School of Engineering and Computing

This research group focuses on advancing knowledge at the intersection of data science, machine learning, deep learning, remote sensing, and nuclear data science. Key efforts include developing deep learning models for remote sensing image processing, addressing challenges such as ground object detection, semantic segmentation, hyperspectral band synthesis, and natural disaster assessment. The group also leverages generative AI to build trustworthy AI systems, ensuring transparency, reliability, and fairness in high-impact mission-critical applications.

In the field of nuclear data science, we apply data-driven and physic-informed scientific machine learning methods to improve the analysis and modeling of nuclear processes, contributing to advancements in radiation monitoring, environmental safety, and energy systems. Broader data science applications include predictive modeling, time-series analysis, and anomaly detection, with relevance to environmental monitoring, urban planning, disaster management, and nuclear research.

Our mission is to create scalable, interdisciplinary solutions that bridge the gap between theory and practice, contributing to both scientific innovation and societal impact across multiple domains.

Contacts
Yan Lu (yan.lu@cnu.edu)
Keith Perkins (keith.perkins@cnu.edu)
William Phelps (william.phelps@cnu.edu)

The cybersecurity research group has a broad and diverse focus on topics in cybersecurity, spanning from technical interests to interests on the human and behavioral side. Specific areas of research include security in augmented and virtual reality (AR/VR), trustworthy vehicular networks and components, FANETs (Flying AdHoc Networks), intelligent intrusion detection systems (IDS), policy and, cybersecurity education pedagogy. This research draws on a variety of advanced technologies such as Machine Learning, Blockchain, Large Language Models, and Public Key Infrastructure (PKI), Head worn AR/VR devices, and mobile/efficient on-board intrusion detection functionality.

Contacts
Chris Kreider (chris.kreider@cnu.edu)
Michael Lapke (michael.lapke@cnu.edu)
Yan Lu (yan.lu@cnu.edu)
Ayan Roy (ayan.roy@cnu.edu)
Abhishek Phadke (abhishek.phadke@cnu.edu)

Capable Humanitarian Robotics and Intelligent Systems Lab (CHRISLab)

At CHRISLab, we focus on advancing autonomous robotic systems to enhance human capabilities and safety. Our work seeks to remove humans from hazardous environments, such as during rescue operations, and to develop assistive technologies to enable humans to flourish. We specialize in formalizing robotic system capabilities to enable automatic and verifiable synthesis of high-level behaviors from user-defined tasks. Through this, we aim to simplify the deployment of robotic systems, making them accessible to non-engineers who can easily adapt them to new tasks.

We maintain and contribute to several open-source ROS 2 tools, including the FlexBE system for hierarchical finite state machines. Our long-term goal is to enable "correct-by-construction" synthesis of hybrid systems that integrate hierarchical finite state machines, behavior trees, task planning, and AI/machine learning to solve challenging problems in robotics.

Contacts
David Conner (david.conner@cnu.edu, robotics@cnu.edu)

Autonomous Systems and Controls

Dr. Ghaffari's research focuses on dynamic modeling, safe control, and path planning with applications to autonomous aerial and ground vehicles. We develop novel solutions using control and optimization theory for existing and emerging problems related to autonomous robotics, precision mechatronics, and distributed energy systems. In particular, our research translates operational safety features, such as obstacle avoidance and actuator constraints, into programmable software components, which are easy to integrate with existing control systems. We leverage drone technology and our control algorithms to advance urban air mobility and precision aerial manipulators for agricultural applications. We also investigate the vehicle's interaction with the environment and analyze accidental impacts on the vehicle's safety and stability. We design and utilize hardware-in-the-loop systems to verify and validate our algorithms.

Dr. Abhishek Phadke’s research focuses on integrating resilience mechanisms in modern day Cyber Physical Systems. Target systems include UAV Swarms, Heterogeneous robot teams, Electric grids and Vehicular micro clouds. Resilience is a concept that is fuzzily described as the ability of a system to withstand disruptions and continue functioning. However this is difficult to program in real world control systems. UAV swarms are an ideal target system for examining and integrating resilience as they work in a dynamic environment where multiple disruptions seek to impede swarm operation. Results are validated using hardware and simulation platforms. Novel approaches to UAV control such as the integration of LLM (Large Language Models) for UAV operations is also being explored.

The research follows a two pronged approach. While methods to strengthen UAV control are studied, a second perspective looks at understanding external disruptions such as wind patterns, obstacle geometry and equipment disruptions. Through this, a better understanding of resilience implementations in vehicular cyber physical systems is expected.

