Companies are collecting mountains of data to understand their environmental footprint. But to turn that data into sustainable action, they’ll need help from cutting-edge artificial intelligence.
Illustrations by Timo Lenzen
I n a city brimming with skyscrapers, One Vanderbilt still manages to stand out. At 1,401-feet, the two-year-old office tower is one of the tallest buildings in Manhattan. It’s also one of the greenest. The building was constructed with 90 percent recycled steel rebar. It has a state-of-the-art cogeneration system that keeps its energy use low and a 90,000-gallon rainwater collection system that recycles water for irrigation and cooling. As a result, it boasts one of the highest levels of LEED certification. “We see environmental sustainability as a social obligation. It’s not just a trend,” says Laura Vulaj, senior vice president of hospitality and sustainability at SL Green, the real estate company that owns the tower.
While One Vanderbilt is well ahead of the pack on sustainability, Vulaj knows that she and her team are nonetheless going to have to pick up the pace. In New York City, a rigorous new climate law is requiring landlords to dramatically reduce their environmental impact. That comes on top of state and federal regulations, United Nations targets, as well as requests from board members, investors, and potential tenants for data on environmental, social, and governance (ESG) performance.
Fulfilling those demands is particularly complex for a landlord like SL Green, which has hundreds of tenants and a vast portfolio of properties. According to Vulaj, requests for environmental compliance data have increased tenfold in recent years, and each new framework requires different reporting methodologies. So today, getting a full picture of SL Green’s environmental footprint, and figuring out how to shrink it, is a tall order. It requires aggregating and analyzing a mountain of data from a variety of sources across multiple buildings. “There’s energy data. There’s water data. There’s waste data,” Vulaj says. “Data is pouring in for every building, and it’s living in so many different areas. You get data fatigue.”
Vulaj’s data challenges are shared by leaders at many companies focused on sustainability. This year, according to a poll conducted by the IBM Institute for Business Value, more than half of CEOs ranked sustainability among their top concerns. Yet 44 percent of CEOs said they lack the ability to translate sustainability data into insights that help them meet environmental targets. “It’s coming to a head,” says Kareem Yusuf, Ph.D., general manager of IBM Sustainability Software. “Society cares a lot more, investors are using this to inform where they place their dollars, and regulation is only going to become more apparent.”
Faced with such challenges, companies like SL Green often come to IBM for help. “Often the conversation starts with, ‘I need to get a handle on this,” says Yusuf. “‘How can I make sense of all this data? I can’t do it with spreadsheets anymore.’” Yusuf’s solution for these organizations is simple: AI. “Machine learning can look at data, bring it together, and make sense of it—and then, most importantly, place it in front of you in a way that allows an informed, intelligent decision to be made,” Yusuf says. “It’s operationalizing sustainability.”
More and more companies are following this advice. According to an IBM study, two-thirds of IT leaders say their company is currently planning or already in the process of using AI to manage the complexity of data for sustainability. According to Witold Henisz, vice dean and faculty director of the Environment, Social and Governance Initiative at the Wharton School, that’s a huge shift. In years past, many companies kept such minimal sustainability data that they could effectively relegate it to a single spreadsheet column. Now, he says, the scale of the sustainability data companies are collecting requires more sophisticated technology. “This is a big data problem,” he says.
Bjarne Jørgensen, executive director of asset management and operations at Danish civil infrastructure operator Sund & Bælt, came to understand that problem well in 2020, when he began looking into how to preserve the Great Belt (Storebælt) Link, an 11-mile system of bridges and tunnels connecting the Danish islands Zealand and Funen. When it was built in the 1990s, Jørgensen says, the system looked indestructible. But it turned out to be no match for the ravages of the North Sea and climate change, which have deteriorated the system with fierce winds and tidal surges. “Our focus was on prolonging the lifetime of the bridge, which also reduces its carbon footprint, because if you have to rebuild something, it releases more carbon,” Jørgensen says.
To preserve the system, Sund & Bælt needed to know its health in real time. This meant processing between 12,000 and 14,000 data points collected from moisture-detecting sensors and a fleet of drones inspecting 300,000 square meters of concrete. And the data itself could only go so far. “Data doesn’t necessarily improve your decisions,” says Jørgensen. “You have to see into it and find the essence in order to use it.” It’s a common complaint. “The promise of big data analytics is that we’re going to gain insight, which is going to help performance and address the climate transition,” says Henisz. “But it’s not a crystal ball. A lot of analysis has to be done.”
That analysis, in the case of the Great Belt Link, relied heavily on AI. Using IBM Maximo Civil Infrastructure and Maximo Application Suite for intelligent asset management helped Sund & Bælt generate penetrating, real-time analysis on the condition of the bridges, tunnels, and other critical infrastructure components. Harnessing the power of AI to analyze visual inspection data on rust, corrosion, displacement, and stress, alongside maintenance records, design documents and 3-D models, provides Jørgensen’s team with crucial insights not only on the current health of the bridge but also on the potential impact of changing environmental conditions.
The results have been game-changing for the team managing the Great Belt Link. IBM’s AI has accelerated and streamlined workflow processes, including the timing of inspections. It has also quickened the decision-making power of engineers in the field and allowed them to plan further ahead.
To its surprise, the Sund & Bælt team found that the Great Belt Link system, which had been expected to last 100 years, could significantly lengthen its lifespan using AI. “We now know that if we keep getting better information about the health of the concrete and steel, then we can reach 200 years,” says Jørgensen. Those extra hundred years will save the company the cost of new construction and reduce its carbon footprint by 750,000 tons, twice the mass of the Empire State Building—an achievement for both the business and the environment. “Thankfully, they go hand in hand,” Jørgensen says.
For SL Green, the promise of that win-win motivates the company’s ongoing pursuit of even more sustainable buildings. In the coming year, Vulaj says, the company will develop targets to achieve carbon neutrality at buildings like One Vanderbilt—both because the science demands it and because SL Green customers expect it. “Regardless of what the legal mandates are, we feel we have an obligation to reduce emissions and improve the energy efficiency of our buildings,” she says. To ensure its success, SL Green is ditching its sea of spreadsheets and shifting to Envizi, an IBM software suite that will allow the company to manage all its ESG indicators—including energy use, carbon emissions, and environmental and social responsibility metrics—in one place, making it easier to analyze, operationalize, and report.
Once those systems are integrated, Vulaj hopes, One Vanderbilt will stand out even more in the New York City skyline for its green bona fides. But to create a truly sustainable world, buildings like One Vanderbilt will have to become more commonplace, which means more companies will have to supercharge their sustainability initiatives. According to Henisz, companies are currently spending $35 billion per year on financial data, but only $1 billion on ESG data. On one hand, he says, you could look at those figures and focus on the fact that ESG data is just 3 percent of the total spend on financial data. Or you could recognize, as he does, that “there’s a lot of runway to do more.”
