80264 research outputs found
Sort by
Formalizing Human-to-Robot Delegation in Sequential Tasks
Full text linkRobotic systems have achieved widespread adoption across sectors ranging from automotive manufacturing to service domains such as hospitality robots and autonomous delivery systems. Yet in collaborative settings—such as search and rescue, construction, or healthcare— effective human-robot teamwork often depends on the human’s ability to delegate tasks appropriately, especially when the robot plays a passive or reactive role. To do so, humans must understand both the robot’s capabilities and limitations, as well as the demands of the environment. This thesis aims to formalize, model, and analyze human-to-robot (H2R) delegation in sequential decision-making tasks, addressing several fundamental questions: What is the optimal delegation policy given capable robots? Is the user’s current delegation policy optimal, or can it be improved? How does human delegation adapt over time, particularly with increased robot usage experience? Is the robot’s policy structured to support continued, effective use?
To explore these questions, we introduce a formal human-to-robot (H2R) delegation framework for environments where humans remain actively engaged. This framework is paired with a reinforcement learning algorithm that computes optimal delegation strategies to robotic teammates. To reflect realistic human behavior, this framework incorporates latent decision factors—intent and workload—that influence decision-making but are not directly observable. To ground our analysis in real-world human behavior data, we present a multimodal dataset of H2R delegation in a search-and-rescue–inspired environment. We also develop a factorial model of human behavior that jointly predicts actions and estimates latent states, using a combination of imitation learning and variational inference. Finally, we introduce tools for measuring human states during data collection: a novel workload assessment method that enables both direct and continuous measurement, and an active learning framework for efficiently annotating latent states post hoc under constrained expert supervision. The methods and insights developed in this thesis inform the design of training protocols, interfaces, and adaptive robot behaviors that better accommodate natural human delegation and will help humans more effectively leverage their robotic teammates
Improved Scale Prediction for High Calcium Containing Produced Brine at High Temperature and High Pressure Conditions
No full textIncreasing demand for gasoline drives the oil and gas industry into aggressive production. Production associated with deepwater sites further intensifies the challenge associated with high temperature and high pressure (HTHP) issues. These conditions —ranging from 300 to 500°F, with pressures up to 20,000 psi and total dissolved solids (TDS) exceeding 300,000 mg/L—pose significant operational challenges. Of particular concern is the role of calcium, a prominent ion in produced water, which not only forms common scales such as calcite (CaCO3), gypsum (CaSO4·2H2O), and anhydrite (CaSO4) but also affects the solubility of other minerals, including barite (BaSO4) and halite (NaCl). Therefore, accurate scale prediction under HTHP conditions, particularly in mixed-electrolyte brines, is essential for effective scale management.
To address this challenge, a flow-through apparatus was developed to measure mineral solubility under HTHP conditions. Anhydrite solubility was investigated in 0–4 m NaCl and 0.1–2 m CaCl2 brines at 150–200°C and 1,500–21,000 psi. Similarly, barite solubility was measured in the same brine systems over a broader temperature range (25–200°C) and pressures up to 22,000 psi. The resulting data enabled the development of a robust thermodynamic model based on Pitzer theory, capable of accurately predicting solubility across a wide range of conditions. The model demonstrates strong performance, with standard errors of ±0.017 SI units for anhydrite and ±0.014 SI units for barite. Notably, it incorporates binary (Ca–SO4, Ca–Ba) and ternary (Ca–Cl–SO4, Ca–Ba–Cl, Na–Ca–SO4) interaction parameters, which significantly improve prediction accuracy in high-calcium brines.
In parallel, barite and Ba–Sr–SO4 nucleation kinetics were examined to improve predictions of scale formation rates under elevated Ca concentrations ([Ca2+] = 0.025–1.33 m) and Sr2+/Ba2+ ratios of 0–8. Experiments were conducted at 25–70 °C and SI of 1.5–2.33. Based on these kinetic data, semiempirical models were developed to estimate barite and Ba–Sr–SO4 solid solution induction times. These models enhance predictive capabilities for scale onset and support cost-effective inhibitor dosage and the minimum inhibitor concentrations (MIC) needed for complex brine compositions.
