Frostburg State Student Presents Groundbreaking AI-Driven Robotics Research at IEEE International Conference

Frostburg State Student Presents Groundbreaking AI-Driven Robotics Research at IEEE International Conference

Jun 9, 2026 11:00 AM

When Jeremy Perando first became interested in robotics as an elementary school student competing in robotics programs, he never imagined that journey would lead to presenting original research on artificial intelligence and robot navigation at an international conference.

Now a graduate in the computer science program at Frostburg State University, Perando recently presented his first-author research paper, "Constrained Behavior Tree Generation for Safe LLM-Driven Robot Navigation," at the 2026 IEEE International Conference on Electro/Information Technology (IEEE EIT 2026) in La Crosse, Wis. The paper was co-authored with Dr. Xunyu Pan, assistant professor of computer science and director of the Frostburg Laboratory for Artificial Intelligence and Robotics (FLAIR).

The conference, held May 21-23, brought together researchers, educators and industry professionals from around the world to share advances in electrical engineering, computer science, artificial intelligence, robotics and intelligent systems.

Perando's achievement marks a significant milestone for Frostburg State University. He is the first undergraduate student in the history of FSU's Department of Computer Science and Information Technologies to publish a first-author paper at an IEEE international conference.

"Jeremy gave a very good presentation and received thoughtful questions and positive feedback from the audience," said Pan. "In our session, he was the only undergraduate presenter among researchers that included doctoral students and faculty members. That speaks to the quality of his work and the level of professional research experience he has gained at FSU."

Perando presented during the conference's Spatial Intelligence, Navigation and Environmental AI session, where his research focused on improving the safety and reliability of AI-controlled robots.

The project builds on previous work conducted in FLAIR, a research laboratory established by Pan to expand opportunities for student research in artificial intelligence and robotics. The new study addresses one of the field's biggest challenges: ensuring that AI systems can safely make decisions while navigating real-world environments.

"We added a number of safety features to the system," Perando explained. "Our more advanced checking system is the novel part of the project. Based on an instruction, the robot determines what it should be doing and then verifies that its actions are actually leading toward that goal and that the goal is achievable."

Pan said the research fills an important gap in current AI robotics studies. "When you publish at the IEEE level, you need to contribute something new," he said. "I reviewed the existing research and found examples of behavior trees being used for robot navigation, but I had not seen anyone focus on constraining those behavior trees to provide safer control of AI-driven robots. That became the foundation of our work."

For Perando, the research connects directly to a lifelong passion for robotics. "I've been involved with robotics since elementary school," he said. "I competed through middle school and high school and gained experience with robots, computer vision and related technologies. When I came to Frostburg, there really weren't many opportunities in robotics, so being able to work with Dr. Pan on projects that combine AI, robotics and computer vision has been a great experience."

The research also points toward future applications. The team is exploring ways to run AI models directly on robots rather than relying on cloud-based systems, reducing delays and improving performance in environments with limited internet connectivity.

"One potential next step is running the model directly on the robot using specialized hardware," Pan said. "That would allow the robot to make decisions more quickly and continue operating even when a network connection is unavailable."

Perando's paper represents an early scholarly achievement for the NSF-supported Appalachian Innovation Institute (AI²), which promotes research, innovation and workforce development in emerging technology fields including artificial intelligence, robotics, cybersecurity and advanced manufacturing. FLAIR serves as one of the institute's research laboratories and provides students with hands-on opportunities to participate in cutting-edge projects.

"This work aligns closely with AI²'s mission to advance use-inspired research in emerging technologies," Pan said. "We want to build opportunities for students, strengthen research at Frostburg State and attract more students who are interested in robotics and artificial intelligence."

The paper is Perando's second professional publication. In 2025, he co-authored "Enhancing Robotic Navigation with Large Language Models" with Pan, which was published through Electronic Imaging 2025.

As FLAIR continues to grow, Pan hopes more students will become involved in research. "We recently established a dedicated research suite for the lab, and we're looking to expand opportunities for students interested in AI and robotics," he said. "If students are interested in this type of research, we encourage them to reach out and learn more."

For Perando, the conference experience reinforced the value of undergraduate research. "It was exciting to present our work and engage with researchers in the field," he said. "It showed me how much opportunity there is to contribute to new developments in AI and robotics, even as an undergraduate student."

Perando's travel to present at IEEE EIT 2026 was supported through an FSU Foundation Opportunity Grant.