Electrical, Computer, Software, and Systems Engineering Research Projects

PI: Dr. Radu Babiceanu
Co-PI: Dr. Omar Ochoa, Co-PI: Dr. Keith Garfield, Co-PI: Dr. Krishna Sampigethaya

Aviation and aerospace cybersecurity is of critical importance to the Nation. As a key component of the overall U.S. transportation infrastructure, it protects people and contributes to American prosperity and leadership. This project promotes workforce development in this vital sector by building on undergraduate and graduate cybersecurity programs at Embry-Riddle Aeronautical University (ERAU), where both ERAU campuses (Daytona Beach, FL and Prescott, AZ) have a history of collaborative education and research activities within the aviation and aerospace cybersecurity domain. Known locally as "Cyber Eagles," the project will advance the collaboration ecosystem across education programs and research centers to prepare students for productive cybersecurity careers and leadership roles in federal and state agencies. The program will recruit diverse scholars and create a supportive environment through effective mentorship, a well-developed curriculum, student involvement activities, and research experiences. These project components will help establish a pathway that enables students to participate in an environment where they can excel and enter a rewarding career in government aviation and aerospace administration agencies.

The project aims to develop a high-skilled workforce to cover the Nation’s needs in the area of aviation and aerospace cybersecurity, focusing on the safety-criticality aspects of airborne systems and the protection of associated hardware and software assets. The project will fund 20 scholarships to students over a five-year period. Student scholars will benefit from the strong ties that ERAU has with Federal and state aviation and transportation administration agencies and the aviation and aerospace industry. Scholars will have the opportunity to meet and learn from top cybersecurity engineers and managers from government and industry through aviation and aerospace-themed projects, events, and symposia hosted by ERAU. Furthermore, the project will take advantage of on-site expertise at ERAU in all computation and communication services related to flight operations, including airborne hardware and software, avionics equipment, and network and communication data links among aircraft, ground stations, radar systems, and satellite systems. This expertise places the scholarship students in a unique position to contribute to cybersecurity protection during the design, development, and operation stages of systems specific for the aviation and aerospace domain.

This project is supported by the CyberCorps® Scholarship for Service (SFS) program, which funds proposals establishing or continuing scholarship programs in cybersecurity and aligns with the U.S. National Cyber Strategy to develop a superior cybersecurity workforce. Following graduation, scholarship recipients are required to work in cybersecurity for a Federal, state, local, or tribal Government organization for the same duration as their scholarship support.

PI: Dr. Massood Towhidnejad,
Co-PI: Dr. Omar Ochoa, Co-PI: Radu Babiceanu, Co-PI: Dr. Jay Pembridge

The next generation of engineers will need essential technical and professional skills to solve the complex problems facing society. Changes to how departments operate, the curriculum, and teaching practices in engineering programs are required to better prepare students for the profession. Efforts to implement these kinds of changes are often slowed down by department cultures or faculty attitudes about the amount of time and work that would be involved. In this project the Electrical Engineering and Computer Science (EECS) Department at Embry-Riddle Aeronautical University will implement an innovative approach to become a department that responds quickly to student and industry needs. This approach will apply agile development methods typically used in industry to deliver the best products faster. Agile methods involve working on teams in short cycles which allow shared work responsibility, frequent feedback, and adjustments between cycles. The EECS Department will use the Scrum agile method to organize how the department carries out its normal operations. The department will also embed Scrum agile product development into courses across the curriculum. The new approach will allow faculty to achieve quicker changes and implementation of prioritized items for the department. Examples of prioritized items will include incorporating more evidence-based practices in courses such as just-in-time teaching, case-based teaching, active learning, and peer instruction; fostering inclusive learning environments; updating course materials; revising department procedures; and recruiting diverse students and faculty. Consequently, both faculty and students in the department will gain expertise with this agile professional skill. The project will investigate how the changes to department operations enhance faculty and student experiences. The findings would help inform other engineering departments about practices to improve the education of a diverse student population to be well-skilled engineers for the workforce.

