62588 research outputs found
Sort by
Artificial Intelligence-Aided Strategies for Structural Health Monitoring and Control
No full textStructural resilience is an essential prospect for engineers to obtain and ensure the structure\u27s ability to withstand or recover from an operational or adverse demand. Seeking endurance of structures can emerge as structural health monitoring to help detect the potential damage, deterioration, or failure to allow a timely intervention or maintenance. It can also emerge when controlling the dynamic responses of a structure due to dynamic loading scenarios. In this dissertation, data-driven approaches to facilitate structural fatigue assessment and structural active control are proposed and performed successfully through comprehensive case studies. In the context of structural fatigue assessment, an artificial neural networks-based methodology is proposed for strain estimation from acceleration measurement along the Gene Hartzell Memorial bridge in Pennsylvania, USA, to circumvent the necessity of strain gauge instrumentation for long-term monitoring. A study on measurement uncertainty, its correlation with the strain estimation framework performance, and the robustness of this framework for a range of added noise levels is carried out. Moreover, the scalability of the proposed strain estimation framework is studied using autoencoders to enable the generalization of the framework beyond the original input-output configuration. In the scope of structural active control, RL-Controller, a data-driven reinforcement learning-based methodology is introduced to enable finding the optimal active control solution for a variety of given structures, loading scenarios, and control systems. The performance of the RL-Controller is then verified for three linear and nonlinear building case studies
The Impact of the COVID-19 Pandemic on the Psychiatry and Psychology Services Provided to Traditional Medicare Beneficiaries
No full textImportance: Understanding how the COVID-19 pandemic affected mental health care delivery is critical. A growing body of evidence has reported the negative effects of mental distress in older adults which was exacerbated by the pandemic. These unprecedented years had a further impact on the mental health professional workforce.Purpose: This work examined the changes in psychiatry and psychology services received by traditional Medicare beneficiaries before and during the pandemic. Methods: Publicly available data was used from the Medicare Physician and Other Practitioners datasets. The change in the number of total services by mental health provider was calculated between 2019 and 2021 among practitioners who billed Medicare Part B in 2019, 2020, and 2021. Descriptive analysis and regression analysis were performed. Results: The number of unique mental health professionals decreased by 14.9% from 2019 to 2021; furthermore, all provider types demonstrated a decrease in the number of total services except for nurse practitioner and physician assistant. The change in the number of services per provider was affected by the total number of services and the number of beneficiaries per provider with dual enrollment. Conclusions: Decrease in the number of mental health providers during the COVID-19 pandemic highlighted the challenge of attracting and retaining mental health providers to serve Medicare beneficiaries. Future efforts should address the shortages of mental health providers to improve access and equity in mental healthcare delivery for Medicare beneficiaries, as well as examining a state-specific change in mental health providers
DoppelgAIngers - Project Summary
No full textDoppelgAIngers displays a false "mirror" image that looks eerily like you: dressed similarly, posed similarly, but who does not exist."Deep fakes" are images or videos that have been manipulated with AI to misrepresent the truth. These images spread virally online, giving credence to conspiracy theories and stoking political, social, and cultural tensions. These images force us to question the nature of visual truth. How can we investigate what we know to be true? How can we reconcile the reality of the physical world with the subjective nature of images? How can we recognize manipulated images when we see them?
DoppelgAIngers is a web-based interactive piece and series of images. This work raises questions about truth in imagery and highlights the uncanny nature of AI-generated images. The piece functions like a "camera" or "mirror" but instead of reflecting a true image, it displays an image that has been passed through AI as if in a game of telephone. The image or video input is described by AI and that image description is fed back into AI as an image prompt. For instance, a person stands in front of a web camera to be met with an unsettling AI doppleganger — a person who looks eerily like them, wearing a similar outfit, posed in a similar way, but who does not exist.
