Professor Kevin Leahy Awarded Office of Naval Research Young Investigator Award for $707K
Professor Kevin Leahy from the Department of Robotics Engineering has earned the Office of Naval Research (ONR) 2025 Young Investigator Program (YIP) Award for $707,951 over the next three years. His project, “Graph Learning for Distributed Situational Awareness with Limited Communication and Localization,” supports the Cooperative Autonomous Swarm Technology (CAST) initiative under ONR’s Sea Warfare and Weapons portfolio (Code 33), under the guidance of Program Officer Christine Buzzell. Professor Leahy’s work combines graph neural networks (GNNs) with marine robotics to solve a critical challenge: coordinating teams of autonomous underwater (UUV) and surface vehicles (USV) in environments where communication and localization are deeply constrained. His team aims to determine the minimum information exchange necessary, when it’s essential, and how it affects coordinated behavior. “Underwater environments demand a rethink of traditional swarm robotics,” explains Professor Leahy. “Cameras and GPS are unreliable submerged—acoustic sensing dominates, and communication bandwidth is extremely limited.” His approach equips robots with the ability to infer missing information, enabling effective collaboration even with sparse communication. This year’s award is especially significant, as it marks Professor Leahy’s final year of eligibility for ONR’s early-career funding. After earning his Ph.D. in 2017, he became ineligible for many other DoD programs—yet recognized this opportunity to secure essential support for hardware-driven research. The CAST program seeks to advance cooperative autonomy in marine environments, where conventional swarm technologies struggle. Professor Leahy’s research bridges the gap by adapting strategies from land and aerial robotic systems to underwater applications—closing a vital technological divide. Under this award, Professor Leahy’s lab— two full-time Ph.D. students and multiple summer research assistants—are deploying multiple UUVs and USVs. These pilot trials focus on real-world sensing, communication, and decentralized control, laying essential groundwork for deeper theory development during colder months. This summer’s field deployments will establish the robots’ sensing and control capabilities. Over the fall and winter, attention will shift to mathematically grounding emerging theories. The lab plans to return to field tests next summer—now backed by rigorous insights and validated infrastructure. The potential applications extend beyond defense. Professor Leahy emphasizes the work’s significance for Bayesian search, a strategy used in the recovery of Air France flight 447, and for underwater infrastructure monitoring and marine ecological studies. His tools promise robust, scalable autonomy even when communication is unreliable. When reflecting on how vital the ONR award has been, Professor Leahy notes that the work could not have been done without the award. Deploying a multi-robot testbed—especially underwater—is costly and complex. “Creating our testbed wouldn’t have been possible without this award,” he says. Professor Leahy offers this advice to early-career researchers: Talk to multiple program officers. “By talking to many of them, I could target the program that was the best fit for my interests.” His outreach efforts were key to aligning his work with CAST’s mission.
Professor Kevin Leahy from the Department of Robotics Engineering has earned the Office of Naval Research (ONR) 2025 Young Investigator Program (YIP) Award for $707,951 over the next three years. His project, “Graph Learning for Distributed Situational Awareness with Limited Communication and Localization,” supports the Cooperative Autonomous Swarm Technology (CAST) initiative under ONR’s Sea Warfare and Weapons portfolio (Code 33), under the guidance of Program Officer Christine Buzzell.
Professor Leahy’s work combines graph neural networks (GNNs) with marine robotics to solve a critical challenge: coordinating teams of autonomous underwater (UUV) and surface vehicles (USV) in environments where communication and localization are deeply constrained. His team aims to determine the minimum information exchange necessary, when it’s essential, and how it affects coordinated behavior.
“Underwater environments demand a rethink of traditional swarm robotics,” explains Professor Leahy. “Cameras and GPS are unreliable submerged—acoustic sensing dominates, and communication bandwidth is extremely limited.” His approach equips robots with the ability to infer missing information, enabling effective collaboration even with sparse communication.
This year’s award is especially significant, as it marks Professor Leahy’s final year of eligibility for ONR’s early-career funding. After earning his Ph.D. in 2017, he became ineligible for many other DoD programs—yet recognized this opportunity to secure essential support for hardware-driven research.
The CAST program seeks to advance cooperative autonomy in marine environments, where conventional swarm technologies struggle. Professor Leahy’s research bridges the gap by adapting strategies from land and aerial robotic systems to underwater applications—closing a vital technological divide.
Under this award, Professor Leahy’s lab— two full-time Ph.D. students and multiple summer research assistants—are deploying multiple UUVs and USVs. These pilot trials focus on real-world sensing, communication, and decentralized control, laying essential groundwork for deeper theory development during colder months.
This summer’s field deployments will establish the robots’ sensing and control capabilities. Over the fall and winter, attention will shift to mathematically grounding emerging theories. The lab plans to return to field tests next summer—now backed by rigorous insights and validated infrastructure.