Contacts
Azad Ghaffari (azad.ghaffari@cnu.edu)
Abhishek Phadke (abhishek.phadke@cnu.edu)

Communicational Collaborative Agents (COCOA Lab)

At COCOA Lab we seek to understand the role communication plays in enabling (software) agent collaboration in multi-agent systems. From message meaning to protocol specification to team formation, we look at communication through the spectrum of infrastructure, language, protocol, and social context to analyze collaboration and build tools to formalize it.

Contacts
Roberto Flores (roberto.flores@cnu.edu)
Abhishek Phadke (abhishek.phadke@cnu.edu)

We investigate and develop frameworks and methodologies in support of software engineering pedagogy. Our projects include:

Gooey: A Lean JUnit Testing Library for Java Swing Applications

Gooey is a programmatic, capture-and-test framework for Java Swing applications. Its integration to JUnit lets developers write automated tests that expose the components of displayed windows and allow manipulating them to verify expected results.

Gooey is currently being extended to support the testing of JavaFX applications.

Assignment Development Tools (ADTools)

Developing programming assignments that provide effective feedback to students is a resourceand time-intensive activity. As a result, instructors tend to give fewer assignments than what might be desired for pedagogical effectiveness.

ADTools provide a framework and tools to streamline and automate the assignment development process. The current framework enables JUnit methods to be written using only textual specifications of pre and post conditions, while the tool can generate complete Web-CAT JUnit classes given a spreadsheet describing the test cases. Our current work involves automatic generation of test cases, and an assignment specification language and translator.

Contacts
Roberto Flores (roberto.flores@cnu.edu)
Anton Siochi (siochi@cnu.edu)

We focus on the hardware and software systems that provide the foundation of our networked world. We address challenges that arise across a broad spectrum of technologies and at various layers of the protocol stack.

One emphasis is on applications of software-defined radios (SDRs) in networking and wireless communications. We leverage the flexibility of cognitive radios to improve spectrum sharing and utilization in novel wireless networking protocols. Further projects in this area include post-disaster networks, real-time localization using SDRs and investigating 5.8 GHz ISM band indoor signal propagation. In the fields of mobile computing and the internet of things our focus is on novel applications and the evaluation of platforms and systems.

Contacts
Keith Perkins (keith.perkins@cnu.edu)
Anton Riedl (riedl@cnu.edu)

For almost two decades, the nuclear physics group has made significant contributions at nearby Thomas Jefferson National Accelerator Facility (JLab). The group is supported through a National Science Foundation operating grant, with a focus on the program of experiments that will be carried out following the upgrade of JLab to a 12 GeV beam energy. In addition, the group members are supported in their research efforts through joint staff scientist appointments at JLab.

Experimental Physics Program

Dr. Edward Brash is involved in the successful form factor program where he is a spokesperson for two approved experiments that will take place in the 12 GeV era at JLab.

Dr. David Heddle develops visualization software for complex event topologies, which is of central importance to optimized event reconstruction and physics analyses.

Dr. Peter Monaghan’s research is a balanced blend of experiment, theory and hardware development. He contributes to the research program at JLab by running experiments in Halls A and C and in the improvement of parton distribution functions in describing real data with the CTEQ-JLab collaboration.

Dr. William Phelps is involved in Hall B at Jefferson Laboratory and in the EPIC collaboration at the Electron Ion Collider at Brookhaven National Laboratory. He has been involved in photoproduction and quasi-real photoproduction experiments in CLAS and CLAS12 in Hall B and GlueX in Hall D. He is a co-spokesperson of one experiment. Since 2014 he has been involved in the construction, calibration, and maintenance of the High Threshold Cherenkov Counter. He has also made significant contributions to CLAS12 offline software, in particular to the GROOT plotting and fitting package.

Theoretical Physics Program

Dr. Alberto Accardi applies the methods of Quantum Chromodynamics, the theory describing the strong nuclear interactions, to study how quarks and gluons build up protons and neutrons that build up the atomic nuclei - in brief, he studies the inner structure of visible matter. His approach combines theoretical physics methods with numerical analysis of large data sets on high-energy scattering processes such as electron-proton collisions at the nearby Jefferson Lab. Much of the numerical and data analysis is done within the CTEQ-JLab collaboration that includes colleagues from CNU, JLab and around the world.

Hardware and Software Development

The nuclear physics group has taken on responsibility for the construction, testing and commissioning of the Coordinate Detector, which is a crucial component of the JLab Hall A SuperBigBite experiment detector package. Monaghan leads the hardware development side, while Brash is leading the data acquisition, analysis and simulation software efforts. Heddle leads a team of researchers on the development of software for the newly constructed CLAS12 spectrometer in Hall B, with a focus on event visualization.