The objectives of this project will be to radically transform the EECS department into an agile department that: 1) develops students into engineers with agile skills desired by industry, and 2) develops an agile faculty culture which models the use of agile practices for students. Faculty will work collectively in Scrum teams to innovate the practices, policies, and culture of the department. Students will use Scrum in individual and team projects throughout the middle two years of the curriculum to progressively build their expertise for the culminating capstone courses in the senior year. The research study will use an explanatory case study design guided by social cognitive theory. Quantitative and qualitative analyses will be performed using data from interviews with faculty and students, feedback from stakeholders, and artifacts from Scrum teams. Research results could lead to transformations in engineering education by offering a model on the novel use of Scrum as an agile organizational practice and its influences on the collective efficacy of faculty. This project is jointly funded by the Division of Undergraduate Education and the Division of Engineering Education and Centers reflecting the alignment of this project with the respective goals of the divisions and their programs.

PI: Dr. Omar Ochoa
Co-PI: Dr. Massood Towhidnejad, Co-PI: Radu Babiceanu, Co-PI: Debarati Basu

This project will contribute to the national need for well-educated scientists, mathematicians, engineers, and technicians by fostering student success and supporting the retention and graduation of domestic, high-achieving, low-income students with demonstrated financial need at the Embry-Riddle Aeronautical University, a non-profit private institution. Over its six-year duration, this project will fund scholarships to 25 undergraduate students to pursue four-year bachelor’s degrees in Computer Science, Software Engineering, or Computer Engineering. Subsequently the scholars will pursue a one-year accelerated master’s degree in one of the following areas: Software Engineering, Electrical, and Computer Engineering, or Cybersecurity Engineering. First-year students will receive up to five years of scholarship support. The project will increase student persistence in STEM fields by linking scholarships with a newly created effective ecosystem that combines evidence-based practices such as faculty mentoring, academic advising, participation in the learning community, professional development activities, guidance in acquiring internships and research opportunities. With the help of mentors, the scholars will create individual development plans outlining their career goals and steps toward achieving those goals. The project will also include the evaluation of the impact of the ecosystem on supporting the academic success of scholars and the identification of best practices and lessons learned. This project will significantly contribute towards creating a model that actively engages students from groups underrepresented in STEM fields of study, broadens participation in STEM, and infuses 25 talented and diverse engineers with advanced degrees in engineering into the American workforce.

The overall goal of this project is to increase undergraduate and graduate STEM degree completion of domestic, low-income, high-achieving undergraduates with demonstrated financial need in STEM field. Three specific aims guide the project. First is to deliver financial support to domestic, low-income, high-achieving students who will pursue an undergraduate and accelerated master’s degree in engineering. Second is to leverage evidence-based practices to foster student success, increase retention and degree attainment. Third, and finally, is to evaluate the impact of the newly created ecosystem in supporting the academic success of scholars in engineering, and disseminate best practices and lessons learned. Little is known about the factors that affect the academic success of domestic, low-income, high-achieving undergraduate students in engineering fields at a private institution, and how factors such as gender, ethnic background and discipline impact their success, which is the focus of this project. Two research questions will be investigated in this project: (a) Does the academic success of scholars improve across the years by being part of this project? (b) What were the factors effecting the academic success of the scholars, and what are the accomplishments, best practices, and lessons learned from implementing the ecosystem for the scholars? This project is funded by NSF’s Scholarships in Science, Technology, Engineering, and Mathematics program, which seeks to increase the number of low-income academically talented students with demonstrated financial need who earn degrees in STEM fields. It also aims to improve the education of future STEM workers, and to generate knowledge about academic success, retention, transfer, graduation, and academic/career pathways of low-income students.

PI: Dr. Ilhan Akbas
Co-PI: Dr. Laxima Niure Kandel

Establishing cyber analytical capabilities that are common between various elements of the aviation ecosystem is an essential capability that needs to be developed and matured to allow efficient and synchronized use of common data sets, analytical tools and communication backbones across the entire aviation ecosystem. To address these needs, the FAA NextGen Organization has established the CSDS research program with emphasis on discovery, assessment, adaptation, demonstration and transfer of cyber technology to enhance information cybersecurity for elements of the aviation ecosystem. The research is focused on Artificial Intelligence and Machine Learning (AI/ML) techniques to address these aviation ecosystem cybersecurity needs using customizable algorithms and tools. Our collaborators in this project include Astronautics, Collins Aerospace, GE Aerospace,  Gulfstream Aerospace, Massachusetts Institute of Technology Lincoln Lab, The Boeing Company, The Port Authority of New York and New Jersey United Airlines.

