As wildfires rage in the west and the density of smoke is blocking out the sun in many areas, attention is turning to the fight against climate change and whether AI can make a difference.
Climate Change AI is among the initiatives working on it. The group of AI researchers is exploring solutions to climate change and related issues such as food insecurity and human displacement. The group in the past month has discussed what type of startups can combat climate change; they are assembling a dataset wish list from researchers to inform the data used to train models, according to an account in VentureBeat.
One effort is from WattTime, a nonprofit organization that aims to reduce the carbon footprint of a household by automating when electric vehicles, thermostats and appliances are active based on when renewable energy is available. Algorithms to determine those times are trained using data from the continuous pollution monitoring system of the EPA. The technology is available in California, which produces one-third of its power from renewable energy.
“Nobody knows about the U.S. continuous emissions monitoring system, but it’s been live since the ’70s, and it’s why organizations like mine can write increasingly sophisticated AI algorithms to integrate more renewable energy and do what we do,” stated WattTime Cofounder and Executive Director Gavin McCormick. WattTime received a grant last year from Google.org’s AI Impact Challenge, to see whether computer vision can be used to track power plant emissions outside the US from satellite images.
Moreover, McCormick worked with former US Vice President Al Gore on the July announcement of Climate Trace, a coalition attempting to build a tool to track human-cause greenhouse gas emissions from all over the planet.
The Intergovernmental Panel on Climate Change (IPCC), said to be the world’s leading scientific body on the climate crisis, has advised that emissions of accumulated greenhouse gases (GHGs) need to be cut in half by 2030 and reach net-zero emissions by 2050, to avoid the worst effects of global warming.
“We are honored to announce that a powerful new tool will soon be joining the climate fight,” Gore wrote in a recent account in Medium. “Climate TRACE is a coalition creating a high-tech solution to independently detect emissions and where they’re coming from, everywhere in the world, in real time. It’s a feat that’s never before been possible — until now.”
In addition to WattTime, founding members include Blue Sky Analytics, CarbonPlan, Carbon Tracker, Earthrise Alliance, Hudson Carbon, Hypervine, OceanMind, and Rocky Mountain Institute.
The coalition plans to leverage advanced AI, satellite image processing, machine learning, and land- and sea-based sensors to monitor GHG emissions from every sector and in every part of the world. The intent is to monitor human-caused GHG emissions worldwide with a granular focus down to specific targets including power plants, ships and factories.
The group hopes to present its emissions data in Glasgow, Scotland, next year when countries are scheduled to gather to renew their commitments to the Paris Agreement, signed in 2016 to address greenhouse gas emissions mitigation. Under US President Donald Trump, the US announced in June 2017 it would be withdrawing from the agreement.
The current system of tracking greenhouse gas emissions is a patchwork of infrequent self-reporting by companies and countries, without reliable third-party verification, and long lags in reporting. “We can only manage what we can measure,” Gore wrote. “Countries, companies, and leaders worldwide want to solve the climate crisis, but lack the tools to do so quickly and effectively.”
The global sensor network to include satellites and ground- and sea-based instruments, all connected to an AI engine built for the purpose, “Emissions have nowhere to hide,” Gore stated. Scientists, regulators, the news media, citizen activities, investors and business leaders will see where the GHG emissions are coming from, and whether it’s increasing or decreasing.
Recent technology advances have made this solution possible. The group built a smaller version of the project to measure power plant emissions; the experience shows the range of components. The effort combined imagery from: multiple satellite constellations (like the European Space Agency’s Sentinel 2 mission), AI algorithms from experts in computer vision (such as Pixel Scientia Labs), data pipeline engineering (Google.org), power plant databases (World Resources Institute), remote sensing (Valence Strategic), power systems modeling (WattTime), weather adjustments and power plant cooling systems (Carbon Tracker), and more.
“We envision a future in which low- and zero-carbon energy is the norm. We believe Climate TRACE will be an integral part of making that future become reality, and we’re getting right to work,” Gore wrote, then issuing a challenge to the industry: “Today, we issue a call to action: If you’re working in a field that touches on emissions monitoring — whether you have AI expertise, satellite sensor networks, or other global sensor or emissions data networks — we want to hear from you.”
Efforts to combat climate change have been taking place and deserve recognition.
Google is using machine learning to lower the energy use of its data centers. The London-based AI unit DeepMind is using information collected by sensors to reduce energy use for cooling for up to 40%, according to a recent account in Thrive. Plans are to apply the system to clean energy output so Google can better manage its conventional energy needs.
The company Hypergiant of Austin, Texas, is growing algae so it can absorb carbon dioxide and give off oxygen. Naturally-growing algae grows out of control. The company’s scientists have developed an AI unit called the Eos Bioreactor that can regulate the growth of the algae and optimize its carbon-absorbing properties. The size of a refrigerator, the unit is said to be 400 times more effective at capturing carbon than trees in the same area.
Green Horizons, an IBM research initiative, is using cognitive computing and the internet of things (IoT) to analyze climate change data. Cognitive computing is being paired with IoT to predict pollution rates in Beijing. The system uses machine learning to ingest data from sources such as meteorological satellites and traffic cameras to constantly learn and adjust the predictive models. It is able to forecast pollution 72 hours in advance, with an accuracy down to the nearest kilometer on where the pollution is coming from and where it will likely go.
Beijing is using this methodology with the goal of reducing pollution levels ahead of the 2022 Winter Olympics. It can use the predictions to implement policies such as temporarily restricting industrial activity or limiting traffic and construction. It is modeling hypothetical scenarios to test the effectiveness of the interventions.
Comments