Skip to content
Fast Company Executive Board

The Fast Company Executive Board is a private, fee-based network of influential leaders, experts, executives, and entrepreneurs who share their insights with our audience.

Leveraging AI and real-time data in the fight against climate change

Researchers and scientists are turning to artificial intelligence (AI) as a potentially powerful tool in the fight to slow the pace of climate change.

Leveraging AI and real-time data

[Adobe Stock / florynstudio3]

Profile photo of Subbu Iyer

The impact of climate change continues to mount—from headline-grabbing extreme weather events like wildfires and hurricanes to more subtle, but equally heartbreaking, pervasive health consequences such as food insecurity and what the World Health Organization calls “waterborne and foodborne disease risks.”

And as a recent Planetary Health Check indicates, the outlook is trending in the wrong direction, as six of the nine “planetary boundaries” (the “key systems and processes that contribute to stable, healthy life on Earth,” according to the World Economic Forum) have already been breached.

As global efforts to slow the pace of climate change and lessen its effects continue to increase (albeit with limited success to date), researchers and scientists are turning to artificial intelligence (AI) as a potentially powerful tool in the fight.

THE DARK SIDE OF USING AI TO FIGHT CLIMATE CHANGE

Before diving into how AI is helping mitigate and address climate change, we need to first recognize the impact of AI itself on the environment. The tremendous amount of energy required to power AI is taking its toll: The International Energy Agency estimates that by 2026, energy demand from data centers will double from 2022 levels, reaching the same level as the electricity consumption of Japan.

To minimize AI’s energy consumption, technologists must work together to ensure the sustainability of the software and systems that power this AI future. Some of the areas include: 

  • Optimizing AI algorithms by creating more efficient AI training models and implementing techniques that minimize process requirements.
  • Enhancing data center efficiency by reducing the number of CPU cycles required to accomplish tasks, shrinking server footprints, improving the efficiency of software platforms to lessen carbon emissions, and ultimately, lowering the amount of energy needed to power data centers.
  • Improving hardware efficiency through the use of specialized hardware designed for energy-efficient AI processing. 
  • Optimizing workload management with batch processing, load balancing, and off-peak energy consumption.

THE REQUIREMENTS FOR POWERING REAL-TIME DATA FOR AI

Of course, a sustainable AI system still requires the efficient ingestion of multiple streams of data from potentially hundreds of thousands of sources: satellites, sensors, video and audio feeds, and more.

With that in mind, what’s often missing from the discussion around the use of AI for climate change work are the characteristics of the robust infrastructure (one that meets the sustainability suggestions above!) needed to efficiently process vast datasets and make it available for AI ingestion. As more data sources come online, the system must be able to scale to support them. The sheer volume and diversity of data necessary to power these AI systems can be overwhelming.

Given the various formats from multiple sources, real-time AI applications are also seeking to achieve efficiency through multiple data models—graph, document, vector, and key value—to support the wide range of ingested data sources. As mature multi-modal systems become more prevalent, AI applications are taking advantage of the flexibility they offer and the improved search control they provide. 

FROM PREDICTION TO ACTION: AI AS EARTH’S ALLY

The technological considerations for AI as a partner in the fight to contain and mitigate the effects of climate change are evolving as rapidly as the use of AI itself. As early as the 1980s, environmental scientists were adopting AI techniques and methods to improve weather forecasting. Today, we see AI being used across many aspects of the climate change discussion—for example, to process and analyze enormous amounts of complex data; monitor environmental changes both in real time and with an eye toward long-term trends; predict the effects of change or proposed policies; and, finally, to accelerate discovery. 

Here are just some of the real-world applications of AI in climate change today:

  • Environmental Monitoring: The United Nations Environment Programme (UNEP) uses AI to drive real-time analysis and future predictions on topics such as carbon dioxide atmospheric concentration, rising sea levels, and changes in glacier mass.
  • Active Risk Mitigation: The California Department of Forestry and Fire Protection uses computer vision to identify problem spots before they erupt. Meanwhile, in agriculture, AI is being used to improve food demand forecasting and management, and detect and prevent diseases and parasites, among many other use cases.
  • Policy Forecasting And Climate Simulations: NASA’s Earth System Digital Twins effort uses artificial intelligence to visualize and forecast various “what-if” scenarios to model future climate impacts based on a number of inputs, including proposed environmental policies.

One Aerospike customer, a company that is at the intersection of real-time data and climate efforts, uses an AI platform that processes inputs from almost one million unique public data sources to manage and respond to climate-related and other high-impact events. Using these sources, the company can detect early signs of natural disasters such as earthquakes, floods, hurricanes, and wildfires, enabling organizations, governments, and first responders to respond faster and more effectively to mitigate the impact.

Through a partnership with the United Nations Global Pulse program, they’re helping the UN accelerate humanitarian and crisis response efforts in emergencies. For example, if a region suddenly lost access to its water supply, UN staffers would be alerted so they could redirect water delivery trucks to the affected areas. 

Consider the use cases outlined above: Without accurate, real-time data, AI’s value in fighting climate change drops precipitously. In many cases, AI is being used to identify emerging threats or drive major policy decisions. In the most extreme real-time scenario, that of emergency response, the immediacy of the data can literally be the difference between life and death for thousands. 

A WAY FORWARD

It’s clear from these stories that AI’s value to climate change research and response continues to grow. As scientists, researchers, governments, and activists increasingly think of new ways to deploy AI in the fight against climate change, equal attention must be paid to ensuring such systems can manage real-time streams efficiently and accurately to help minimize energy consumption and create a more sustainable AI future.