PI: Dr. Eduardo Rojas

Antennas are key components of ubiquitous wireless communication, radar, and navigation systems that affect widespread societal needs, such as aerospace systems, healthcare, and space exploration. Most of the antennas used in a variety of applications including cellular phones to unmanned aircraft systems (UAS) are based on flat planar structures or wire geometries that are developed using traditional manufacturing technique. This approach does not allow designers the opportunity to fully leverage the geometry, space, and materials available to design better performing antennas. The focus of this CARRER development project is on an emerging antenna fabrication technique that combines additive manufacturing (AM) and pulsed laser machining that has the potentials to fundamentally alter the existing state of the art. The proposed research will allow engineers to implement smaller, efficient, lighter, and reconfigurable antenna embodiments in three-dimensions (3D) for future applications with increasing complexity. The research proposed in this project is fully integrated with an education and outreach plan. The educational plan will impact the next generation of professionals by exposing high school students to hands-on activities and videos to explain basic antenna engineering concepts. The videos will be made by accomplished engineers in the engineering field to have a strong role-model-based motivational component to stimulate them to pursue STEM careers. An advanced cellular phone-based teaching tool that allows engineering undergraduate students to visualize complex 3D concepts in electromagnetics and antenna engineering is also proposed.

The overall goal of this project is to pursue the discovery of the next generation of antennas with reconfigurable performance while conserving size, weight and cost. Research initiatives include: (a) the investigation of novel additive manufacturing processes for the fabrication of conformal 3D multiple curved antennas based on laser-enhanced direct print AM (LE-DPAM) with femtosecond laser machining and 5-axis kinematics, (b) the study of bio-inspired 3D superior antenna geometries that are not possible to manufacture using traditional methods but are conceivable using LE-DPAM, (c) the development of design methods based on a novel 3D to 2D conformal mapping technique, (d) the study of embedded material- and IC-based reconfigurability mechanisms including the use of electrically tunable inks that can be deposited on conformal surfaces, as well as IC-based switches for reconfiguration of antenna feeds and loads, and (e) the investigation of the structure-property relationships of commercially available and custom-formulated inks that provide excellent electromagnetic performance while addressing the needs for aviation and space environments.

PI: Dr. Laxima Niure Kandel
Co-PI: Dr. Ilhan Akbas

This funding institutes a Research Experience for Undergraduates (REU) Site at Embry-Riddle Aeronautical University (ERAU). Each year, over the summer, ten highly motivated undergraduates will conduct an intense 10-week Unmanned Aerial Vehicles (UAV) cybersecurity research program complemented by professional development activities that prepare them for future cybersecurity careers and graduate schools. Students will research existing UAV cyber threats and mitigation strategies and explore new techniques and algorithms to safeguard UAV systems. The REU program will focus on providing unparalleled opportunities for undergraduate students, especially those from underrepresented and minority groups and from institutions with limited resources, by engaging them in real-world cybersecurity research of UAVs. Through small-group, high-quality mentoring practices, the REU training will not only aid in enhancing the safety and security of UAVs in personal and commercial applications but will also build research confidence among REU participants.

The overall objective of this project is to immerse undergraduate students in research-intensive training in the cybersecurity field and encourage them to think creatively and independently through hands-on project activities. REU participants will be engaged in faculty-led projects such as UAV cyber-attacks, UAV cyber defense mechanisms, privacy protection methods for UAV communications, and Physical Layer-based cybersecurity. They will participate in activities that range from literature reviews, technical seminars, and workshops to the preparation, presentation, and dissemination of research findings. The three major goals of the REU Site are: (1) to expose undergraduate students to a variety of cybersecurity projects that are bound to build the interest, skills, and knowledge necessary to pursue cybersecurity careers; (2) to increase the number of underrepresented undergraduates in cybersecurity and STEM fields through diversity recruitment emphasis, and (3) to provide undergraduate students with strong professional skills for their future careers and graduate schools. The REU Site will leverage ERAUs? state-of-the-art facilities, research labs, and faculty expertise to promote interest in cybersecurity and develop research skills of the undergraduate students which, in turn, will contribute towards cybersecurity education, training, and workforce development.

PI: Dr. Ilhan Akbas

The overall purpose of this project is to inform rulemaking and standards development regarding potential Right of Way (RoW) concepts for manned and unmanned aircraft in the low altitude environment. 

