Earthly Business
At the Nicholas School of the Environment, researchers and entrepreneurs are joining forces to solve environmental problems.
Growing up, Dalia Patiño-Echeverri spent four months each year on a remote family farm in the mountains outside Bogotá, Colombia, where she was raised. Removed from the city hum, she spent halcyon days riding horseback amid dairy cattle and fruit trees and reading bound volumes of Reader’s Digest from the 1940s.
“For someone visiting like me — for a little girl — it was perfect,” says Patiño-Echeverri, Gendell Family Associate Professor in the Division of Environmental Social Systems at Duke’s Nicholas School of the Environment.
But she also understood the rigors of her rural setting, which had no electricity for many years. She recalls women rising before dawn to light the fire for cooking and heating water. “It was so clear how life can be really difficult without electricity,” she says.
Things changed on the farm the day 30 men with mules arrived after an arduous journey up a narrow mountain path. The animals bore heavy machinery — components of a transformer that would eventually channel electricity to the homestead.
Once the lights went on, the house became a social hub, attracting far-flung visitors eager to watch La Fiera, a Mexican soap opera, on the family’s T.V.
“It was like going to the movies in the theater,” Patiño-Echeverri recalls. “I remember men would come with their kids on their shoulders. They would walk two to three hours and then go back at dark.”
Through those experiences on the farm and in Bogotá — a city of glaring wealth disparities — she found her calling: solving energy problems for human and planetary good. More specifically, Patiño-Echeverri studies how to deliver energy in the most efficient, affordable way possible, while minimizing waste and risk — a process called optimization.
“Doing it right has profound consequences. And not doing it right also has profound consequences,” she says.
The Nicholas School is an ideal setting for her work, which is already seeing real-life applications.
Last year, Patiño-Echeverri joined forces with Jesko von Windheim, Chair of Business and Environment, to launch GridSeer, an AI-driven software and analytics platform designed to improve energy system management through smarter forecasting and robust risk analysis. The aim is to optimize how energy supplies meet demand.
In March, their team formalized a collaboration with Duke Facilities Management to test GridSeer technology on the Durham campus. The project is emblematic of the Nicholas School’s burgeoning ethos of innovation — one that embraces different disciplines and departments. The school is a font of research with commercial potential, according to von Windheim, but faculty frequently need help getting their ideas across the startup line.
“Quite often, unfortunately, in the academic sector, there’s this belief or expectation that if a faculty member has a promising commercial idea, that they’ll go out and turn it into a company,” says von Windheim, who is also Lynn Gorguze-Scott Peters Professor of the Practice of Environmental Entrepreneurship and Innovation. But establishing a commercial path typically takes about 2-3 years, he explains. Given their multiple responsibilities — from teaching to advising to conducting research — faculty hardly have time to hone or apply the necessary business acumen.
That’s where von Windheim and colleagues come in. With expertise in entrepreneurship, marketing, accounting and law, his network offers a guided path to business development for busy researchers with brilliant ideas.
“Think of a world where we just leave Dalia on her own and say, oh, okay, do GridSeer on top of everything. It’s just not smart, right?” von Windheim says. But with his team handling business matters, “she can provide her intellectual capital and her advice and, honestly, her star power on the technical side when it’s really needed.”
Smarter Forecasting
In 1999, the author Alan Weisman published Gaviotas, the true story of an eponymous ecovillage located in a war-torn region of eastern Colombia. Founded in 1971 by an Italian visionary named Paulo Lugari as a model of sustainable development, the village was to be entirely self-reliant on farmed food and renewable energy. The community still exists half a century later, a testament to human ingenuity against inhospitable odds.
The unlikely success of Gaviotas resonated with Patiño-Echeverri, who cites Lugari and the book as inspiration for her path into engineering. As a Ph.D. student at Carnegie Mellon, she helped develop a computer model for energy system planners interested in minimizing the cost of replacing fossil fuels with renewable energy sources, such as solar and wind.
After joining the Nicholas School in 2008, she created simulations to better understand major power system operations such as PJM, an organization that manages the largest wholesale electricity market in the United States and oversees energy transmission across a large swath of the nation’s grid.
Through her research, Patiño-Echeverri has become intimately familiar with the major challenges facing energy system planners during a time when the climate is rapidly warming, and air pollution from fossil-fuel burning tops the list of global environmental health threats.
“We have urgency to make energy even cleaner,” Patiño-Echeverri says, but “you cannot make it more expensive, because people will die. It is not uncommon for American families to reduce or forgo food or medicine to pay for their energy bills. Or they may keep their homes at unsafe temperatures that ultimately increase their health risks.”
Keeping energy affordable, she adds, is “as important as not making it dirtier. And the energy has to be there — it has to be reliable.”
