Faculty & Staff Resources

The present von Neumann architecture in which data is cycled between memory and processors is inefficient and costs energy. In this regard, memristors or resistive switching devices which maintain (memory) and change their electrical resistance (switching/computing) in the same device can realize in-memory computing and become potential candidate for next-generation data-centric computing. An ideal oxide-based memristor requires both low switching energy (current/voltage) and good retention. Unfortunately, this has not been realized in the same device in conventional memristors using active metal cations or anions as mobile species. Furthermore, a sharp jump/drop of current over 1~2 order of magnitude is typically seen during switching in existing memristors, which is detrimental to the device for use in training machine learning and neural network for high precision neuromorphic computing. Recently, a memristor based on a new type of mobile species of Ru has realized highly desirable low-current and energy, gradual switching, and good retention simultaneously. However, the underlying mechanism has not been fully understood. This proposal aims at elucidating the mechanism of resistive switching in Ru-based memristors, and identifying new types of mobile species/host materials beyond Ru that can realize even better switching performance, which enables people to choose different materials combinations for different applications. Through integration between research and education, this proposal also aims at improving STEM selection and retention for high school and undergraduate students, and promoting K-12 outreaches through the annual “Engineers Week” in local museums and libraries in DFW area, and through the Texas Pre-Freshman Engineering Program (TexPREP).

The UTALL has been operational since thesummer of 2022 and was established with start-up funding from the university to PI Brown. TheUTALL has state-of-the-art instrumentation capable of performing the most commonly usedluminescence dating methods, as well as several unique and novel instruments of wide interest tothe geology community. Presently, however, the UTALL is operated solely by the PI andstudents. The proposed technician support would enable the UTALL to: expand access to thebroader Earth Science community, increase the scope of research applications, perform methoddevelopment work, and improve student and visitor training. Additionally, the technician wouldhelp maintain facilities, develop more accessible data archiving practices, and engage incommunity outreach activities.

 Unfortunately, the disconnect between the logical constraints that form a model and the examples results in the user getting feedback about the holistic behavior of their model, but does not help the user understand the logical constraints themselves. This stance is supported by a recent user study that found users struggle with refining a model based on observed examples. Our vision is that a live programming environment, which interweaves writing the model and evaluating the model in tandem, is the answer to all of these challenges. If a user can witness the impact of their constraints in real time, then examples become constructive feedback. Moreover, the IDE advancements needed to establish a live development environment will bring finite morel finder IDEs in line with the state-of-the art, which all include features to aid in composition, not just enable composition. All together, these advancements can make software modeling approachable to the average software architect.

Portland cement concrete (PCC) is the backbone of our nation’s transportation infrastructure, thus any strategy for fixing it must intimately involve PCC. Although PCC can have a service life of more than 100 years, in practice this is atypical for a variety of reasons (e.g., material degradation and delayed repair). Extending the service life of PCC is the single-most effective mechanism for reducing life-cycle costs and other socio-economic impact on the transportation built environment. The Tier-1 UTC, Center for Durable and Resilient Transportation Infrastructure (DuRe-Trans), will focus on “Improving the durability and extending the life of transportation infrastructure”. The consortium driving the Center’s mission is comprised of a multidisciplinary team of leading researchers from five universities from across the nation who will carry out an ambitious program to revitalize the nation’s transportation infrastructure, including highways, roads, airfields, bridges, tunnels, and railways. The consortium is well-qualified to address the following strategic topics: I) Durability; II) Construction; and III) Finance. The project is divided into the following Themes: I) Inspection, Maintenance and Preservation, II) Sustainability and Longevity, III) Health Monitoring, IV) Sustainable Materials and Structures for Climate Change Mitigation, V) Advanced Materials and Technologies for Construction and Retrofit, VI) Construction Methods and Management (CMM), and VII) Innovative Revenue and Finance. The Center will lead the nation’s efforts to develop and deploy the next generation of durable concrete-based materials. The performance of these materials will be rigorously tested in both laboratory and field conditions, evaluating the materials’ exposure to various environments (e.g., marine and frost). This research will support the development of standard guidelines for formulation and deployment of the next generation of durable materials. The Center will also dedicate resources for comprehensive research on advanced structural retrofitting and repair solutions for existing infrastructure. Further, the Center will develop and implement cutting-edge, in-place and remote-sensing technologies for structural health monitoring (SHM). SHM tools will provide the ability to sense and signal damage in infrastructure elements, with suitable spatial and temporal resolution with a high degree of accuracy. Such tools will be supplemented with advanced data-driven models to perform life-cycle cost analysis, asset management, and performance characterization. These data-driven models will employ state-of-the-art artificial intelligence techniques, including those based on deep learning methods for data inference, prediction, and optimization. The DuRe-Trans Center will contribute to the development of a workforce trained in interdisciplinary scholarship in order to address the nation’s complex transportation needs. The Center will fast-track the adoption of novel and durable construction materials, the development of standards for reconstruction and retrofitting of aging infrastructure, and the improvement of infrastructure safety, economic efficiency, and service life. Such efforts will help achieve the extended service life of infrastructure, which will create quality jobs, improve communities’ resilience to natural disasters, and provide sustainable industrialization and social good.