There are various RoW standards, which apply to specific types of UAS. However, there is ambiguity for other UAS and rules have yet to be developed for interactions between two unmanned aircraft or for UAS swarms. RoW rules impact UAS Detect and Avoid (DAA) requirements and the development of industry standards. This research project explores RoW for diverse UAS operations and make safety-based recommendations for consideration by FAA and UAS standards bodies. Our collaborators in this project are University of North Dakota and University of Kansas.

PI: Dr. Richard Stansbury

ASSURE or the Alliance of System Safety for UAS through Research Excellence is a multi-university center designated by the Federal Aviation Administration (FAA) as its Center of Excellence for Unmanned Aircraft Systems established in 2015. As a core and founding member of ASSURE, ERAU sponsorship to conduct research enabling the integration of unmanned aircraft systems (UAS), advanced air mobility (AAM), and urban air mobility (UAM) in the National Airspace System (NAS). New funding opportunities come available 1-3 times per year.
ERAU has completed or is conducting research tasks addressing the impact of maintenance induced failures on UAS safety; the function allocation of systems operations between automated systems, remote pilots, and support crew, surveillance criticality for detect, and avoid systems; impact of UAM air traffic on air traffic controllers; data analysis to determine the impact of UAS on the NAS, UAS flight data recorder requirements, etc.

PI: Dr. Shafika Moni

This research project focuses on designing efficient privacy-preserving authentication schemes to reduce the communication and computation overhead related to authenticating entities in VANET. We have exploited Merkle Hash Tree and Modified Merkle Patricia Trie (MMPT) to overcome the performance limitations of the conventional approach of authentication, such as the ECDSA algorithm for efficient authentication of Road Side Units (RSUs) and Vehicles in VANET. A detailed security analysis is carried out to demonstrate the effectiveness of our authentication scheme against message modification attacks, replay attacks, and message injection attacks. Performance evaluation also shows that the proposed scheme has a significantly lower authentication overhead than other related schemes. 

PI: Dr. Shafika Moni

Most of the Public Key Infrastructure (PKI) based security and privacy solutions for VANETs use pseudonyms where each vehicle gets multiple identities to improve privacy significantly. A vehicle needs 720 pseudonyms in 24 hours and 262,800 pseudonyms in 1 year, according to the US-based SAE J2735 standard. Trusted Authority (TA) revokes all the pseudonyms assigned to a malicious vehicle and stores them in the CRL. However, the overhead of maintaining such a large volume of identities is overwhelming for traditional CRL-based solutions. In turn, it incurs a higher delay to update and broadcast the CRL periodically. We have designed a novel approach by leveraging the Cuckoo Filter to reduce the storage, computation, and communication overhead associated with the CRL in VANET. The cuckoo filter contains only one entry for all pseudonyms of a revoked vehicle, thereby minimizing the overhead associated with CRL verification. Our scheme also provides an efficient lookup operation for vehicles and Road Side Units (RSUs) in a Vehicle to Infrastructure (V2I) scenario.

PI: Dr. Bryan Watson

Electrical distribution needs to protect society by providing reliable power, even under changing conditions. The current approach to design electrical distribution grids often focuses on steady state design requirements or response to a subset of potential faults. Even small and gradual changes in loading, however, can cause voltage transients and lead to major blackouts due to voltage collapse. As electric demand increases and infrastructure operates near its design limits, these events are likely to become more common. While designers can examine slowly changing load transients, this occurs after creating a model of the proposed grid, which can be costly. Thus, this research examines the following gap: A cost-effective approach is needed early in the electrical distribution design process to screen candidate architectures for their expected response to slowly changing operating conditions. 

There is an opportunity to examine unexpected voltage collapse through the lens of ecosystem critical transitions. Critical transitions occur when an ecosystem shifts suddenly from one stable configuration (e.g. forest) to another (e.g. grassland) due to slowly changing environmental conditions (e.g. annual rainfall). The mathematical framework established to evaluate and classify critical transitions has been well studied but has not been used to design electrical distribution. The central hypothesis examined in this proposal is: If we screen initial electrical distribution architectures with graph theory (Ecological Network Analysis), then the resulting designs will have improved critical transition performance over non-screened architectures. Critical transition performance has two aspects: 

1) superior ability to absorb additional loading before voltage collapse (i.e. margin to critical transition), and 
2) transition to desirable, stable secondary configurations following voltage collapse, rather than cascading throughout the system and causing a complete blackout (i.e. type of Bifurcation).

