Artificial Intelligence, an ally against climate change
Written by Curtis M. Wong
The extinction of species, the rise in temperatures and major natural disasters are some of the consequences of climate change. Countries and industries are aware and work to combat the planet’s accelerating pollution. Are there any viable solutions? According to some researches, using big data and machine learning could help drive energy efficiency, transforms industries such as the agriculture and find new eco-friendly construction materials.
Greta Thunberg is 16 and lives in Sweden. Until February last year, she was just another student concerned about the environment. Today, she’s become the world’s most influential climate activists, after founding ‘Fridays for Future‘, a movement that encourages school students to take time off from class to participate in demonstrations against their countries’ governments for breaching environmental laws. Greta believes that we’re heading towards a natural disaster and the planet’s destruction if we fail to change our habits as a society. And that includes everyone, from large industries to the individual citizen.
“Younger generations are already demanding solutions”
Greta reports directly to the bank’s management board. His avowed mission is “to ensure the bank systematically positions people at the heart of its decision-making processes.”
Articles, reports and studies all point in the same direction: global warming is a reality and is a product of human activity. The FAO (UN Food and Agriculture Organization), estimates that average temperatures will rise between 1.5 and 4 degrees in the coming decades. Greenhouse gas emissions resulting from the combustion of fossil fuels and industrial activity have grown steadily in recent years, especially in five regions: the U.S., China, Russia, the European Union and India.
This growth threatens thousands of species with extinction. Also, drastic changes in climate patterns have caused extreme weather events (heat waves, hurricanes…) to double over the past 20 years according to Economic Losses, Poverty & Disasters, a report published by the United Nations Office for Disaster Risk Reduction (UNDRR).
Despite these discouraging prospects, it is still not too late. There is growing consensus about the decisive role that data and computational sciences (big data and artificial intelligence) can play in helping find solutions to change habits and fight climate change. “Big data technologies allow us to work with swaths of data from different sources, while machine learning technologies allow us to process them and come up with predictive models. There is no doubt that they can both help us analyze and predict climate change. Together, it is clear that they can be extremely useful. However, right now, the interpretation and action still depend on us,” said Berenguer Bríquez, an AI expert specializing in the energy sector and runner up in the AI and Machine learning category at Datathon Iberia 2019 de AWS.
At Climate Change AI, an organization made up by volunteers from academia and industry, computational science researchers from leading US colleges have discussed the challenges we are facing and how computation can help mitigate them. The list of experts includes Andrew Ng, co-founder of Google Brain, DeepMind founder and CEO Demis Hassabis and Jennifer Chayes, managing director at Microsoft Research. AI can focus on and yield positive environmental results in a number of areas, including energy, transport and logistics, agriculture and construction.
Greater energy efficiency
One of the niches where AI can be pivotal is the energy sector, whose forms of production and distribution remained virtually the same for decades. Today, thanks to algorithms, we have automated distribution networks capable of performing real-time smart assessments weighing in demand and supply levels, and of detecting potential errors. “AI can help boost energy efficiency in many small tasks and aspects of our lives – with smart home and building automation devices – and in operations and logistics – with the so-called Industry 4.0. From a macro standpoint, it will allow us to generate more synergies as it enables new ways of organizing and planning high volumes of resources and assets,” said Bríquez. Thanks to algorithms and machine learning, it is already possible to anticipate the electricity demand of a city or a manufacturing plant with months in advance, and to distribute power to small local populations more efficiently with buildings equipped with their own solar panels.
An example of how big data’s potential benefits in the renewable energy industry are Google’s efforts to boost wind farm efficiency in its fleet of wind farms in the U.S. Thanks to the algorithms developed by Alphabet’s Deepmind, researchers were able to predict wind farm energy output 36 hours in advance, building on the advanced weather forecast technologies now available and taking into account fluctuations in energy market activity. Based on these forecasts, the technology suggested a series of actions to cover its customers electricity needs while selling the power at an optimal price.
To be sustainable or no to be, that’s the question
Each individual can play a part in the fight against climate change, to one degree or another. With small, daily gestures, it is possible to contribute to the conservation of the environment. Large corporations, through different debt instruments, bonds, and even specific transactions, also play a role in supporting this campaign, a campaign that concerns every corner of society. More specifically, what impact do businesses have on the environment, and what can they do to offset it?
More responsible transportation and logistics
Another sector where AI can help improve efficiency while helping curb global warming is the transportation and logistics sector. According to data from the IPCC (Intergovernmental Panel on Climate Change), between 1970 and 2004, the transportation and logistics sector increased its greenhouse emissions by 120 percent.
Resorting to algorithms, transportation companies can predict more accurately the level of demand on a per-location basis, planning ahead for events, avoiding risks and creating solutions. Thus, companies resorting to road transportation to fulfill its orders can optimize its efforts and resources, cutting costs and issuing less CO2.
This is the case of multinational companies such as DHL, one of the global logistics services providers, whose software can juggle up to 58 different parameters to define, days in advance, optimal schedules for their cargo airplanes. Thus it can optimize travel operations and curb its carbon emissions.
New sustainable materials for construction
Today artificial intelligence technologies are contributing to speed up the development of new, more eco-friendly construction materials. At research team at the University of Jaén, in southern Spain, is studying a viable ways for manufacturing bricks from plant waste and other materials, such as steel. The goal is to replace materials such as concrete, thus reducing the tremendous amounts of carbon emissions that result from its manufacturing process.
Agriculture, a problem and a solution at the same time
The increase in the amount of land used for agricultural processes has caused the sector’s CO2 emissions to double over the past 50 years, to 5,000 million tons in 2014, according to a FAO report. The use of fertilizers, deforestation and methane produced by livestock are the main drivers of this trend.
But computer science can help making agriculture much more environmentally friendly. “Agriculture is changing thanks to AI and the Internet of Things. We now have livestock and crop motorization technologies, satellite images for plague control and plan sowing and harvesting cycles,” said Bríquez. These innovations will contribute to reduce fertilizer volume and to manage corps in a more effective manner. The goal is to produce more and regenerate land health.