137091 research outputs found
Sort by
Next Generation Earthquake Monitoring: Harnessing Deep Learning for Enhanced Seismic Phase Detection and Association
No full textAs seismology enters the era of big data, the exponential growth in data volume and processing needs surpasses the capacity of traditional seismic monitoring workflows. The recent success of machine learning applications across various scientific domains has made a paradigm shift in image processing and simple task automation. Within this context, this thesis presents a modern earthquake monitoring workflow with deep learning integrated into different fronts. In the first study, I utilized a deep learning phase picker - EQTransformer and a template matching method for foreshock discovery. I performed a detailed study of the foreshock sequence preceding the 2010 magnitude 6.7 Yushu, Qinghai earthquake in the Tibetan plateau and successfully identified 120 foreshocks with magnitude ranging from -0.7 to 1.6. The foreshock sequence started with an magnitude 4.6 foreshock that occurred approximately 2 hours before at a fault plane roughly perpendicular to the mainshock rupture zone. The observations suggest that extensional step-overs and conjugate faults along major strike-slip faults play an important role in generating short-term foreshock sequences. In the second and third studies, I introduced a novel phase association and location framework tailored for a global-scale seismic monitoring network. The global seismic phase association remains a challenging task due to several factors, such as an inhomogeneous sparse seismic network, the high volume of phase arrivals (comprising both true and false picks), and the large solution space inherent for the global scale. I crafted a framework to tackle these challenges by combining an ensemble deep learning locator, advanced sampling strategies, beam search, and an OcTree grid search algorithm. Through comprehensive evaluations with synthetic and real-world datasets, I demonstrated the framework's effectiveness in associating seismic phases, even in scenarios with multiple events, noise, and overlapping events. During a 9-day trial period in May 2010, this framework recovered up to 93 \% of the events cataloged in the analyst-curated Unconstrained Global Event Bulletin (UGEB) catalog when applied to the phase arrival dataset at the International Data Centre (IDC), while effectively handling up to 88 \% of false picks, despite only using P-waves. Finally, I presented an integration of full waveforms and velocity models through an auto-encoder network and an Eikonet-style deep-learning surrogate model. This work contributes to the modern earthquake monitoring workflow by leveraging deep learning across various aspects of seismic research in the era of big data.Ph.D.Earth and Atmospheric Science
Effects of Inter-Layer Dwell Parameters on Hardness Profiles in Thin Wall Mild Steel WAAM
No full textWire arc additive manufacturing (WAAM) is an emerging technology, and as such the ideal parameter sets for various metals, structures, and applications are still in development. This study examines the inter-layer dwell parameter, which determines how long a deposited layer is left to solidify before the next layer is deposited. The solidification time, interpass temperature reached, and reheating cycles due to subsequent layer depositions directly impact the geometry and material properties of the final structure. A literature review has shown that both the dwell parameters of a constant inter-layer dwell time and of a constant interpass temperature are frequently used in WAAM applications throughout various structures and materials. This study directly compares the two methods of dwell by implementing an in-situ forward looking infrared camera to monitor the inter-layer temperature during depositions. Two sets of experiments were conducted: one set which varied the constant interlayer dwell time between 10 seconds to two minutes, and one set which varied the interpass temperature from 500 °C to 150 °C. Rockwell B Hardness tests were then taken on all wall samples, and geometry was measured. This allowed for direct comparison between the two dwell methods and the relationship between interlayer dwell parameter and resultant hardness, geometry, and total deposition time to be considered. The findings of this study indicate that there is a threshold in interpass parameter control where the improvements to final properties may not be justified by the increase to total deposition time. This is emblematic when comparing the one- and two-minute constant interlayer dwell time samples where the width decrease was 5.4%, average hardness value increase was 0.6 HRB, and increase in total deposition time was 70%.M.S.Mechanical Engineerin
Helium Burning: Affordances of stellar nucleosynthesis to explore timbral plasticity, form and spatialization in computer music
No full textPresented at the 30th International Conference on Auditory Display (ICAD 2025)Helium Burning (2024) is conceived as a generative system, programmed in Max MSP, that addresses the affordances of stellar nucleosynthesis processes to explore timbral plasticity in a computer music composition practice. The triple-alpha process by which three helium nuclei fuse into a carbon nucleus in the core of a red giant star is examined in the piece. A comprehensive four-layer timbral morphing process encompasses methods of transformation of microsounds such as granulation, transposition and timestretching of sound files as well as iterative re-synthesis procedures. As the piece progresses in time, it actualizes the transmutation of timbral substances, evoking the transmutation of atomic nuclei in the core of a star. The successive phases of the comprehensive granular synthesis method are the building blocks of musical structure; form emerges from the timbral morphing process. Thus, Helium Burning delves into astronomical processes as archetypes for conceiving open-ended musical forms. Finally, spatialization is approached as a performative analogy, connoting the essential condition of the observer as an integral part of a quantum phenomenon
"Fellow Travelers": Professionals, Surveillance, and Social Architecture in Postwar America