The objective of this research is to develop and validate a new approach to design-for-transient resilience that provides additional insights, is less expensive, and can be used early in the design process.

PI: Dr. Bryan Watson

Modern aerospace systems often approach problems by connecting many smaller agents, rather than using a single, more expensive platform. For example, it is often advantageous to have a fleet of lower-cost UAVs searching an area than a single, highly capable platform (airship). These sophisticated networks, however, are vulnerable to cascading faults.  For example, errors in data from a single UAV could lead the entire search party away from their intended target. Although recognized as a vulnerability for multi-agent systems, current fault-mitigation methods have significant limitations. Centralized monitoring methods are too computationally expensive and do not work well at large scale, while solutions that rely on agents reporting their own failures may not work in situations where the units are under attack or experiencing certain types of faults (e.g. communication failures). Additionally, current approaches often have strict assumptions that may not apply in real-world systems. As a result, large-scale aerospace systems are at risk of individual agent failures that can spread throughout the entire network, causing problems with system operation, and putting personnel in danger. This proposal examines the issue of faulted-agent mitigation through the lens of Biologically Inspired Design. The objective of this research is to investigate and evaluate a new biologically inspired approach to increase multi-agent system resilience. The Ophiocordyceps camponoti-rufipedis (OCR) or Zombie Ant Fungus provides an example of fault resilience in nature. The fungus infects the ant's nervous system and alters their behavior, ultimately leading to death. However, ant colonies have developed a unique foraging and organizational structure that contains the spread of the fungus. The central hypothesis is that an examination of colony response to OCR will allow derivation of information sharing protocols to increase multi-agent system resilience to fault propagation.

PI: Dr. Bryan Watson

Modern aerospace systems need a new approach for swarm consensus that is distributed, operates with local knowledge, and uses simple agents. The overarching goal of our research is to advance our understanding of bed bug behavior and use this understanding to improve performance of aerospace swarms. The first step is to understand individual bed bug response to stimuli (CO2, heat, light) and individual neural characteristics, before considering group dynamics. The objective of this research was to establish a collaboration between biologists and engineers at ERAU to design and implement a test-platform to enable new data collection for bed bug movement. This collaboration begins by examining individual bed bug response to CO2 concentration. Our central hypothesis is that if we record bed bug response to CO¬2 exposure, then we will be able to improve our understanding of collective decision making because the bed bugs coordinate their response to environmental conditions. The research involved five undergraduate students from three campuses.

PI: Dr. Rumia Sultana

We consider a secret sharing model where a dealer shares a secret with several participants through a Gaussian broadcast channel such that predefined subsets of participants can reconstruct the secret and all other subsets of participants cannot learn any information about the secret. Our first contribution is to show that, in the asymptotic blocklength regime, it is optimal to consider coding schemes that rely on two coding layers, namely, a reliability layer and a secrecy layer, where the reliability layer is a channel code for a compound channel without any security constraint. Our second contribution is to design such a two-layer coding scheme at short blocklength. Specifically, we design the reliability layer via an autoencoder, and implement the secrecy layer with hash functions. To evaluate the performance of our coding scheme, we evaluate the probability of error and information leakage, which is defined as the mutual information between the secret and the unauthorized sets of users channel outputs. We empirically evaluate this information leakage via a neural network-based mutual information estimator. Our simulation results demonstrate a precise control of the probability of error and leakage thanks to the two-layer coding design.

PI: Dr. Ilhan Akbas

To better understand the behavior of multiple sea turtle species along Florida’s Space Coast, we teamed up with Northrop Grumman and the Brevard Zoo to launch a drone-based surveillance effort. The Turtle Tech project, leveraging two different unmanned aircraft systems (UAS), aim to provide conservation insights by fine-tuning the operations and computer vision systems for identification of individual sea turtles, including their species, gender and even unique markings.

