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
Agriculture and the Environment: Impact of Climate Change and Conservation Policy
Wen, Yuanyuan (Virginia Tech, 2026-11-11)
Agricultural production depends on the environment. Climate change gradually shifts grow- ing conditions, requiring producers to adapt, while conservation policies can reshape land use and affect people beyond the land enrolled in the program. This dissertation examines these two dimensions of environmental change related to agricultural production, focusing first on how producers adapt to climate change and then on how conservation policy gener- ates broader environmental benefits.
The first chapter studies perennial crops, which supply nutrients that annual crops lack and store more carbon per acre, yet have received far less attention in the climate literature.
Trees and vines live for decades, so growers cannot switch varieties or change growing sea- son. Moreover, perennial crops have different weather needs during different phases within a season. We construct measures of temperature and precipitation over each crop's four phe- nological phases—dormancy, bloom, fruit development, and fruit maturity—and estimate short-run and long-run responses of yields together using a California county-year panel of almond, walnut, wine grape, and other grape yields from 1951 to 2019. Warming during walnut dormancy and during almond bloom lowers yields in the long run as well as the short run, which means adaptation has not undone the damage. Warming during almond fruit maturity raises long-run yields, a case where growers appear to have turned higher temperatures to their advantage. Wine grape losses in dormancy show up only in the short run. Since adaptation has worked differently by crop and by phenological phase, this result suggests targeted adaptation effort.
The second chapter turns from climate change to conservation policy. China's Grain for Green program pays farmers to convert sloped cropland to forest. We examine whether the program has generated broader societal benefits, in particular improvements in air quality.
Using a 0.1-degree grid panel covering all cropland in 2001 and a difference-in-differences design that compares grids by how intensely the program was carried out, we find that a one-percentage-point increase in program-induced forest coverage lowers annual PM2.5 by roughly 0.08 g/m3. The reduction is about half as large after 2015. Valued with benefit transfer, the improvement carries health gains large enough to matter for the program's efficiency. Because air pollutants travel, these gains reach households that never enrolled in the program, a public benefit that program evaluations have so far left out.
Eyes on the Road! – A Market Assessment Investigating the Characteristics, Capabilities, and Limitations of Production Head-Up Displays
Walters, Jacob; Anderson, Gabrial T. (National Surface Transportation Safety Center for Excellence, 2026-11-11)
This report provides a structured descriptive assessment of head-up display (HUD) implementations and availability across the 2025 U.S. light-vehicle market. It combines a broad market scan with a smaller comparative review of vehicles for which sufficiently detailed materials were publicly available, with the goal of identifying common implementation patterns, documented differences, and recurring design tendencies.
At the market level, the evidence collected indicates that HUDs are now an established premium-to-upper-mainstream feature, but they are still more often used as an equipment differentiator than as a universally standard interface. Within the 20-vehicle review sample, availability was most often tied to trim level, while package-based or add-on pathways were less common. Across the reviewed vehicles, manufacturers overwhelmingly described forward-view, windshield-oriented HUD systems. In practice, this suggests convergence on a common display architecture, with most differentiation occurring in packaging, size, displayed content, color treatment, and adjustment logic rather than in projection concept.
Taken as a whole, the reviewed materials point toward a consistent interpretation of current HUD strategy. Across the sample, the HUD is typically positioned as a forward-view driver-information surface for glanceable, high-priority content and is also used as a usability feature and trim-level differentiator. The evidence in this report does not support a claim of full market uniformity, but it does support the narrower conclusion that recent differentiation is concentrated less in basic projection architecture than in packaging, interface tuning, content mix, and presentation detail.
Fleet Automation: Introducing Automated Driving Technologies to Fleets
Miller, Andrew M.; Britten, Nicholas; Pugliese, Brian (National Surface Transportation Safety Center for Excellence, 2026-11-11)
This research, funded by stakeholders of the National Surface Transportation Safety Center for Excellence (NSTSCE), investigates practical methods for analyzing the integration of Automated Driving Systems (ADS) into commercial vehicle operations (CVOs). Given the complexity of integrating automated technology into trucking fleets, the project applied a hybrid approach combining hierarchical task analysis (HTA), cognitive work analysis (CWA), and contextual inquiry (CI) methodologies to assess operational tasks, workflows, and human factors considerations within a controlled intermodal port environment.
Through extensive site evaluations, interviews, operational mapping, macrocognitive analysis, and Human Reliability Analysis (HRA), the research team identified critical tasks, risks, personnel interactions, and organizational dynamics relevant to ADS implementation. The findings highlighted critical fleet characteristics, such as fleet size, operational complexity, cargo type, and environmental factors, essential for successful ADS integration.
Building upon these insights, the project developed the Automation Implementation Strategies (AIS) tool—a web-based Qualtrics survey that provides a decision-support framework designed for CVOs. The AIS tool provides fleet management with tailored automation recommendations based on four primary areas: fleet operations, localization strategies, automation system behaviors, and human-in-the-loop considerations.