Delivering on these factors — reliability, affordability and environmental sustainability — is central to GridSeer’s mission. Its technology is designed to improve how energy system operators account for uncertainty when planning and allocating, or dispatching, energy supplies. Relying on renewable energy, for instance, means grappling with the vagaries of wind and sunlight.
Traditionally, most energy system operators forecast one most likely future supply-and-demand scenario — what Patiño-Echeverri calls a “best guess”— and maintain backup reserves in case they’re wrong. As a simple example, the probability of a snowstorm hitting North Carolina in July might be next to zero, whereas a heat wave is highly likely. With those considerations in mind, energy operators would plan for the latter possibility.
This type of forecasting can be more expensive or polluting, however. Insufficient reserves could mean booting up an expensive and polluting combustion turbine, according to Patiño-Echeverri, but maintaining more reserves than needed is also expensive.
GridSeer instead uses a modeling technique that assigns probabilities to an array of possible futures based on weather patterns and historical energy system performance. This approach enables system operators to pre-position their power plants and storage systems to operate cost-effectively under all plausible scenarios. In this way, they hedge their bets against less likely, but potentially costly, events.
Although probabilistic forecasting isn’t new to the energy world, GridSeer is ahead of the curve, according to Patiño-Echeverri. Through machine learning, the platform quickly crunches copious data to provide timely forecasts that feed into planning software, which operators can use to nimbly mix and match supply with demand. The planning tool is also customizable, allowing users to specify how often and how far into the future they wish to forecast, the risk they are willing to tolerate, and the cost they are willing to bear.
In 2020, Patiño-Echeverri’s team won a four-year, $2.4 million federal grant from the Advanced Research Projects Agency-Energy, or ARPA-E, to begin developing the technology that drives GridSeer.
After several years, the Patiño-Echeverri lab — known as GRACE, for “Grid that is Risk-Aware for Clean Electricity” — was clearly onto something. But to qualify for additional funding, the team had to show their platform could be commercialized. That’s when they called von Windheim.
Meeting of the Minds
Von Windheim has a proven track record as an entrepreneur and tech business founder. In the late 1990s, he served as vice president of Cronos Integrated Microsystems Inc., which manufactured devices used in fiber-optic communications. In 2000, Cronos executives brokered a deal to sell the business for $750 million, and von Windheim promptly became the general manager for the new business unit.
Following stints as the founder and CEO of a thermal energy startup and, later, as the CEO of a biomedical company, von Windheim joined the Nicholas School, where he met Aaron Franklin, a professor in the Pratt School of Engineering. Franklin had created an electronic sensor to precisely monitor tire tread wear, with the goal of improving tire safety while also reducing the occurrence of premature tire replacement. Drivers could then confidently save money on their tires while keeping still-good rubber out of landfills.
Recognizing commercial opportunity, von Windheim worked with Franklin to found a startup called Tyrata and solicit venture capital. After successfully transitioning the initial design into technology that monitors treads for large fleets — and therefore millions of tires under management — they sold Tyrata to tire industry juggernaut Bridgestone in 2023.
As von Windheim’s sees it, Tyrata embodies his business mantra: people, planet, profit.
“Although I think you have to start with profit — profit is essentially the grease that makes things work — at the end of the day, here at the Nicholas School and at Duke in general, there’s a focus to ensure that it’s not just profit for profit’s sake, but profit that helps people and makes the planet a better place,” von Windheim says.
Partnering with Patiño-Echeverri to pursue a smarter energy management system was a no-brainer. Together, they assembled a multidisciplinary team including market business development specialist Akash Aithal (MEMP’23), systems engineer Mauricio Hernandez (Ph.D.’24) and systems architect Dimitris Floros. A cadre of outside legal, marketing, sales, business operations and finance professionals — including technology translation experts Karl von Gunten and John Fuscoe, who have taught at Duke — provide additional support.
After incorporating GridSeer, the team submitted a $400,000 grant proposal to ARPA-E for “technology-to-market” funding, with an initial focus on optimizing energy management at Duke. They were immediately approved for the award.
“There are lots of projects funded by ARPA-E,” Patiño-Echeverri says, but “if they don’t have a Jesko person, a business-and-environment person that knows how to create a company and how to see value in what is mostly an academic exercise — how to see something that can be put into a market and commercialized — then there is no company, and there is no product and there is no follow-up grant.”
On-Campus Testing
Members of the GridSeer team gather as Duke Facilities campus operations engineer Cody Stevens, with the tape measure, and a control technician, crouching, check the level sensors at Duke Pond, a wetland ecosystem and reserve water source for a chiller plant that serves the buildings cooling system at Duke.
When it comes to GridSeer clientele, von Windheim sees a potential base in the commercial and industrial, or C&I, sector, which includes universities, data centers and manufacturing campuses. These entities are more willing to work with new technology than large utilities companies, he says, and are much more motivated by cost savings. Plus, success at one university might inspire others to follow suit.