PI: Dr. Ilhan Akbas

Validation and Verification (V&V) of Artificial Intelligence (AI) based cyber physical systems such as Autonomous Vehicles (AVs) is currently a vexing and unsolved problem. PolyVerif is an open-source solution focused on V&V researchers with the objective of accelerating the state-of-the-art for AV V&V research. PolyVerif provides an AI design and verification framework consisting of a digital twin creation process, an open-source AV engine, access to several open-source physics-based simulators, and open-source symbolic test generation engines. PolyVerif’s objective is to arm V&V researchers with a framework which extends the state-of-the-art on any one of the many major axes of interest and use the remainder of the infrastructure to quickly demonstrate the viability of their solution. 

PI: Dr. Ilhan Akbas

Autonomous Vehicle Validation and Verification AV V&V testing produces multi-variate time series data as output, which is evaluated to determine testing coverage. The recent surge in interpretable Artificial Intelligence (AI) research has resulted in Python interfaces for modern interpretable AI implementations. In this project, various modern interpretable AI implementations will be applied to AV V&V testing data to interpret parameter impact, and generate an informative report of AV V&V scenario using data generated from a traffic simulator and AV V&V test scenarios.

PI: Dr. Daewon Kim
Co-PI: Dr. Eduardo Rojas

This grant supports fundamental research on a radical transformation of additive manufacturing through digitally connecting machines, humans, and manufactured products. Additive manufacturing has enabled a new paradigm shift from conventional design for manufacturing approaches into manufacturing for design. A fundamental change in additive manufacturing is necessary as we enter a new era of intelligent future manufacturing beyond additive manufacturing. A promising solution is the convergence of wireless embedded sensors with artificial intelligence (AI) and machine learning (ML) data processes, which can transform the way people interact with manufacturing processes, factory operations, optimizing efficiency, and anomaly system detection that could provide critical information about evaluated components and systems. This project opens a new transitional door to perceptive and cognitive additive manufacturing, enabling true internet of things and digital twin, connecting devices and machines in factories with robots, computers, and humans, and every product we manufacture in factories. The grant will also support educational activities to upskill the manufacturing workforce, K-12, undergraduate and graduate students, and the public, significantly influencing diverse populations of all ages and backgrounds.

Transformation to cyber-physical production manufacturing demands advanced process monitoring through distributed sensing beyond the current state of digitally connected machines and robots collaborating with humans. This project seeks to enable unprecedented wireless fingerprinting and sensing of additively manufactured parts by embedding wireless sensors and performing predictive analysis and health monitoring using AI and ML techniques. This project proposes a holistic approach involving four core research tasks: 1) to study the effects of embedding sensors during additive manufacturing; 2) to design embeddable acoustic sensors and insert them during the manufacturing process to read physical parameters; 3) to prove that embedded passive sensor signals can be sensed wirelessly using millimeter-wave antennas, and 4) to quickly monitor and evaluate the state of manufactured products using ML algorithms. This project has the potential to enable next-generation cyber-physical production systems.

PI: Dr. Richard Stansbury

The aviation industry uses flight data recorders (FDR) and cockpit voice recorders (CVR) to investigate accidents and incidents. FDRs record sensor data to provide information about an aircraft’s technical status, while CVRs record sounds from the cockpit to draw conclusions through crew communications and environmental sounds. The American National Standards Institute (ANSI) Unmanned Aircraft Systems Standardization Collaborative (UASSC) standardization roadmap v2.0 indicates that there are significant gaps regarding these flight recorders for UAS. Therefore, the purpose of this project is to close these gaps and define appropriate requirements for FDR and CVR for UAS in the national airspace.

The project is divided into subtasks. The first major step is the literature review of current data recorder standards, technologies, and their requirements for UAS and UAM aircraft. The requirements of various government organizations and institutions are analyzed in this step. The next step is to examine the requirements found. Within this task, it is investigated how applicable the existing requirements are to various categories of UAS. If there are problems adapting these requirements, the corresponding standards will be adjusted. The research will especially focus on test procedures for crash survival, methods for data recording, and the minimum data required.

PI: Dr. Richard Stansbury, Co-PI: Dr. Christopher Herbster

As a cross-disciplinary research project, the EECS department and the Meteorology program have teamed up to apply new machine learning concepts toward prediction of flight delays based upon current and historic aviation weather. The study specifically focuses upon the application of Physics Informed Machine Learning (PIML).

PIML blends traditional scientific mechanistic modeling with ML techniques like deep learning through innovative methodological solutions to address domain-specific challenges and extract insights from scientific data sets. Using the tools from both scientific computing and machine learning, PIML can be used to develop new methods for scalable, domain-aware, robust, reliable, and interpretable learning and data analysis.

