VTechWorks
VTechWorks provides global access to Virginia Tech scholarship, including journal articles, books, theses, dissertations, conference papers, slide presentations, technical reports, working papers, administrative documents, videos, images, and more by faculty, students, and staff. Faculty can deposit items to VTechWorks from Elements, including journal articles covered by the University open access policy. Email vtechworks@vt.edu for help.
Open Access Policy
Virginia Tech's open access policy enables researchers to deposit the accepted version of scholarly articles with no embargo.
Theses and Dissertations
Virginia Tech was first in the world to require ETDs in 1997, and continues to add scans of older theses and dissertations.
Open Textbooks
More than 50 freely available and openly licensed textbooks are among our most downloaded items.
Communities in VTechWorks
Select a community to browse its collections.
Recent Submissions
Urban Sprawl: Effects and Mitigation
Eckman, Colin; Konduru, Vibhu; Lee, Amy; Noel, Suana; Ulicny, Sarah (Virginia Governor's School for Agriculture, 2025-07-20)
This literature review details the negative effects of urban sprawl and strategies to mitigate them. The purpose of this review is to identify the consequences of urban sprawl on agricultural land and to explore possible solutions for improving global health and advancing progress toward meeting the United Nations’ Sustainable Development Goals and the topics connected to the USDA’s Agriculture and Food Research Initiative (AFRI) priority areas. In the review, the focus was to identify and analyze at least two effects that Urban sprawl has on rural or agricultural land and to analyze the techniques used to help lessen the effects on the land. The literature review included case studies from multiple countries to incorporate a variation of global viewpoints and to incorporate the findings applicable to a larger audience. Multiple scholarly sources and peer-reviewed journals published within the past ten years were used to include reliable and updated information in the review. The effects focused on are food insecurity, the global south, and pollution, an effect that occurs globally. The strategies researched to mitigate urban sprawl effects are the use of conservation easements, notably abundant in the United States (US), and urban development boundaries, common in China. Food insecurity is a pre-existing issue in the global south, and the loss of agricultural land amplifies it. This issue is intertwined with the second Sustainable Development Goal, which is to end hunger. Pollution surges when urban sprawl causes increased runoff and vehicle emissions, closely related to the third Sustainable Development Goal, ensuring healthy lives. The conservation easements and urban development boundaries work to prevent rural, agricultural, and forest land from being developed through the use of economic and legal tactics. These methods are closely linked to the USDA AFRI priority area of Agriculture Economics and Rural Communities. This literature review works to provide an overview of the effects and methods, also working to provide connections to global and country efforts to better the communities that suffer from urban sprawl. The main purpose is to inform people about the harm of urban sprawl and suggest methods that policymakers should implement to protect agricultural land.
PERM1-Mediated Metabolic Crosstalk Between the Heart and Skeletal Muscle in Pressure Overload-Induced Heart Failure
Gusinac, Rebekah Thomas (Virginia Tech, 2026-01-23)
Heart failure is a complex syndrome with high mortality, as nearly 50% of patients die within five years of diagnosis. Among its systemic complications, cardiac cachexia–a condition characterized by severe unintentional weight loss due to cardiac dysfunction– serves as an independent predictor of mortality. The early stage of cachexia involves a vicious cycle between the heart and skeletal muscle driven by metabolic dysregulation; however, its underlying bioenergetics remain unclear. PERM1, a striated muscle-specific regulator of mitochondrial bioenergetics, is highly expressed in the heart and skeletal muscle. We previously demonstrated that PERM1 is downregulated in failing hearts; however, whether its downregulation also occurs in skeletal muscle during the progression of heart failure is unknown. To address this, wild-type mice underwent transverse aortic constriction (TAC) for 8 weeks. Cardiac function and body composition were assessed by echocardiography and NMR, and PERM1 expression and metabolomic profiles were analyzed by Western blotting and gas chromatography-tandem mass spectrometry (GCMS).
TAC reduced systolic function and downregulated PERM1 to a comparable extent in both tissues. Global PERM1 knockout (KO) mice exhibited lean mass loss with an increase in adiposity and no change in body weight, indicating sarcopenic phenotype and not cachectic phenotype. Partial loss of PERM1 in heterozygous mice accelerated systolic decline and mortality and modulated metabolomic programs linked to ketone handling, branched and medium chain fatty acid oxidation, malate-aspartate shuttling, amino acid anaplerosis, nitrogen recycling, and membrane/cofactor biosynthesis. In vitro, PERM1 silencing in C2C12 myotubes induced a compensatory shift toward glycolysis. AAV-PERM1 preserved systolic function and remodeled metabolomes modestly in the heart and robustly in skeletal muscle as compared AAV-GFP controls. In summary, this study provides the first coordinated PERM1 downregulation and distinct metabolic alterations, which may contribute to systemic myopathy. These findings highlight PERM1 as a potential regulator of metabolic crosstalk between the heart and skeletal muscle during the progression of heart failure.
Enhancing Chinese Heritage Language Education Through Technology: A Two-Part Design and Evaluation Study
Ding, Ye (Virginia Tech, 2026-01-23)
This dissertation addresses critical gaps in Chinese Heritage Language (CHL) education by exploring the role of technology in bridging literacy-oracy disparities and supporting instructional practices. The project comprises two distinct but interconnected manuscripts that collectively aim to enhance the learning experiences of heritage students and the professional capabilities of CHL educators.