No full textThis paper examines the surveillance of architectural professionals during the McCarthy era in the United States and explores how the heightened political suspicion of the period influenced their lives and architectural endeavors. In the late 1940s and 1950s, a political consensus that equated Americanism with militant anti-communism dominated all aspects of American life. Amidst this climate, the Federal Bureau of Investigation (FBI) intensified its investigation of American citizens involved in what the agency considered “subversive activities,” including many prominent figures of mid-century Southern Californian modernism, such as Gregory Ain, Rudolph Schindler, Richard Neutra, Robert Evans Alexander, and Garett Eckbo. Focusing on the experiences of Gregory Ain, this paper raises critical questions about the classification of these architects as subversive, extending beyond their political activism. Drawing from declassified FBI files and other historical archives, such as the California House Un-American Activities Committee records, this paper reveals discernible patterns among these professionals. Architects under surveillance were more than “fellow travelers” in political terms; they were interconnected through professional associations, shared architectural interests and concerns, and mutual attitudes toward architecture’s social significance. This paper also investigates the Community Homes project, a cooperative housing initiative designed by Ain that advocated racial integration and progressive design principles, which drew the attention of the FBI. Relevant records indicate concerns about the political nature of the project and its allegedly communist connections. The implications of this research extend beyond the specific architects under surveillance. It argues that their surveillance served as a form of censorship, stifling their socially conscious voices and progressive architectural visions, with enduring repercussions that shaped the history of urban and housing development in Southern California. Understanding the complete scope of the perceived risks associated with these professionals is indispensable for a comprehensive grasp of the intricate interplay between architecture and politics in the early Cold War
Challenging Erasure: Collaborative Architectural Documentation and Historic Interpretation
No full textPresented on February 13, 2025 at 12:30 p.m. in the Tech Square Research Building, Ballroom.Danielle Willkens is an Associate Professor at Georgia Institute of Technology's School of Architecture and a practicing designer, researcher, and educator who is particularly interested in bringing architectural engagement to diverse audiences through interactive projects. Her experiences in practice and research include design/build projects, public installations, and on-site investigations as well as extensive archival work in several countries.Runtime: 55:26 minutesHow can we study and reveal the hidden, or repressed, histories within the built environment that illuminate a more complete and accurate record of our shared history? This talk will feature ongoing documentation, visualization, and historic interpretation work at modern civil rights sites in the southern U.S., and how expanded narratives can pave the way for a more resilient future and inclusive future
Power Delivery and Thermal-Aware Electronic Design Automation Solutions for High-Performance 3D ICs
No full textThe objective of this research is to explore 3D IC design methodologies and develop Electronic Design Automation (EDA) solutions to evaluate and mitigate power delivery and thermal challenges in high-performance 3D ICs with emerging interconnects and advanced CMOS technology. Specifically, we implement tool flows to enable 3D IC physical design with various design and technology assumptions based on existing EDA software and customized algorithms. Using a novel 3D RTL-to-GDS design flow, we improve the performance of heterogeneous 3D ICs, investigate emerging computing systems such as logic-on-memory stacked 3D CPU, and explore the Performance, Power, Area (PPA), power delivery, and thermal impacts of 3D integration.
With a comprehensive analysis of 3D IC power delivery and thermal integrity, we propose EDA methods to quickly analyze thermal sustainability, explore design trade-offs, and generate floorplan and cooling solutions for large-scale 3D computing systems. In addition, we study the trend of power delivery issues in 3D ICs with technology scaling. We identify key power delivery challenges and bottlenecks considering various technology nodes and applications, propose robust 2D/3D Power Delivery Network (PDN) structures and design strategies, and evaluate the system-level impacts on PPA and power integrity.Ph.D.Electrical and Computer Engineerin
Degradation impact on the differing Cross-Linking densities and types between SIS and PGD
No full textThis study investigates how varying crosslinking chemistries and post-thermal curing (PTC) conditions in poly(glycerol dodecanoate) (PGD)-based composites with small intestinal submucosa (SIS) influence thermal behavior, degradation profiles, and mechanical properties for soft tissue repair applications. Shape memory behavior was optimized through PTC, enabling recovery at physiological temperatures, while swelling assays confirmed that SIS thickness significantly impacts crosslinking density. Degradation studies revealed that acrylated PGD (APGD) degraded more slowly than methacrylated PGD (MPGD), correlating with retention of vinyl groups. Mechanical testing further showed that SIS thickness enhanced initial performance but made composites more susceptible to tensile deterioration during degradation—especially in APGD samples. MPGD composites demonstrated post-degradation stiffening, suggesting internal network restructuring. These findings provide critical insight into tuning scaffold properties for minimally invasive surgical delivery and soft tissue regeneration by manipulating polymer chemistry, SIS integration, and processing conditions.UndergraduateBiomedical Engineerin
Distributed Cooperative Estimation and Modeling of Fields with Applications for Underwater Sensor Networks
No full textEnvironmental phenomena, such as algal blooms, wildfire propagation, pollutant dispersion, and more, evolve as continuous spatiotemporal fields that demand real-time monitoring for effective management, rapid response, and/or policymaking. While modern sensing and modeling frameworks have improved environmental monitoring through satellite imagery, sensor networks, and deep learning, underwater environments present unique challenges due to severe communication constraints, costly infrastructure, and the absence of shared research platforms. This dissertation addresses these challenges by co-designing accessible aquatic hardware infrastructure with distributed estimation algorithms for real-time monitoring of spatiotemporal fields.