To date, the research team has developed a robust aviation big data set on aviation operations within the national airspace system correlated across several FAA provided data sources such as flight information, airport flow control, airport status information, etc. Weather data has also been correlated from airport METAR reports and predictions made by the GFS weather prediction model.

The project initially received funding from the ERAU FIRST program. UNIDATA, an organization under NCAR/UCAR, has awarded ERAU funds to upgrade our high-performance computing resources in the Meteorology lab to provide sufficient compute and storage resources. 

PI: Dr. Radu Babiceanu

The verification process of safety-critical systems must ensure system design performs all intended functionality within the required output ranges and safety limits. It must also ensure that no intended functionality is present having a risk larger than the stated development assurance level. The objective of the AI/ML-based system is to assist with the detection of unintended behavior during operations that results in enhanced online hazard analysis and risk mitigation. Validation and verification techniques must be developed for these systems with the future goal of adopting them in airborne operations.

PI: Dr. Radu Babiceanu

The first research uses the pilot cybersecurity event and risk assessment station located in the Cybersecurity Engineering Lab (LB 131). The station includes a Force Dynamics 401CR flight simulator and a digital twin for scenario development and analysis, and it allows for human systems research on aircraft crew response to external stimuli. The research results are intended to be used to build a training module for aircraft pilots.

PI: Dr. Radu Babiceanu

This research covers unmanned systems deployment in uncertain adversarial environments. Resilient logistics operations call for a holistic and crosscutting approach to proactively address both real-time and persistent adversarial events in several operational areas to outfit mobility platforms, networks, and C2 digital twin to support continued uninterrupted operations. The research proposes the development of robust mobility platforms for UAV deployment and remote maintenance in adversarial environments with predictive logistics guarantees, including platform reliability evaluation, and remote inspection.

PI: Dr. Omar Ochoa

The Semantic Web provides a wealth of high-quality, structured, and contextual data, which can be used to train machine learning models. This can lead to the creation of models, i.e., the engineering of Machine Learning, that adhere to non-functional requirements, which include considerations such as safety, security, and reliability, which are key elements of Software Engineering. These requirements do not concern a system's functionality, but rather its quality attributes. By incorporating these concepts into the engineering of machine learning models, one can strive to create models that are secure, reliable, and exhibit the desired quality attributes. Furthermore, Verification and Validation, or V&V, is integral to successful software engineering, by ensuring that a system is implemented correctly and meets specified requirements. In engineering Machine Learning, it's equally important to define processes and methods to thoroughly test and validate models to ensure they're performing as expected and providing accurate results. Together, the fusion of Software Engineering principles into Machine Learning Engineering, aided by the Semantic Web's capabilities, can bolster trustworthiness in machine learning systems. This trustworthiness ensures that the systems can be relied upon to behave as expected. In essence, by combining these fields, one can develop machine learning systems that are reliable, secure, interpretable, and trustworthy, upholding the core principles of Software Engineering. Our research group focuses on the most recent developments in these areas, i.e., Knowledge Graphs and Large Language Models, to accomplish these goals.

PI: Dr. Thomas Yang

In modern wireless communications, scenarios often arise in which the receiver is required to perform detection of multi-user transmissions on the same channel or suppress co-channel interferers. In these scenarios, signal separation techniques based on statistical properties can be highly effective. However, for wireless systems operating in highly dynamic environments (such as mobile and vehicular communications), the rapidly time-varying channel condition remains a major challenge for block-based signal processing, in which the estimation of statistical properties is performed through averaging over a block of data samples. When the channel parameters change with time, long blocks mean substantial variation of mixing matrices within each block, which inevitably degrades the source separation performance. On the other hand, short blocks render the estimation of signals’ statistical properties inaccurate and biased, thus resulting in poor estimation performance.

We addresses the above-mentioned challenge via the adoption of signal separation algorithms specifically designed for dynamic channel conditions, and artificial data injection applied to short processing data blocks in wireless receivers. Through theoretical and simulation studies, we concluded that the data injection method has great potential in improving signal detection accuracy and/or processing speed for multi-user detection in wireless receivers under dynamic channel conditions. The physical layer security of these mobile communication systems is also being addressed. The research is supported by Air Force Research Laboratory’s Information Directorate (AFRL/RI).