The headlines are typically overrun with stories about how artificial intelligence (AI) is taking everyone’s jobs.
But AI isn’t as dark as the films and news headlines often make out, and this powerful technology is helping solve one of the biggest challenges faced by the planet and all the species on it: climate change.
Already, there is a wide range of compelling AI climate use cases, from predictive maintenance software decreasing emissions in the manufacturing sector to computer vision and machine learning imagery helping scientists track environmental changes.
Of course, no new technology is perfect. As such, AI climate solutions have many challenges to solve, particularly around cybersecurity and privacy.
But in the meantime, how is it helping?
How Can Artificial Intelligence Help Fight Climate Change?
Tackling climate change will inevitably require “intelligent solutions” such as AI, according to Albert Plugge, professor of ESG transformation and digital innovation at Nyenrode Business University in the Netherlands.
Plugge sees three areas in which AI can help firms meet their ESG goals and ultimately curb the effects of climate change, with the first being AI-powered water management solutions that send surplus water to water-deprived areas.
“AI may provide automated controls for decision-making, taking environmental factors like weather influences into account,” he explains.
Secondly, Plugge believes that AI can play a useful role in underground carbon dioxide emissions storage. Namely, the technology could help experts determine the most optimal temperature for storing emissions and tracking their subsidence impact. Subsidence is the sinkage of ground under buildings, causing them to collapse.
Thirdly, he says AI could enable organizations to predict future environmental risks arising from biodiversity measures, such as ensuring water sent to dry lands to improve their biodiversity doesn’t damage people’s homes.
How Can AI Help Reduce Carbon Emissions in Manufacturing?
Manufacturing and engineering companies are increasingly using AI-based predictive maintenance software to meet their net-zero targets, according to Malavika Tohani, research director of operational excellence at Verdantix.
She explains that predictive maintenance software allows manufacturers to make their machines more efficient and last longer by mitigating mechanical breakdown. She adds that this can decrease energy consumption, wastage, and greenhouse gas emissions in the industry.
“We are already seeing some diversified downstream energy firms implement AI platforms that leverage digital twins, artificial intelligence, and machine learning to recommend ways to improve the energy efficiency and performance of their assets,” she explains.
How Can Big Data and AI Help the Environment?
AI is perfect for solving significant challenges like climate change due to its ability to analyze large, multi-source datasets, according to independent technology consultant and Blinq Blinq founder Doug Stevenson.
“Whether it’s parsing data from weather satellite imagery, sensor networks monitoring energy use and emissions, or simulations predicting future impacts, the volume and complexity of information involved is precisely the type of challenge that AI excels at handling,” he explains.
By applying computer vision and deep learning technology to environmental imagery, he says scientists can monitor the effects of deforestation, algal blooms, and other ecological changes as they happen.
He continues: “Being able to automatically detect anomalies or subtle patterns across massive datasets could help uncover insights that may have otherwise gone undiscovered.”
Stevenson also expects AI technologies to improve the optimization of smart grids so that they use less energy and emit less carbon emissions.
He adds: “And as predictive modeling improves, we’ll have better forecasting to aid disaster preparation and mitigation.
How Can AI Optimize Renewable Energy Systems?
Organizations can use AI to optimize their systems, improve efficiencies, and assess potential risks in the renewable energy industry. Maria Opre, a cybersecurity expert and senior analyst at EarthWeb, saw these things firsthand when she helped a leading solar power company perform risk assessments.
She explains that running historic production and consumption data through machine learning technology allowed the company to predict the output and demand of solar power. Consequently, it now uses less fossil fuel-based backup power.
“By integrating forecasts into day-ahead scheduling, the AI optimized the flow of solar energy onto the grid when it was most needed and diverted excess to battery storage for later use,” she continues. “This preemptive matching of supply and demand minimized waste and curbed purchasing from fossil fuel plants that would otherwise provide extra capacity during peaks.
The use of AI has been fruitful for the solar power company, with Opre explaining that it saw a “15% decrease in greenhouse gas emissions per megawatt-hour compared to their traditional operations” in 18 months.
But AI isn’t just useful for improving existing renewable energy systems’ efficiency and environmental impact. According to Opre, it could also lead to the faster development and deployment of novel renewable technologies.
“Through machine learning, researchers are working to design wind turbines and solar panels better optimized for performance based on real-world data from installations worldwide,” Opre explains.
“AI can help engineers develop more efficient renewable components informed by insights from tens of thousands of existing systems operating under an immense variety of conditions. Faster, data-driven innovation cycles supported by AI may translate to quicker scale-up of renewable capacity globally.
How Can AI Be Used With Other Emerging Technologies?
Using AI with technologies such as 5G, edge computing, and the Internet of Things (IoT) will make the world more equipped to deal with climate change, according to Steve Carlini, VP of innovation and data center at Schneider Electric.
Carlini says these innovations are improving everything from resource allocation to traffic management. He expects them to improve the sustainability of cities and businesses by “increasing operational efficiencies and empowering us to make valuable changes to help reduce our water, land, and energy consumption.”
However, Carlini says powerful data centers will be needed for AI and other digital technologies to reduce emissions in areas such as manufacturing, transportation, real estate, and energy.
“These powerhouses can not only enable operational efficiencies and processes while increasing capacity to meet the demand of ever-increasing AI workloads, they can do so without increasing their own carbon footprint,” he says.
“Better still, AI data centers can continue reducing their carbon emissions over time using smart management for renewable power sourcing and innovative cooling technologies.”
What Challenges Are There to Consider?
While AI can help solve climate change in many different ways, some potential challenges must be considered and overcome.
Taimur Ijlal, information security leader at Netify, warns that a range of cyber security and privacy risks could arise from AI climate solutions.
“As we invest more in digitalizing climate solutions and building connected networks to monitor resource use and infrastructure, it expands the attack surface,” he says.
“Vital systems becoming dependent on AI models that could potentially be manipulated or services disrupted raises the stakes. Additionally, sensitive location data, individual energy profiles, and other personal information collected requires robust protections.”
As companies rush to adopt these technologies, he encourages them to ensure “resilience from the start”. They can do this by implementing industry best practices like “differential privacy in datasets, defensive training of models, and segmented, air-gapped deployments to limit harm from threats.”
Stevenson warns against relying solely on technology to fight climate change, urging governments, organizations, and individuals to consider “social and policy dimensions” simultaneously.
“Things like incentivizing reductions in high-emitting industries or changing consumer behaviors still come down to decisions by governments, organizations, and individuals,” he concludes.
“The right regulations and initiatives must complement what the models and algorithms provide.”