Our group is also involved in a collaborative effort with Old Dominion University, the University of Virginia and JLab in the development of the CLAS12 polarized target to be used in longitudinal spin structure and other experiments in Hall B.

Learn more about experiments and theory at Jefferson Lab

• Hall A (Brash and Monaghan)
• Hall B (Heddle and Phelps)
• Hall C (Brash and Monaghan)
• CTEQ-JLab Collaboration (Accardi)

Contacts
Edward Brash (edward.brash@cnu.edu)
David Heddle (david.heddle@cnu.edu)
Peter Monaghan (peter.monaghan@cnu.edu)
William Phelps (william.phelps@cnu.edu)
Alberto Accardi (alberto.accardi@cnu.edu)

Christopher Newport is an active member of the Laser Interferometer Gravitational-Wave Observatory (LIGO) Scientific Collaboration (LSC), whose continuing mission is to discover new gravitational wave signals from astrophysical events such as colliding black holes and neutron stars. Starting with the initial detection of gravitational waves (GW) from a binary black hole merger called GW150914, and now continuing with the first detection of GWs from a binary neutron star merger associated with a gamma-ray burst (GRB) called GW170817, the era of true GW astronomy has begun. Our group focuses on three specific areas within the field of GW astronomy, and undergraduate and master's students are actively involved in all three areas of research:

Detector Characterization

Dr. Fisher works to improve the ability to distinguish actual detections of GWs from noise in the LIGO interferometers as part of the Detector Characterization (DetChar) group within the LSC. This effort is critical to increasing the number of detected GW events, because the noise in the LIGO instruments limits the distance to which GW events can be detected. Dr. Fisher's group is exploring machine learning approaches to identifying transient noise events.

Multi-Messenger Astronomy

Dr. Fisher focuses on multi-messenger astronomy using gravitational wave signals as a tool for studying transient astrophysical phenomena. The detection of GWs from the binary neutron star merger, GW170817, when combined with the coincident GRB detection, GRB 170817A, and the electromagnetic follow-up campaign that discovered associated afterglows across the electromagnetic spectrum, has had a significant impact on the understanding of the progenitors of short GRBs, kilonovae, and the equation of state of neutron stars. This collection of observations and discoveries have demonstrated the capabilities of multi-messenger astronomy with the addition of GWs. Dr. Fisher's group is searching for GWs associated with newly detected GRBs or fast radio bursts (FRB), which are extremely short-lived and extremely energetic bursts of radio waves first detected by radio telescopes in 2007. The progenitors of FRBs are completely unknown, and many astronomers are actively trying to detect more of these events to learn what might cause them.

Contacts
Ryan Fisher (ryan.fisher@cnu.edu)

Smart Grid and Renewable Energy Sources

The electric power industry is going through a transformation, moving from a centralized, producer-controlled system to one that is less centralized and more consumer-interactive (“Smart Grid”). This transition leads to changes in various aspects of power system analysis, including modeling and control design.

Our research group develops and applies technologies, tools and techniques to make the future electricity grid more efficient, reliable and secure. One particular focus is on modeling and control design of distributed generation (DG) units and renewable energy sources (RES) in microgrid systems with the goal to achieve high energy efficiency, low environmental impact and uninterruptible outputs.

We also investigate task scheduling algorithms for energy management systems, to optimize the use of power by categorizing and prioritizing loads in a way that an efficient load dispatch is achieved. Our goal is to develop algorithms that simultaneously improve power system reliability, operability and intelligence. We focus on systems capable of real time “intelligent” monitoring through the use of distributed smart devices and on optimized handling of power consumption by analyzing smart devices data.

Multi-Agent and Cyber Physical Systems

Multi-agent and cyber physical systems are key components of modern distributed smart grids, introducing the actual “intelligence” into the power systems and, thus, enabling the coordination between power sources and demands and the optimized distribution of power.

Our group investigates the potential of such decentralized techniques as well as their implications for cybersecurity. For example, we have developed optimized design procedures for local controllers for voltage and frequency regulation in islanded microgrids under stochastic conditions that result from the variability in distributed and renewable sources. We have also derived defense strategies to improve the resiliency of such systems against attacks on the underlying communication infrastructure.

Contacts
Farideh Doost Mohammadi (farideh.d.mohammadi@cnu.edu)
Hessam Keshtkar (hessam.keshtkar@cnu.edu)
Abhishek Phadke (abhishek.phadke@cnu.edu)