By connecting AI images to personal identity and understanding of self, the work intends to raise conversation around ethics and media literacy. It can be used as a tool for conversation, community engagement, and education around AI images
Data-driven Advanced Computational Materials Modeling
No full textDeveloping scalable, interpretable AI tools integrating generative models to revolutionize computational materials discovery and enhance STEM education.This project addresses challenges in computational materials modeling by creating scalable, interpretable AI tools. Current machine learning (ML) models face high computational costs, difficulty handling complex crystal symmetries, and a "black box" nature that limits interpretability. To address these issues, we aim to integrate nonnegative matrix factorization (NMF) with generative AI. NMF offers structure-aware capabilities, reducing computational costs and incorporating material symmetries, while generative AI produces high-fidelity synthetic datasets to augment limited data. Additionally, uncertainty quantification is included to enhance model reliability. This hybrid approach aims to develop efficient tools for high-throughput materials discovery and property prediction. The project outcomes include scalable algorithms, enhanced interpretability for clearer insights into materials, and advanced methodologies for materials discovery. Open-source tools, user-friendly interfaces, and educational materials will ensure broader accessibility, supporting AI literacy and fostering collaboration. This effort benefits both Lehigh\u27s research community and broader STEM fields by advancing materials modeling and empowering diverse students through accessible AI tools and resources.</p
Exploring Hydrogen Impurities and Their Interactions in Ga2O3 and SnO2
No full textHydrogen has been found in previous studies to be an n-type dopant in Ga2O3 and SnO2 that gives rise to unexpected changes in conductivity.[1,2] Hydrogen passivates shallow donor and acceptor impurities in conventional semiconductors by acting as an acceptor in n-type materials and a donor in p-type materials.[3] However, in semiconducting oxides H can act as an n-type dopant and act as a source of conductivity.[4] Fourier Transform Infrared (FTIR) spectroscopy is a powerful tool for studying the vibrational properties of hydrogen-related defects in semi- conductors[5], and can reveal information about the microscopic structure, atomic composition, and thermal stability of a defect.The beta phase of Ga2O3 has a monoclinic crystal structure and touts an ultra-wide bandgap energy of 4.9 eV.[1,6,7] Infrared (IR) lines observed in β-Ga2O3 at 2540, 2547, 2577, and 2585 cm−1 have been assigned to a Ge-D complex, a VGa-2D center, a Si-D complex, and an Fe-D complex respectively with corresponding H frequencies at 3425, 3447, 3477, and 3490 cm−1 respectively.[8–11] We have performed experiments to help assign the various O-H and O-D lines to specific impurity complexes. Using the experimental data presented in this dissertation and previous work done by others[10,12,13], Professor Fowler has presented plausible models for the OD-impurity complexes investigated here based on a VGa(1) perturbed by an additional impurity.The alpha phase of Ga2O3 boasts a bandgap energy of 5.3 eV, surpassing the band gap of the beta phase.[7,14] An epilayer of α-Ga2O3 was grown with a sapphire substrate since they share a similar corundum crystal structure.[15–17] Our α-Ga2O3 samples have been implanted with either H, D or both H and D. IR lines have been observed at 2431, 2440, and 2472 cm−1. The line at 2431 cm−1 has been assigned to an OD-VGa complex in α-Ga2O3.[18] The line at 2440 cm−1 is due to an OD-VAl complex in the sapphire substrate. The thermal stabilities of the three lines have also been studied. Putting our results together with an accumulation of prior theoretical and experimental work by others[19–26] had helped us to propose plausible defect structures and assignments for the observed IR lines.The last material investigated in this dissertation is SnO2 which has a rutile crystal structure.[27] The H present in the IR line at 3156 cm−1 has been observed previously to be mobile at room temperature on the timescale of weeks.[28,29] Weinvestigated the diffusion of H and determined an activation energy of 0.53 ± 0.05 eV from a plot of ln(k) vs T −1. Plotting the fraction of the initial defects annealed at early times with the square root of time yielded an activation energy of 0.45 ±
0.06 eV.[30] Studies done with TiO2 [31–36], which shares the same rutile structure as SnO2, show a pathway for H to diffuse along the c-axis that we agree with.</p
Does non‐native diversity mirror Earth\u27s biodiversity?
