hopes to participate in the creation of “the circular economy” where the full lifecycle of every product and industrial output is accounted for, including its environmental impact, and waste products are eliminated or recycled to the maximum extent possible, Morse said. “One challenge is that, if you look at the actual math, there’s no agreement around the world on how you measure these things,” she said.
Amazon’s Climate Pledge Fund recently announced investments in three green innovation companies, Infinium, Resilient Power, and CMC Machinery, each of which has products that could reduce the carbon footprint of Amazon’s ecommerce and logistics operations. For example, with CMC Machinery’s process for rightsizing packages, Amazon projects it will be able to cut the number of plastic air pillows it uses to cushion small products shipped in large boxes by 1 billion next year. Resilient Power is working on improving urban electrical infrastructure to support more charging stations for electric vehicles like Amazon delivery trucks. Infinium combines clean-burning hydrogen with captured carbon dioxide to create a transitional fuel that can be used in jet engines and diesel vehicles.
“We’re supporting tech and products that will allow us to meet our goals around our climate pledge and be more sustainable in general,” said Matt Peterson, Head of the Climate Pledge Fund.
Amazon also has machine learning initiatives aimed at meeting its climate goals. And while Amazon didn’t particularly push a digital story in conjunction with the Climate Pledge Fund announcements, but that doesn’t mean there isn’t one. For example, Resilient Power President Josh Keister said via email that software helps his company “bootstrap off of orders and grants without too much capital intensive hardware prototyping,” adding that “once [the] equipment is installed, big data and machine learning on AWS also increases reliability and provides faster feedback to the design process. One example would be to use the sound or vibration from a fan to schedule preventive maintenance prior to an actual issue occurring.” Infinium says it developed its own homegrown GIS mapping application, which it uses to analyze pipeline and electrical transmission infrastructure, optimize the use of renewable power sources and minimize carbon dioxide emissions.
What about power-hungry datacenters?
The datacenters where such models run are also large energy consumers with their own carbon footprint. The more electrons we send racing through processors, the more electricity they consume and the more heat they generate, and the more cooling systems must compensate, consuming yet more power. These effects can be mitigated somewhat with efficient designs or the use of natural cooling, which is why cold climate nations like Norway say they’re ideal datacenter sites.
However, the contribution of datacenters to global warming is often exaggerated. I came across several articles implying that datacenters are responsible for 2% of the world’s carbon emissions before tracking down a 2018 article in the scientific journal Nature that appears to be the source of that statistic. Actually, the scientists gave that as an estimate for the entirety of computing and communication technologies — including the power that goes into every PC, iPhone, telecommunications switch, and home router — putting the carbon footprint of computing and communications technologies about on par with the airline industry’s use of jet fuel.
The Nature article estimated the contribution of datacenters, specifically, at about 0.3% of the world’s carbon output. On the other hand, it warned that under some pessimistic scenarios they could grow to consume 20% of the world’s electricity, particularly if power-hungry applications like cryptocurrency mining proliferate. AI and machine learning models consume disproportionate amounts of electricity, as well, as do supercomputers.
One of Huang’s claims for E2 is that it will demonstrate breakthrough energy efficiency as a supercomputer. That remains to be seen, but meanwhile, the world needs all the brainpower it can muster to battle climate change — including AI-magnified brainpower.
VentureBeat's mission is to be a digital town square for technical decision-makers to gain knowledge about transformative enterprise technology and transact. Discover our Briefings.