The first manuscript, a design case study, evaluates the implementation of the I Chinese Reader mobile application at a Chinese language school in Virginia through a mixed-methods approach involving 142 students, the study assesses the app's impact on reading proficiency, vocabulary acquisition, and student engagement. Findings indicate significant improvements in reading scores, particularly among intermediate learners, and highlight the importance of culturally relevant content and dialect support features. The study identifies key design principles for heritage learner applications, including the necessity of addressing disconnects between sound and writing systems and providing offline functionality for rural access.
The second manuscript adopts a design and development research methodology to create a research-based CHL Teacher Toolkit. Recognizing the challenges of "material overload" and the lack of structured guidance for technology integration, this study details the systematic design, development, and formative evaluation of a digital resource for K–12 CHL teachers.
The toolkit operationalizes theoretical frameworks such as TPACK and SAMR into practical lesson plans and decision-making matrices. Formative evaluation with practitioners confirmed the toolkit's usability and effectiveness in helping teachers select appropriate digital tools to scaffold literacy and affirm student identity.
Together, these manuscripts contribute to the fields of instructional design and heritage language education by providing empirical evidence on the efficacy of mobile-assisted language learning and practical resources for educators navigating the digital landscape.
Developing an Automated Practice Environment and Feedback Engine for Guided Instruction in the Deformable Bodies Course
Basak, Arinjoy (Virginia Tech, 2026-01-23)
Mathematical problem solving is a fundamental competency demanded by employers and an integral component of engineering education.
We seek to improve instruction in mathematical concept-based engineering courses.
To this end, we created an interactive problem solving environment that provides targeted feedback on a student's problem solving attempts by having them work mathematical exercises in a dedicated intuitive software interface.
We provide syntactic feedback that is independent of the problem statement, as well as semantic feedback.
Semantic feedback is achieved by systematically comparing the ground-truth solution for the exercises provided by the instructor with the student's attempt and reconstructing appropriate error messages based on the differences detected between the solutions.
Our systematic approach is based on analyzing the dependencies between equations used by the student, and comparing either individual equations or a reduced representation of a sequence of equations with each other to identify and report differences.
We conducted a study to evaluate the quality of the feedback provided by the system.
We also conducted surveys on usability of the interface.
Based on the studies conducted, we prove the effectiveness of the system and the engine both as a tool for providing rapid and accurate assessment of correctness of exercise attempts.
Are Vision Large Language Models Road-Ready? Benchmarking and Adapting VLLMs for Safety-Critical Driving Video Understanding
Zeng, Tong (Virginia Tech, 2026-01-23)
Vision Large Language Models (VLLMs) have demonstrated impressive capabilities on general-purpose image and video understanding tasks, such as captioning and visual question answering. However, their effectiveness in specialized, safety-critical domains like autonomous driving remains largely unclear. Autonomous Driving Systems (ADS) must reason reliably about complex, dynamic environments and rare but high-risk events (e.g., crashes and near-crashes), whereas existing multimodal benchmarks are dominated by routine scenes, focus on narrow sub-tasks, and often rely on evaluation protocols that are vulnerable to guessing and position bias. To address these gaps, this thesis introduces DVBench, a pioneering benchmark for safety-critical driving video understanding with VLLMs. DVBench is built from naturalistic driving videos in the SHRP~2 study and organized by a three-level hierarchical ability taxonomy aligned with widely adopted ADS scenario frameworks (PEGASUS and NHTSA). It covers 4,000 five-second clips spanning normal driving, crashes, and near-crashes, and approximately 10,000 multiple-choice questions that probe 25 fine-grained perception and reasoning abilities, including environmental conditions, road and lane semantics, hazard assessment, event severity, and fault analysis. To obtain more reliable measurements than standard single-trial evaluation, we further propose GroupEval, which rotates answer positions and requires consistent correctness across permutations to mitigate position bias. Using DVBench and GroupEval, we systematically evaluate 14 state-of-the-art VLLMs (0.5B--72B parameters). No off-the-shelf model exceeds 40% accuracy, and all exhibit substantial gaps between relatively strong low-level perception abilities and much weaker high-level safety reasoning, indicating that current VLLMs are far from deployment-ready as autonomous decision-makers. We also study the effect of injecting textual domain knowledge (e.g., definitions of driving terms) and observe model-dependent but generally modest performance gains. To probe domain adaptation, we conduct a fine-tuning study on Qwen2-VL models using DVBench-style multiple-choice questions. Supervised fine-tuning on 2,800 carefully curated instances yields accuracy gains of up to 10.94 percentage points (up to 43.59% relative) under GroupEval. Notably, a fine-tuned 7B model achieves 36.04% accuracy, surpassing a 72B off-the-shelf baseline while using an order of magnitude fewer parameters. Ability-level analysis shows that fine-tuning particularly improves visually grounded environmental and hazard-related abilities (e.g., atmospheric conditions, lane positioning, risk and severity assessment), while nuanced geometry and maneuver evaluation remain challenging. Overall, this thesis provides: (i) a safety-critical, taxonomy-driven benchmark for driving video understanding; (ii) a robust evaluation strategy for reducing position bias in multiple-choice testing; and (iii) empirical evidence that targeted fine-tuning can substantially narrow---though not close---the gap between general-purpose VLLMs and the stringent requirements of mission-critical driving applications. We release the DVBench benchmark, evaluation toolkit, and fine-tuning scripts and checkpoints at: https://github.com/tong-zeng/DVBench.git, to support reproducible research and future progress in VLLM-based traffic safety understanding.