No full textAim Human activities have introduced numerous non‐native species (NNS) worldwide. Understanding and predicting large‐scale NNS establishment patterns remain fundamental scientific challenges. Here, we evaluate if NNS composition represents a proportional subset of the total species pool available to invade (i.e. total global biodiversity), or, conversely, certain taxa are disproportionately pre‐disposed to establish in non‐native areas. Location Global. Time period Present day. Major taxa studied Global diversity. Methods We compiled one of the most comprehensive global databases of NNS (36,822 established species) to determine if NNS diversity is a representative proportional subset of global biodiversity. Results Our study revealed that, while NNS diversity mirrors global biodiversity to a certain extent, due to significant deviance from the null model it is not always a representative proportional subset of global biodiversity. The strength of global biodiversity as a predictor depended on the taxonomic scale, with successive lower taxonomic levels less predictive than the one above it. Consequently, on average, 58%, 42% and 28% of variability in NNS numbers were explained by global biodiversity for phylum, class and family respectively. Moreover, global biodiversity was a similarly strong explanatory variable for NNS diversity among regions, but not habitats (i.e. terrestrial, freshwater and marine), where it better predicted NNS diversity for terrestrial than for freshwater and marine habitats. Freshwater and marine habitats were also greatly understudied relative to invasions in the terrestrial habitats. Over‐represented NNS relative to global biodiversity tended to be those intentionally introduced and/or \u27hitchhikers\u27 associated with deliberate introductions. Finally, randomness is likely an important factor in the establishment success of NNS. Main conclusions Besides global biodiversity, other important explanatory variables for large‐scale patterns of NNS diversity likely include propagule and colonization pressures, environmental similarity between native and non‐native regions, biased selection of intentionally introduced species and disparate research efforts of habitats and taxa
Formative evaluation of Peace Spaces in a middle school
No full textTeaching youth how to self‐regulate is a focus of a trauma‐informed school approach. Peace Spaces, including Peace Corners and Peace Rooms, are a trauma‐informed intervention used to teach self‐regulation skills to students. The current study used a Consensual Qualitative Research design to identify and describe the thoughts, concerns, and experiences of teachers who implemented Peace Spaces within a middle school in a small urban center of Pennsylvania. Results of qualitative data analyses provided insight into teachers\u27 perceptions of strengths and challenges to the implementation of Peace Rooms. Additionally, using the school\u27s archival data, the study examined how students were using the Peace Room. Domains from interviews with teachers included: (1) How Peace Spaces are Used; (2) Reasons for Using Peace Spaces; (3) Benefits of Peace Spaces; (4) Peace Space Strategies; (5) Challenges of Peace Spaces; (6) Attitudes about Peace Spaces; and (7) Future Directions. Themes from analysis of archival data indicated that students used the Peace Room for one or a combination of five stressors, with Peer‐Related Stressors and Mental Health as the top reasons. Collectively, these results suggested Peace Rooms were useful to students who needed a space to self‐regulate and/or talk through their problems with a trusted adult. Finally, the school\u27s data indicated a downward trend in the number of referrals for discipline. , Practitioner points Teachers see strengths and offer recommendations for implementing peace spaces in schools. Students use Peace Spaces to learn/practice self‐regulation skills or to talk to a trusted adult. Peace Spaces in schools can serve to help students with emotional and behavioral problems
Integrated uplift, subsidence, erosion and deposition in a tightly coupled source‐to‐sink system, Pagliara basin, northeastern Sicily, Italy
No full textHow tectonic forcing, expressed as base level change, is encoded in the stratigraphic and geomorphic records of coupled source‐to‐sink systems remains uncertain. Using sedimentological, geochronological and geomorphic approaches, we describe the relationship between transient topographic change and sediment deposition for a low‐storage system forced by rapid rock uplift. We present five new luminescence ages and two terrestrial cosmogenic nuclide paleo‐erosion rates for the late Pleistocene Pagliara fan‐delta complex and we model corresponding base level fall history and erosion of the source catchment located on the Ionian flank of the Peloritani Mountains (NE‐Sicily, Italy). The Pagliara delta complex is part of the broader Messina Gravel‐and‐Sands lithostratigraphic unit that outcrops along the Peloritani coastal belt as extensional basins have been recently inverted by both normal faults and regional uplift at the Messina Straits. The deltas exposed at the mouth of the Pagliara River have constructional tops at ca. 300 m a.s.l. and onlap steeply east‐dipping bedrock at the coast to thickness between ca. 100 and 200 m. Five infrared‐stimulated luminescence (IRSL) ages collected from the delta range in age from ca. 327 to 208 ka and indicate a vertical long‐term sediment accumulation rate as rapid as ca. 2.2 cm/yr during MIS 7. Two cosmogenic 10 Be concentrations measured in samples of delta sediment indicate paleo‐erosion rates during MIS 8–7 near or slightly higher than the modern rates of ca. 1 mm/yr. Linear inversion of Pagliara fluvial topography indicates an unsteady base level fall history in phase with eustasy that is superimposed on a longer, tectonically driven trend that doubled in rate from ca. 0.95 to 1.8 mm/yr in the past 150 ky. The combination of footwall uplift rate and eustasy determines the accommodation space history to trap the fan‐deltas at the Peloritani coast in hanging wall basins, which are now inverted, uplifted and exposed hundreds of metres above the sea level