This summer, researchers and diplomats from across the globe gathered on Johns Hopkins University’s Homewood campus for the Energy Technology and Science Bridge event—a forum designed to support collaborative approaches to today’s energy challenges. Co-hosted by the Ralph O’Connor Sustainable Energy Institute (ROSEI), the National Science Foundation’s Electric Power Innovation for a Carbon-free Society (EPICS) Center, and the Embassy of Belgium in the United States, the event featured the latest energy research from several Johns Hopkins engineers, including three from the Whiting School of Engineering’s Department of Civil and Systems Engineering (CaSE): Yury Dvorkin, Magdalena Klemun, and Jan Drgona.
The U.S. director of EPICS, Yury Dvorkin, also one of the event’s organizers and associate professor in the departments of Civil and Systems Engineering and Electrical and Computer Engineering, helped bring together Hopkins engineers with diplomats and representatives from Canada, the Netherlands, Denmark, Finland, France, Spain, Switzerland, Poland, and the EU Delegation to the U.S. to explore potential collaborations in energy programs.
During the forum, Dvorkin emphasized how cross-industry partnerships can accelerate the application of cutting-edge research to real-world problems.
“There’s a mutual interest among researchers and industry professionals who aim to expand energy resource availability to support the AI revolution,” said Dvorkin. “Both groups also hope to use data science and AI to improve energy efficiency, along with power grid resilience and management.”
Dvorkin’s own research draws on AI and optimization methods to tackle some of the energy sector’s biggest challenges, from smarter grid management to integrating renewable energy approaches at scale. During the event, he presented a recent case study on the integration of wind power across large geographical areas. The study demonstrates that the usefulness of this technology increases when wind turbines are sited based on their mutual complementarity. This approach helps avoid idling and maximizes energy generation across all areas, rather than concentrating production in a single location.
Dvorkin’s study also considered the integration of offshore wind power with the mainland power grid and highlighted the value of optionality—the ability to enhance offshore infrastructure at a later stage—to improve system resiliency and efficiency. His presentation showed how optionality can increase overall power production and reduce the burden on the land-based grid to compensate for the intermittent nature of offshore wind power. Working with Professor Dennice Gayme in the Department of Mechanical Engineering, Dvorkin’s research continues with the analysis of the effects of wake modeling on wind farm capacity factors and how advanced control strategies can alleviate these impacts.
The day’s program also included presentations from CaSE faculty on the topics of emission measurement for cloud computing and the use of optimization to improve energy-intensive systems.
Assistant Professor Magdalena Klemun, whose work focuses on data-driven models to track the evolution of energy technologies and policies, presented new research on the challenges of measuring and attributing emissions from cloud computing. As data center power demand grows and digital services cross corporate and national boundaries, emissions tracking has become increasingly complex.
“Over the past 10 years, there’s been a shift towards more standardized and increasingly mandatory carbon emission disclosure,” said Klemun. “This creates challenges for digital services, where energy use is difficult to isolate and emissions responsibilities are fragmented across providers and users located in different jurisdictions.”
She noted that while territorial emissions—those tied to where pollution occurs and reported under the Paris Agreement—are declining in most EU nations, consumption-based emissions can reveal a different picture. “This has long been a focus for physical goods, but digital infrastructure is now increasingly important to understanding climate accountability.”
Klemun’s talk introduced new approaches to better account for the location of digital service use, and discussed implications for how EU countries measure and report ICT sector emissions.
Klemun and Dvorkin are currently working together on the Nexus Award-funded Project INTERSECT, which seeks to understand the cost, reliability, and environmental impacts of data center deployment on PJM, a regional transmission organization.
Associate Professor Jan Drgona shared his expertise in machine learning and advanced control for real-world applications. During the event, he highlighted the need for optimal building control systems.
“Today’s buildings and data centers account for a significant share of global energy use, largely due to inefficient heating, ventilation, and air conditioning systems, referred to as HVAC,” said Drgona. “According to the U.S. Energy Information Administration, buildings account for 70% of electricity use in the U.S. alone. While the knowledge and engineering tools to optimize HVAC systems exist, outdated infrastructure and fragmented deployment still prevent that innovation from reaching its full impact.”
Drgona’s presentation featured the use of AI, specifically scientific machine learning, or SciML, in HVAC control systems, which offers significant benefits over current control methods. SciML modeling and design control is roughly 10 times faster than traditional model-based approaches, it’s coding can be reused to templatize systems, and it can be applied to a broader set of systems, such as buildings, power systems, or data centers.
Additionally, Drgona and Dvorkin are working to expand on recent HVAC cost prediction results in collaboration with the University of Mons to better understand data center energy implications.
During a day of discussions and networking, Dvorkin, Klemun, and Drgona’s research underscored Johns Hopkins’ leadership at the intersection of engineering, AI, data science, and policy. Bolstered by the university’s strategic investments in AI and data science, their collective work is helping to create a foundation for both sustainable energy systems and global energy policies.
Michel Wallemacq, senior economic advisor at the Embassy of Belgium, said that “There’s a strong need for science and diplomacy collaborations, and this event is only the beginning of opportunities to learn from Johns Hopkins.”
CaSE’s curriculum is specifically designed to prepare students to solve civil engineering’s grand societal challenges—complex, large-scale issues that require interdisciplinary approaches, like building future energy infrastructure systems. Research performed by Dvorkin, Klemun, Drgona, and other faculty reinforces the curriculum with real-world problems and advanced approaches for energy systems engineering.
“Our shared research doesn’t stop at the edge of campus or country borders,” said Dvorkin. “Education and research have a central place in shaping sustainable energy policies worldwide. With international partners, we can expand resource availability, deploy AI to improve efficiency from buildings to power grids, and foster solutions to benefit communities everywhere.”