AI Elevates Sustainability

As climate change accelerates, disruptions surge and enterprise sustainability demands pile up, there’s a growing recognition that business as usual isn’t sufficient. It’s becoming increasingly difficult to juggle spreadsheets filled with tens of thousands of suppliers and PDF files that contain snapshots of diverse fragments of business operations.

“There’s a need for greater simplicity and efficiency,” says Christina Shim, vice president and global head of IBM Sustainability Software. As a result, organizations are now turning to artificial intelligence tools to navigate complex sustainability issues, including carbon reduction and various risks associated with weather, water, energy, biodiversity, sourcing and supply chains.

These AI tools, which increasingly tap machine learning (ML) and generative AI frameworks, automate data flows, and spot patterns and trends -- often at a level humans can’t see. They also introduce more advanced capabilities, such as digital twins, that model risk factors and aid in product and packaging reformulation.

Simply running AI itself, however, takes a toll on the environment. AI's potential benefits to sustainability goals must be balanced against those costs. So, as businesses take aim at the 17 United Nations Sustainable Development Goals, establishing a more advanced sustainability framework is essential. “AI offers ways to become more intelligent about sustainability and the environment,” observes Scott Likens, global AI leader at PwC. However, as

Related:Sustainable Partnerships Pay Dividends

Sustainability Gets Smarter

AI is emerging at the center of sustainability. According to a September 2023 IT Sustainability study conducted by IBM, 37% of business leaders report that their firms are already actively using AI for sustainability, and an additional 40% plan to use it soon. Despite some attempts to politicize sustainability, mostly in the US, “Virtually no one is pulling back on sustainability investments and the role of AI is evolving rapidly,” Shim says.

Gaining deeper insight -- and the ability to act on data -- is more important than ever. The real-world impacts and costs resulting from climate change continue to inch upward, while demands for environmental, social, and governance (ESG) reporting and regulatory compliance are growing. At the same time, AI technology can lower energy and water costs, while helping business leaders identify opportunities to innovate and even gain a competitive advantage.

The value of AI is particularly apparent as sustainability data piles up and organizations attempt to gain broader and deeper insights into Scope 3 emissions that they don’t directly control. Shim says she has seen companies that wind up with upwards of 30,000 suppliers and partners packed into a spreadsheet. “Using pivot tables to gain insight into the sustainability data is nearly impossible,” she says. “It’s critical to automate tasks through AI.”

Related:Reusing Waste Heat from Data Centers to Make Things Grow

Unfortunately, no single AI tool, technology, or solution can address the entire challenge -- and AI can’t eliminate the hard work of making actual changes in sourcing, manufacturing, or packaging. “You don’t want to use a hammer to do the job of a screwdriver. It’s important to match AI tools with the right tasks,” observes David Rolnick, an assistant professor in the School of Computer Science at McGill University and co-founder and chair of the global nonprofit Climate Change AI.

Large language models that encompass generative and foundational AI, for example, can condense hundreds or thousands of pages of documents and deliver useful summaries. They can crunch sales data, compact market demand forecasts and pricing data, and illuminate the best business path toward sustainability, including optimizing production and reducing waste. According to IBM, 72% of executives now approach sustainability as a revenue enabler rather than cost center.

Related:Increased Greenwashing Stains ESG Efforts: Report

Yet, humans must stay in the loop. Not only can bad training data lead to hallucinations, biases and inaccurate results, “Someone must be around to analyze the results and make sense of the relevant facts and information,” Rolnick says. In addition, organizations should approach digital twins and other tools thoughtfully. “It’s very easy to use AI in ways that lead to misleading and even counterproductive results,” he adds.

AI Broadens the Scope

The IBM survey also found that a lack of insight from data was the No. 1 sustainability challenge for 44% of CEOs. Yet, the right combination of solutions -- typically addressing things like carbon accounting, data analytics, reporting and supply chain management -- can simplify things. Specialized solutions such as Boston Consulting Group’s CO2 AI, IBM Envizi ESG Suite and Microsoft Sustainability Manager can quantify how products, physical assets, facilities, cloud resources, and IT frameworks impact energy consumption, water use, and carbon emissions.

Deeper data insights drive faster and better decision making but also offer clues about how to combine, integrate and take full advantage of various sustainability technologies. For example, one major telecom company collapsed 10 discrete tools into a single system that slashed energy use by more than 50%, Shim notes. “AI can supercharge the data journey. The key is to evolve from hundreds of data sources spread across multiple systems to a single system of record.”

AI can also unlock hidden value, PwC’s Likens says. For instance, highly accurate simulation models can map out physical footprints for facilities, land use, transportation and logistics frameworks, and even reformulate products and packaging. “A simulation model can help an organization understand its infrastructure, but also the infrastructure of partners. It can analyze and map out the optimal flow of goods through the supply chain,” he explains.

When organizations combine AI with other digital technologies, the capabilities expand, Likens says. IoT devices and other sensors allow organizations to peer into nooks and crannies that reside in supply chains and manufacturing processes, including difficult to track Scope 3 emissions. Satellite data, LIDAR and other forms of imagery and machine vision data can illuminate sustainability opportunities and fuel innovation, he says. “AI is expanding the scope of what’s possible within sustainability initiatives.”

Sustaining Success through AI

AI capabilities continue to advance, Likens says. Tools like IBM Watson X and Microsoft Copilot can translate ESG reports and other documents into different languages or flag issues that may require closer examination. “You might prompt the system to produce different summaries with different points of focus, so that humans can spot what’s truly important and use the information effectively,” Likens explains.

Of course, there’s often a need to educate, train and upskill teams -- and even the C-suite -- about how to use AI effectively. It’s also vital to recognize that the AI landscape is changing rapidly. Avoiding a “paralysis by analysis mentality” is essential, Shim says. “It’s important to get started with AI and get a handle on data with a single system of record. “The goal is to build an AI ecosystem built on cooperation and collaboration,” she says.

Concludes Likens: “AI doesn’t provide a magic button that gives you all the answers. But it is a valuable tool that can augment, extend and redefine an organization’s sustainability efforts.”