To do so, we will (1) facilitate professional development with afterschool AI facilitators to advance their critical consciousness on gendered racism; (2) implementation of gendered racial equity-focused curricula and learning environments; (3) and the co-creation with Black girls to improve gendered racial equity-focused curricula and learning environments. This multi-organizational project engages Black girls and AI facilitators across various knowledge and experiences in AI learning ecosystems. Such a focus attends to many gaps in the research literature pertaining to gendered racialized experiences of Black girls in AI learning ecosystems, specifically, (1) how systemic gendered racism impacts the engagement of this population as AI creators, (2) how Black girls’ individual and collective gendered racialized experiences in afterschool A1 learning ecosystems should inform evidence-based strategies for engaging Black girls in intersectional AI education approaches.

E.A.D.Y- Initial Certification (IC) is to Remove barriers for preservice or practicing teachers, Enrich special education teacher preparation curricula, Advance preservice or practicing teacher leadership skills, Diversify the special education teacher workforce, to Yield positive academic, behavioral, and social emotional outcomes for school-age students with disabilities. Over the course of this proposed project, Project R.E.A.D.Y-IC, we will prepare a total of 30 highly qualified practitioners at the Bachelor’s and Master’s level in special education to hold positions in high needs schools that serve school-age children with disabilities and high-intensity needs.

The RTG program naturally builds on several NSF- and DoED-sponsored mentoring and training programs in Mathematics and articulates well with institutional goals of enhancing interdisciplinary research, education, and community engagement in the area of Health and the Human Condition. Scholars will be trained in mathematical and computational methods for diagnosis, assessment, and treatment of chronic and infectious diseases. A group of 10 UTA Mathematics faculty will fully participate in all program activities, supported by collaborators from UTA Nursing, Bioengineering, Biology, and Psychology, UT Southwestern Medical Center, and UNT Health Science Center. The scholars’ training will be accomplished by joint UTA faculty comentoring from mathematics and the health sciences, RTG peers, UTA support services, and researchers from RTG partner institutions and industry scientists.

The research approach will leverage General Systems Performance Theory(developed by Dr. Kondraske) to address the following objectives: define a set of pilot basic performance resources and associated resource availability measures; define a representative pilot high-level task and associated task performance measures; empirically understand the relationship between resource availability and high-level task performance in a flight simulator; and derive a set of minimum performance resource availabilities. UTA will conduct collaborative research to advance evidence-based practice in aerospace medicine to improve the health and safety of civilian pilots by providing expertise and technical assistance to the FAA in accomplishing human factors and aerospace medical research toward achievement of the above objectives. The Period of Performance is 60 Months after award, although work may be completed prior to the end of the Period of Performance.   Note: A proposal was not submitted in response to a Request for Proposals.  The FAA has provided a Memorandum of Agreement and a Statement of Work aimed at establishing the project.