A primary contribution of this work is the development of μNet, an open-source aquatic testbed that integrates custom acoustic and optical modems, particle-filter-based localization, and miniature underwater robots to enable decentralized sensing experiments. Building on this platform, we develop a comprehensive suite of distributed cooperative filters that transition from relying on known field structures to accommodating unknown field structures. First, we embed known Poisson and advection-diffusion constraints directly into a filter using finite-volume discretization, proving convergence and demonstrating adaptive gradient-based formation control. Next, we extend to unknown parameters through a cascaded architecture that decouples field estimation from online parameter identification of diffusion and flow coefficients while maintaining a distributed implementation. We then generalize the estimation process by augmenting states with arbitrary-order spatial and temporal derivatives, employing extended Kalman filtering to discover governing dynamics from data, and constraining the final estimates by the discovered model. Finally, we enhance the prediction step by embedding a stochastic discontinuous Galerkin solver using polynomial chaos expansion, yielding uncertainty-aware, physics-informed forecasts even during active model learning.
This dissertation bridges the gap between theoretical advances in distributed estimation and practical deployment in resource-constrained aquatic environments. By providing both an accessible hardware platform and a modular suite of algorithms that scale from physics-constrained to data-driven estimation, this work enables accurate, physically consistent field reconstruction under realistic communication and computational limits. The open-source nature of all components—from circuit schematics to algorithm implementations—lowers barriers to entry for underwater research and fosters reproducible, community-driven innovation in environmental monitoring.Ph.D.Robotic
Modeling the emission of energetic neutral atoms at Titan
No full textSaturn's largest moon Titan orbits near the outer edge of the planet's magnetosphere, where conditions vary erratically on timescales ranging from years to tens of minutes. Magnetospheric plasma that rotates with the planet's magnetic field continuously overtakes the moon. Titan's ionosphere causes this magnetic field to pile up and drape around the moon, forming a localized induced magnetosphere. Our present characterization of Titan's induced magnetosphere is largely based on plasma and magnetic field data collected in situ, along the one-dimensional trajectory of Cassini during these flybys. However, it is difficult to place such measurements within the context of the full three-dimensional interaction due to the rapid oscillation of Saturn's magnetosphere. Charge exchange between energetic magnetospheric ions and the moon's neutral atmosphere generates energetic neutral atoms (ENAs), which can be imaged in a manner largely analogous to traditional photography. The Cassini spacecraft took numerous such "photos" of Titan's ENA signature across 126 close flybys. ENA images constitute snapshots of the moon's entire interaction region, observed simultaneously, making these remote measurements advantageous in such a variable environment. It remains challenging to interpret such observations, however, as ENA images contain information on the ambient energetic ion distribution, the electromagnetic environment near Titan, and the moon's atmosphere. A successful disentangling of these influences, which does not yet exist, would provide a major scientific advantage over in-situ measurements alone. Specifically, understanding exactly how the energetic ion dynamics in Titan's induced magnetosphere shape ENA observations is key to deciphering the information embedded in ENA images taken by Cassini. For this purpose, we have developed two new simulation codes which calculate ENA production and detection. Each of our models utilizes a different detection scheme for capturing modeled ENA emissions: first, a hypothetical spherical detector which reveals the entire ENA population, and second, a point-like detector with a finite field of view closely resembling that of Cassini's ENA camera. Using the first scheme, we identify a "belt" of elevated ENA flux that forms a great circle in the plane perpendicular to the ambient magnetospheric field vector. Field line draping attenuates the intensity of ENA emissions into this belt, but does not strongly alter the belt morphology. Using the second model, we generate and compare over 1000 synthetic ENA images of Titan's magnetospheric interaction with data from several Cassini flybys. We find that both the ambient field vector and field line draping can strongly influence the observed ENA signature, and the visibility of the moon's plasma interaction in ENA images is highly dependent on the viewing geometry.Ph.D.Earth and Atmospheric Science