The AIS tool developed in this project emphasizes accessibility and scalability for fleets of all sizes. While large carriers have traditionally partnered with ADS developers, small and midsize fleets often lack the resources to assess automation readiness. The AIS tool addresses this gap by translating complex systems analysis into actionable, context-sensitive recommendations. By enabling fleets to self-assess and explore viable automation pathways, the tool supports broader industry adoption, promotes equity in access to innovation, and contributes to national goals for transportation safety and efficiency.
AI-Powered Weather Detection System for Enhanced Roadway Safety
Thapa, Surendra Bikram; Sarkar, Abhijit (National Surface Transportation Safety Center for Excellence, 2026-11-11)
This study investigates the use of large vision-language models (LVLMs) as a flexible alternative for environmental perception. LVLMs enable image understanding through natural language prompting, reducing the need for task-specific model development. We conducted a comprehensive analysis using the RWVC-BDD dataset, a manually annotated subset of BDD100K with labels for road condition, weather, and visibility. The study examines three modeling paradigms: (1) direct end-to-end LVLM prediction, (2) task-specific concept-based modeling using LVLM-extracted features with classical machine learning classifiers, and (3) combined cross-domain concept modeling that integrates cues across tasks.
Results show that direct LVLM predictions provide a reasonable baseline, particularly for weather classification, but exhibit limitations for more complex tasks such as road condition and visibility, where class imbalance and subtle visual cues reduce reliability. Task-specific concept modeling improves visibility performance and enhances class balance for road condition, demonstrating that structured intermediate representations can stabilize predictions. However, it may reduce performance for tasks requiring holistic scene understanding. The strongest improvement is observed with combined concept modeling, which consistently outperforms both prior approaches and achieves substantial gains in performance across all tasks.
To assess real-world applicability, the framework was further evaluated using traffic camera data from the Virginia Department of Transportation 511 system. Results show strong performance across tasks, particularly under combined concept modeling, highlighting the potential of the approach for infrastructure-based applications such as roadway condition monitoring and weather-aware analysis. Overall, this study demonstrates that LVLMs, when combined with structured concept representations and cross-domain feature integration, provide a promising and scalable framework for environmental perception in transportation systems. The approach reduces engineering overhead, improves interpretability, and enhances performance, making it well-suited for rapid experimentation and deployment in safety-critical applications.
Targeted Modulation of Immune Signaling Pathways in Systemic Lupus Erythematosus
Montano, Hilary (Virginia Tech, 2026-11-10)
Systemic lupus erythematosus (SLE) is an autoimmune disease characterized by hyperactive B cell activity leading to production of autoantibodies and aberrant cytokine signaling. In this dissertation, we focused on improving our understanding of the role of protein signaling pathways in SLE. We modified disease by targeting the adenosine receptor and through germline deletion of targeting the NIK protein. For our adenosine receptor studies, we utilized NZB/W F1 female lupus mice. As these mice develop clinical manifestations of lupus at about 23 weeks-of-age, we injected the mice with an Adenosine 2A Receptor (A2AR) agonist and an Adenosine 2B Receptor (A2BR) antagonist from 21-weeks-of-age to 34-weeks-of-age to determine if modification of the adenosine receptor subtypes would decrease disease severity. Overall, we found that the A2BR antagonist (ADO-5024) was able to improve various disease-associated parameters in these mice, including prolonging survival, while decreasing proteinuria, C3 renal deposition, autoantibody production, and inflammatory cytokine secretion, to be comparable to the standard treatment, dexamethasone. On the other hand, the A2AR agonist, ADO-6001, failed to reduce lupus disease. Treatment with the A2AR agonist, ADO-6001 resulted in disease parameters that were not significantly different from our negative controls. Significantly, ADO-5024 resulted in reduced plasma cell abundance and autoantibody production, demonstrating an important autoimmune role for this signaling pathway.
Using MRL/lpr lupus mice, we targeted the map3k14 gene. We found that Map3k14 deletion on the MRL/lpr background identified contrasting roles for the gene encoding for NIK in systemic immunity vs tissue inflammation. NIK is the master regulator for the non-canonical NF-B pathway which is responsible for many immunoregulatory roles. Although map3k14 KO mice rapidly deteriorated compared to their wild type and heterozygous counterparts, autoantibodies were reduced along with type 2 conventional dendritic cells (cDC2), T helper cells and their subsets, and cytokines like CCL2 and IL-5. However, there was a significant increase of C3, IgG, and SiglecF deposition into the renal tissue leading to elevated proteinuria and reduced renal function. The findings presented in this dissertation will provide a foundation to further examine the role of NIK in immune specific cell populations as NIK may have differential roles on immune cell function. Understanding cell specific pathways regarding NIK function may lead to development of more selective therapeutics for the treatment of SLE.