Perhaps there is no better example of the lack of diversity than in higher education, where it is evident that there are very few black, indigenous or people of color (BIPOC) who aspire to or join the faculty ranks of BIPOC faculty that contribute to STEM higher education, limits perspectives and the likelihood that research will address the needs of all segments of society. Our goal is to increase diversity in STEM professoriate by creating an adaptable model of best practices that provides a strong support system through collaborative mentoring and sponsorship of BIPOC doctoral candidates, postdoctoral fellows and junior faculty in the University of Texas System and across the country.  This proposed project has the support of UT System administrators and is in alignment with the UT System’s mission.  The project is also in alignment with one of the two initiatives of the newly established Alliance for Hispanic Research Serving institutions, a voluntary association of universities that are both Hispanic-Serving Institutions and in the top 5% of universities in the country for research dedicated to increasing the number of Hispanic doctoral students. The University of Texas at Arlington (UTA) will serve as the lead institution. Partnering institutions include the University of Texas at Austin (UT Austin), the University of Texas at Dallas (UTD), the University of Texas-El Paso (UTEP), and the University of Texas at San Antonio (UTSA). All the participating institutions are Carnegie Tier One Universities. Four of the participating institutions are Hispanic Serving Institutions. As a team representing each of the institutions listed above, we have extensive experience in implementing best practices for mentoring BIPOC students, postdocs and junior faculty at our home institutions. We are committed to creating a diverse community of outstanding scholars through inclusive mentoring and creating competitive candidates to hire within the UT System. Specifically, we propose to create an mentoring of best-practices AGEP model to be tested and implemented at UT System institutions that will focus on the development of STEM BIPOC from different states of their career. Toward this goal, we will identify, recruit and organize cohorts of participants and three distinct career stages: doctoral studies, postdoctoral fellowship, and tenure-track faculty appointments.  Our objectives are the following:  Provide inclusive mentoring for UT System AGEP participants around networking opportunities that create a sense of belonging, provide safe spaces to address systemic issues and experiences that adversely affect BIPOC, and make available opportunities to learn “tools for success” to support the maturing of STEM identities.   Gather evidence of the effectiveness of the UT STEM AGEP model for advancement in each one of the three career stages and disseminate results state-wide and nationally.   Promote cultural and policy changes at each participating institution to create an ecosystem supportive of the professional development of BIPOC graduate students, postdoctoral fellows and faculty.  Utilize formative and summative internal and external evaluations to understand the strengths and act on areas of improvement of the UT System AGEP model so that it can be implemented and sustained by the UT System for the long term.   Intellectual Merit: This project will 1) contribute to knowledge of critical interventions that help BIPOC doctoral students acquire key professional skills and successfully complete doctoral programs in STEM fields and proceed to careers in academia, 2) help BIPOC postdoctoral fellows acquire key professional skills to create competitive higher education applications, while contributing to their knowledge in the field of academia, and 3) provide support, mentoring and sponsorship for BIPOC junior faculty so that they can attain tenure in develop into strong mentors themselves in academia. It is anticipated that the results of the training described here will mean that participants will become effective STEM educators amongst the next generation of outstanding STEM leaders.   Broader Impacts: This project will (1) provide a model of best practices for support of BIPOC doctoral candidates, postdoctoral fellows and junior faculty and implemented by UT System into University of Texas institutions.  Texas, having a high population of people from under-represented backgrounds, has the potential to impact STEM higher education in a very significant way with our proposed efforts and thereby serve as a national example for increasing BIPOC success rates in STEM. We will create effective strategies that increases the career success rate BIPOC STEM participants that are applicable to other universities and colleges; (2) provide best practices in leadership training and development for faculty and administrators (3) Increase student preparedness and confidence for completion of their academic programs through mentorship and sponsorship; (4) create dialogue between BIPOC STEM participants and leaders through interactive workshops and conferences (5) contribute to the national goal of increasing the number of BIPOC in STEM with Ph.D. degrees; (6) contribute to BIPOC STEM faculty pool who effectively mentor future BIPOC STEM students. The progress of this work will be monitored through an internal and external evaluation processes and published for the advancement of knowledge for the HigherEd community.  

For example, Q&A systems and chatbots are used for advancing public health, and social sensing systems are used for emergency response, crime prevention, and political campaigns. However, the more applications rely on NLP, the more significant the potential disruption and destruction adversaries can create via attacking them. The computer security research community has progressed in studying the adversarial robustness of deep learning-based systems, especially in image and text classification domains. However, the adversarial robustness of more complex natural language generation (NLG) tasks, such as summarization and question answering systems, and their components, such as ranking algorithms, have been less studied. This CAREER proposal aims to fill this gap by understanding NLG systems’ attack surface and their adversarial vulnerabilities and proposing novel empirical and theoretical methods for increasing their robustness.   In particular, this project includes three interconnected aims for advancing the adversarial robustness of NLG systems, including: (1) developing a framework and proposing novel AI-based optimization methods for implementing and examining the state-of-the-art NLG systems against various attack models, (2) having an in-depth analysis and characterization of vulnerabilities that lead to such attacks, and (3) proposing and developing a set of defensive methods and tools for enhancing the robustness of NLG systems, including detection methods, certified robust training, and trust explainability tools. This research will be integrated with education and outreach by providing research experiences for women and underrepresented groups, incorporating research results into the course content development and curriculum design with the goal of reducing the gap between NLP and cybersecurity programs.