To that end, this past March Duke Facilities Management became GridSeer’s first client (they are not paying for the service). Given Duke’s dedication to net-zero emissions through its Climate Commitment, GridSeer technology could play an important role in further reducing the university’s energy footprint.
“With our team of about 24, we operate and maintain the equipment that provides the utility needs of both the Duke campus and Duke Medical Center,” says Cody Stevens, campus operations engineer for utilities and engineering services within Facilities Management.
For example, the department distributes electricity for buildings, steam for sterilizing tools and chilled water for keeping MRI machines cold, among many other needs. Stevens and team have been exploring ways to improve dispatch, and GridSeer could help fit the bill.
“Do you use one boiler to generate all your needs, or do you use three boilers at higher efficiency to generate your needs? That’s why the forecasts are so important,” Stevens says. “If you know what your demand should be, then you can predict which equipment you need to make it happen.”
The collaboration is still in early stages, so it’s too early to tell how much GridSeer will benefit Duke. However, a preliminary analysis suggested that GridSeer could reduce forecasting errors, which would translate to cost savings and, potentially, energy efficiency improvements, according to von Windheim.
Regardless, the team is already experiencing the gratification that comes with working together on applied research.
“I think the role of academia should be helping society and the environment,” says GridSeer systems engineer Hernandez. As a researcher, “you can have an impact by writing a paper. But I think just doing this in the field will have probably more impact at some point.”
Coastal Connections
The energy sector isn’t the only place where Nicholas School innovation is taking off. Researchers within the Division of Marine Science and Conservation are also exploring ways to commercialize field-tested ideas.
For example, a team led by Brian Silliman, Rachel Carson Distinguished Professor of Marine Conservation Biology, is developing nature-based and engineered approaches to coastal ecosystem restoration as part of a Nicholas School initiative called Duke Restore.
One project involves designing technology that fosters the growth of coral and oyster reefs. Beyond supporting diverse marine life, reefs buffer shorelines from destructive waves.
“We’re coming up with oyster and coral attractor devices that mimic the texture, smells and shapes that attract baby oysters and corals to land on a particular spot. Once they are settled, these devices will then provide those babies with the protection they need to resist being eaten, and eventually grow into adults that then collaboratively form an ecosystem,” says Silliman, who also directs the Duke Wetland and Coasts Center.
For business guidance, Silliman is working with Shannon Parker (MEM/MBA’21), executive in residence at the Nicholas School and co-founder of Carbon Insights, a climate tech startup. Together they recently incorporated the company BioCoastal Systems to help scale novel restoration technologies spinning out of the Silliman Lab.
“What is really unique about the Nicholas School is actually having the environment where you have business faculty sitting right alongside the heavy-hitter scientists,” Parker says. “Starting the conversation about patenting and legal commercialization and licensing and all of this — it’s a huge foreign language to academics.”
“Faculty like Dalia and Brian pour their life into their ideas,” adds von Windheim. “It has taken us a decade to develop the infrastructure, but now at Nicholas, we have created an integrated, trusted business program, making it much easier for our scientists to develop their ideas from science into practice.”
Training Green Business Leaders
Duke students also benefit from the Nicholas School’s in-house entrepreneurial expertise. For example, the Master in Environmental Management Business and Environment concentration teaches graduate students how to implement environmentally sustainable business practices in private-sector companies.
Meanwhile, undergraduates enrolled in Duke Design Climate, a collaboration between the Nicholas School and the Pratt School of Engineering, gain practical experience working with industry experts and Duke faculty who are addressing real environmental challenges.
For example, a team called LightSyn Labs aims to scale low-cost, low-carbon fertilizer based on novel work done by Duke chemistry professor Jie Liu. LightSyn Labs is one of two Design Climate teams selected by the U.S. Department of Energy to participate in the EnergyTech University Prize Challenge, administered by the Office of Technology Transitions.
The collegiate competition challenges students to present a business plan that leverages promising energy technologies. The other team, EFCEM, is exploring how to scale a corrosion- and crack-resistant, low-carbon cement technology developed by ReefCycle, a startup founded by Shannon Parker, Nicholas School executive in residence, and industrial engineer Mary Lempres.
Design Climate also just celebrated the incorporation of its first business spinoff, Connexus. The team uses technologies such as virtual reality to recruit trainees from underserved communities to work in high-demand clean energy trades.
Duke has “smart, passionate, caring students. They’re technically savvy,” says Judy Ledlee, executive director of Design Climate. “I want to help show them the path for how they can start their own businesses. Even as young professionals, they can make a difference in the world.”
Subscribe to the Design Climate Insider newsletter for updates from the program.