The potential applications extend beyond defense. Professor Leahy emphasizes the work’s significance for Bayesian search, a strategy used in the recovery of Air France flight 447, and for underwater infrastructure monitoring and marine ecological studies. His tools promise robust, scalable autonomy even when communication is unreliable.
When reflecting on how vital the ONR award has been, Professor Leahy notes that the work could not have been done without the award. Deploying a multi-robot testbed—especially underwater—is costly and complex. “Creating our testbed wouldn’t have been possible without this award,” he says.
Professor Leahy offers this advice to early-career researchers: Talk to multiple program officers. “By talking to many of them, I could target the program that was the best fit for my interests.” His outreach efforts were key to aligning his work with CAST’s mission.
Latest Announcements - Talent & Inclusion
- How Math at WPI Is Solving a 70-Year-Old Mystery in Muscle ScienceWPI professor Sam Walcott developed a molecular model that explains and accurately predicts muscle force, offering new opportunities for medical innovation. In honor of our upcoming Arts & Sciences Week, WPI is showcasing research that demonstrates how mathematics is advancing medical science. A new mathematical model developed at Worcester Polytechnic Institute (WPI) could enhance our understanding and treatment of heart disease. Created by Sam Walcott, director of bioinformatics and computational biology, the model simulates how microscopic structures within muscle cells generate force, using principles from both physics and biology to describe the interaction of individual molecules. It also reveals how subtle changes at the molecular level can lead to serious cardiac conditions. The research could inform the next generation of energy-efficient prosthetics. Walcott collaborated with Edward (Ned) Debold, professor of kinesiology at UMass Amherst, and Walter Herzog, professor of kinesiology at the University of Calgary. Using rabbit muscle tissue, Debold conducted molecular-scale experiments to study how individual muscle proteins respond under different conditions, while Herzog examined how whole muscle cells generate force. Their combined experiments provided data that Walcott then used for his model. “We have developed a mathematical model that describes how muscle cells generate force by accounting for how the molecules in the cell interact,” Walcott explains. “This connection between the cellular and molecular scale is important because, for example, genetic heart disease often causes subtle changes in one or two types of molecules in the heart muscle, yet drastic changes in heart function.” Walcott’s research has big potential for the future of medical science. For example, a relatively new discovery in the world of muscle contraction research is thick filament activation, which is a kind of “on/off-switch” for muscle molecules. Walcott’s mathematical models account for this process and suggest how it might affect muscle function. When you tense your muscle and stretch it (as if you're beginning to lose an arm-wrestling contest), your muscle can produce more force than without the stretch, a phenomenon called force enhancement. Similarly, if you tense your muscle and shorten it (as if you're beginning to win an arm-wrestling contest), your muscle can generate less force, a phenomenon called force depression. These phenomena, discovered in 1952, lack a molecular explanation. Remarkably, though one might expect force depression and enhancement to arise from the same process, there are differences between them—for example, force enhancement is not associated with an increased number of force-generating muscle molecules, while force depression is associated with a decrease in those molecules. A leading idea for how force enhancement arises is that a molecular "spring" gets engaged as you activate your muscle. When the muscle is then stretched, the spring is also stretched, thereby generating some extra force in addition to the force-generating molecules in the muscle. Walcott and his collaborators proposed that, when the muscle is shortened, the spring contracts. This then decreases the force in the muscle. Thick filament activation proposes that the force-generating molecules switch "off" when force drops, so this drop in force decreases the number of force-generating molecules. This explains both the drop in force observed in force depression and also why the number of force-generating molecules decreases. The model, which was originally designed to describe the Herzog lab’s cellular experiments, was also able to successfully predict the results of the Debold lab’s molecular-scale experiments. This suggests that we can, in fact, connect the behavior of molecules with the function of muscle cells. These discoveries mark an exciting step in the world of medicine and biomechanical design, like heart disease research and prosthetics. “Designing prosthetics requires thinking about how muscles use energy, since one wants the prosthetic to be both functional and efficient,” Walcott explains. “If we understand how muscle molecules interact, we can understand how they use energy and how the muscle overall uses energy.” Walcott’s research was supported by a $1.4 million grant from the National Institute of General Medical Sciences (NIGMS), an institute of the NIH. This project also highlights the interdisciplinary focus of WPI’s Bioinformatics and Computational Biology program, where students and faculty use math and data to explore the frontiers of biological research.
- Last Day to Submit Wellness Day Events for 9/19!Today is the last day to submit Wellness Day events for approval! Requests received after today will not be approved. Submit your event request here.
