AI everywhere: Tangible applications for sustainability teams

Climate Week 2023 in New York featured a series of riveting conversations around new AI applications with potential to accelerate climate action. The intersection of artificial intelligence and sustainability has been on my mind since 2020, when we published an insight on the topic, authored by strategic advisor Karuna Ramakrishnan. To follow up, we convened a webinar with experts employing AI in the realm of waste management (Conor Riffle, Rubicon), finance (Faiz Sayed, Aquantix AI), and nature-based solutions (Nan Pond, formerly of NCX). In 2020, the foundation was being set for AI in the sustainability space, and our research identified multiple ways in which these new applications could accelerate green breakthroughs in the 2020s.

Three years later, it is time for a refresher on the topic. Like AI itself, the AI x sustainability conversation is evolving rapidly, and the new products and applications within it increasingly sophisticated. As a company or organization pondering how to embed these emerging AI tools within operations, data analytics, and customer relationship management, the AI landscape can feel overwhelming. This refresher is meant to clarify some of the main areas AI can enrich your sustainability efforts, and to provide tangible examples of applications to consider.

Some quick background on what got me excited about this topic. Publishing the Global Green Economy Index brings me close to country-level data related to various aspects of the green economy. The traditional “bottom-up” methods for generating these data (e.g. country reporting, sector-based estimates, and modeled datasets) are not always as accurate as required, and lack timeliness and granularity. Recently, “top-down” methods for data collection from satellites, sensors and other technology-enabled tools introduced a new approach to collecting and analyzing these data. AI plays a central role in this process, automating systems for data capture and teaching machines to translate images and observations into datasets related to different green economy topics. These topics include GHG emissions, land-use patterns, and site-specific readings from company assets generating power, manufacturing goods, or extracting raw materials. New initiatives employing AI related to GHG emissions include Climate TRACE, Kayrros, CarbonMapper, and GHGSat; platforms focused on agriculture, land-use patterns, and biodiversity include SkyWatch, Planet, and Gro Intelligence.

These new AI models and applications consume energy, often lots of it. Training one AI model can consume as much electricity as 100 U.S. homes do in an entire year. Any company or organization must consider how integrating AI to their operations will affect both scope 2 emissions (from purchased electricity) and net zero targets. In addition to accounting for the increased electricity to run these applications, there are best practices emerging to further mitigate AI-related energy consumption. The Google “4M approach” recommends selecting efficient “sparse models,” using processors and systems optimized for machine learning training, computing in the cloud rather than on-premise, and map optimization to choose locations with the cleanest energy. By following these practices, Google claims, energy can be reduced by 100x and emissions by 1000x. To generate a baseline, the IBM Cloud Carbon Calculator provides estimates of emissions associated with cloud computing. And remember, there is also a social dimension to AI. People design both AI systems and make decisions about where data centers are located, which can often adversely impact natural resources and local communities. AI ethicist Afua Bruce provides a useful overview of these issues in a Guardian article here.

With an overall AI strategy in place – along with a clear plan on how to deal with the additional electricity use it will entail – companies and organizations can start integrating AI applications in different areas of operations. Many start with tools that don’t necessarily link specifically to sustainability. For example, ChatGPT can be used to more efficiently synthesize information about a wide range of topics. Bard – a conversation, generative AI chatbot developed by Google – can automate touchpoints with customers. These range from pop-ups on websites with tailored information or navigational assistance to performing the traditional role of a customer service chatbot. Companies and organizations monitoring social media engagement and sentiment linked to different posts can experiment with AI tools like Sprinklr. Using AI processes to “socially listen,” Sprinklr can generate faster insights on how different messages are perceived, what type of conversations they are inspiring, and how this benchmarks against previous communications efforts.

As your company or organization cascades a sustainability strategy throughout different departmental units, AI tools can improve the capture, analysis and reporting of key sustainability data. Typically, operations is one of the first departments integrated into a company’s sustainability plan. This is because data associated with a company’s scope 1 (linked to operations) and scope 2 emissions (linked to purchased electricity) are managed by the operations team. Reducing energy consumption and the associated cost-savings are always a company priority, and AI tools can accelerate these efficiency gains. Some examples include C3.ai (monitor & reduce scope 1 emissions) and Brainbox AI (monitor & reduce scope 2 emissions). Scope 3 emissions from supply chains remain a challenge for companies to track, but satellite and sensor-based tools mentioned earlier may help over time. Sector-specific AI tools like KoBold are giving new visibility to the mining sector, while Grey Parrot links AI with waste, adding efficiency and cost-savings to waste management.

As you can see, AI will likely accelerate your company or organization’s ability to gather and analyze sustainability data, but what about managing it in a way that feeds into emerging sustainability frameworks? Given the disjointed reporting requirements for companies over the past few years, sustainability professionals have been challenged to efficiently report on their ESG (environmental, social, and governance) metrics. Some software solutions in this realm are quite broad, connecting insights across governance, risk, compliance, audit and ESG. Examples of these software solutions include Diligent and Workiva. Others are enterprise climate platforms, with a more narrow focus on mapping GHG emissions data to reporting frameworks like the Task Force on Climate-Related Disclosures (TCFD), CDP, and others. Examples of these software solutions include Watershed and Worldfavor. The more climate-focused nature of these platforms also provide companies and organizations with tools to begin reporting on supply chain (scope 3) emissions, a critical area for improvement given that in some sectors, the majority of GHG emissions are scope 3.

In the 2020s, AI is everywhere and holds huge potential to accelerate sustainability and climate action. Finding the right balance between human and AI-centered decision-making will be paramount for all companies and organizations, with data collection, management and reporting at the center of this challenge. COP28 in Dubai later in 2023 will surely feature more discussion and product introductions around AI x sustainability. 

This article is also published on Dual Citizen. illuminem Voices is a democratic space presenting the thoughts and opinions of leading Sustainability & Energy writers, their opinions do not necessarily represent those of illuminem.