- WPI Researchers Design Microbial Tool To Analyze Neuropeptide Function: Advance May Point To New Approach for Peptide TherapeuticsJagan Srinivasan Shruti Shastry Liz DiLoreto Researchers at Worcester Polytechnic Institute (WPI), in collaboration with researchers at Baylor College of Medicine, have developed a simple, scalable method to study how specific neuropeptides affect behavior by programming common lab bacteria to deliver peptides directly to worms. The research also suggests a possible microbial approach for the future design of peptide therapeutics. The study, “Harnessing microbial tools: Escherichia coli as a vehicle for neuropeptide functional analysis in Caenorhabditis elegans,” was published in GENETICS in August 2025. Neuropeptides—small protein messengers that fine-tune brain circuits—are notoriously tricky to evaluate one by one. Traditional approaches often rely on creating transgenic animals or purchasing synthetic peptides, both of which are time-consuming and expensive. The WPI team instead engineered the bacteria Escherichia coli (E. coli) to produce single neuropeptides, then fed those bacteria to Caenorhabditis elegans (C. elegans) worms with a neuropeptide loss-of-function genetic mutation. The researchers then measured whether native behaviors—such as mate-searching, chemotaxis, and pheromone avoidance—were restored. “Our approach turns bacteria into on-demand couriers for the nervous system,” says Jagan Srinivasan, senior author and associate professor in WPI’s Department of Biology and Biotechnology. “When a behavior snaps back only if the matching receptor is present, you get direct, in-vivo evidence for which peptide talks to which circuit—and which ones are redundant versus uniquely powerful.” Because the method delivers intact, sequence-defined peptides through engineered microbes, it suggests a new peptide therapeutic strategy: using microbial “chassis” to produce and deliver short, bioactive peptides in vivo. While this study focuses on worms, the same design principles—sequence control, receptor specificity, dosing through diet—could guide the development of next-generation microbial or probiotic therapies in more complex systems. “We see this as a proof of concept for microbial peptide therapeutics,” says first author Liz DiLoreto, PhD '25. “In true WPI fashion—hands-on and collaborative—our tiny teachers (C. elegans) let us learn the rules fast: which sequences work, how to dose them, and how receptor context shapes outcomes. Those rules can guide adapting the approach to mammalian models.” “What excites me is the accessibility,” adds second author and graduate student Shruti Shastry. “Because the method uses standard E. coli and simple feeding, it’s easy to scale and share, empowering more labs and students to test many peptides and build the design playbook for translational work.” Beyond developing a new toolkit for worm neuroscience, the method opens the door to broader discoveries. Because it cleanly separates individual peptides, it can help researchers identify new peptide-receptor pairs, examine peptide processing and uptake, and investigate how neuromodulators change circuit “states” during complex decision-making.
- The Power of Our Collective Work: A Thank You to WPI Faculty and StaffAs the fall season approaches and A term is underway, we in University Advancement want to express our gratitude to WPI’s faculty and staff. Your commitment and care are central to the experiences of our students and the strength of our community. Thank you for sharing your time, your expertise, and your ideas. Whether you’re collaborating with donors, mentoring students, supporting campus life, or engaging in community -your work and your partnership matter deeply. We know that the impact of our efforts is greatest when we work together. As we look ahead, we are grateful for your partnership—and we look forward to deepening our shared efforts to ensure that WPI continues to be a place where innovation, belonging, and student success are possible for all. With appreciation, University Advancement
- Save the Date: Wellness, Engagement & Belonging Employee SymposiumPlease save the date for our upcoming Wellness, Engagement & Belonging Employee Symposium Thursday, October 16, 2025 Rubin Campus Center | Full-day event This symposium will bring together employees from across campus to explore themes of psychological safety, wellness, and belonging through engaging sessions and meaningful conversations. Keynote Speaker: Dr. Kyle A. Reyes Author of Lenses of Humanity, Dr. Reyes will share insights on human-centered leadership and inclusive community building. His work invites us to reflect through four transformative lenses: See, Become, Belong, and Believe. Featured Sessions Include: Psychological Safety in the Workplace Wellness Strategies for a Thriving Community Leadership Panels on Cultivating Belonging at WPI Lunch will be provided for all registered attendees. You do not have to attend the full day. Bonus: The first 15 registrants will receive a free copy of Lenses of Humanity. Register here! We look forward to seeing you there and building a stronger, more connected WPI community together.
- LabTEK: Research Equipment Repository at WPIAs WPI's research activities continue to expand, there is an increasing need for software tools that support effective communication and inform faculty, students, and staff. Recognizing this, the Research Division and Information Technology Services (ITS) have partnered to create LabTEK, a new resource designed for WPI researchers and their external collaborators. LabTEK Lab Asset Database LabTEK introduces a comprehensive asset database and equipment ticketing system, offering the following benefits: Provides a more complete resource for shared capital research instrumentation. Inform researchers about existing capabilities across the campus. Enables efficient tracking and management of capital lab instrumentation. Supports replacement decisions through integrated reporting tools and synchronization with Workday. Replaces the former labequipment.wpi.edu site with more detailed information and a simplified presentation. Initial Release The Team has completed the initial rollout of LabTEK focusing on a subset of service centers across the WPI campus. Please visit the LabTEK website utilizing the LabTEK icon at www.help.wpi.edu Additional communication will be forthcoming as the Team adds “Phase II” of assets from the overall WPI Fixed Assets. Feedback on the system is welcomed by emailing VPRI-shared-facilities@wpi.edu.