Overall, this study offers a comprehensive framework for predicting both the solubility and precipitation kinetics of critical sulfate scales under extreme field conditions. These advancements enable more effective scale control, optimized inhibitor dosing, and reduced operational and environmental risks in deepwater oil and gas production
Division algebras and extended Frobenius structures in monoidal categories
Full text linkDue to the wide range of applications in logic, programming, and quantum physics, adapting algebraic objects to the monoidal setting has become an active area of current inquiry. This thesis adds to this field of categorical algebra by exploring generalizations of division algebras and extended Frobenius algebras in monoidal categories.
Division algebras were first introduced to the categorical setting in attempts to generalize structure results from classical algebra. Extended Frobenius algebras were introduced by Turaev and Turner in 2006 as a way to extend the correspondence between oriented 2-dimensional topological quantum field theories and commutative Frobenius algebras to the unoriented case. In this thesis, we explore the monoidal analogues of these objects. Concerning division algebras, we are especially interested in determining how analogues of the equivalent definitions of division algebras over a field relate in a variety of monoidal settings. We also find categorical and functorial constructions that interact well with division algebras and extended Frobenius algebras, and we use these constructions to produce examples
6.6 On Trust
Full text linkThis entreaty was created as part of The Spirit of Asilomar and the Future of Biotechnology summit (February 23-26, 2025) in Pacific Grove, CA
Overcoming Mesoscale Recording Challenges with Scalable Microfabrication Techniques
No full textImplantable microelectrode arrays have revolutionized neuroscience, promising future breakthroughs in translational research. However, in-vivo applications are constrained due to the mechanical mismatch between neural tissue and the electrode interface, as the relatively high stiffness of the implant triggers an immune response and neuron loss. Traditional substrate material choices (such as silicon or polyimide) sustain only minor strains before failure, further limiting implant lifetime. The relationship between material properties, such as shore hardness and water absorption, and their impact on the performance of patterned structures requires further investigation for various silicones and other semi-flexible substrates. In our work, we explore the potential of silicone-based substrates to bridge this mechanical gap between neural tissue and substrates by examining different mechanical properties to pinpoint the critical factors for achieving fine metal patterning. For this study, we juxtapose NUSIL MED-6019, a significantly higher durometer silicone compared to the gold standard of Sylgard 184. We hypothesize that systematically controlling the material properties of silicones will enable high-density patterning that can mitigate the thermal stresses induced during photolithography and other thermal processes. These experiments facilitate the creation of diagnostic and therapeutic bioelectronics with greater mechanical and electrical robustness
6.9 NGL Proposal for Spirit of Asilomar Fund
Full text linkThe Spirit of Asilomar Fund EntreatyThis entreaty was created as part of The Spirit of Asilomar and the Future of Biotechnology summit (February 23-26, 2025) in Pacific Grove, CA.This entreaty is a call to ensure biotechnology is a tool for global flourishing by establishing the Spirit of Asilomar Fund. Fear caused the 1975 Asilomar meeting to build barriers that ultimately hindered the transformative potential of biotechnologies to enable global flourishing. The 2025 Spirit of Asilomar Summit emphasized that we are at a critical juncture for Biotechnology. The Summit revealed a continued fear of Biotechnology amongst the general public and a growing lack of trust between the public and biotechnologists. Biotechnology as a tool for global flourishing is only possible by considering its cultural, ecological, political, and historical contexts. The research enterprise must empower all people, not just some, to unlock Biotechnology's transformative power. To nurture these interdisciplinary biotechnological futures, we propose the establishment of a Spirit of Asilomar Fund. The Spirit of Asilomar Fund will support the transition to transform the Biotechnology enterprise towards rights- and community-based, interdisciplinary, and culturally-centered approaches to Biotechnology innovation and governance. Our overall goal is to foster the sustained, responsible and mindful development of biotechnologies for global benefit
Coffee & Quality Case Study #3: Wesley Community Center
No full textThe Kinder Institute for Urban Research and United Way of Greater Houston created a program called Coffee & Quality Case Study that works with designated United Way organizations to 1) identify ways to build and bolster the organization's current data-collecting practices and 2) use data to understand and improve program outcomes. The third Coffee & Quality Case Study focused on Wesley Community Center, a social service agency that supports individuals in the Near Northside and broader Houston community across a number of domains, including basic needs, child development, education, senior support, and financial opportunity
Race, Religion, Marriage, and the Making of Black American Muslims
No full textBlack Americans have the highest rate of religious belief and participation in the United States; yet studies on marriage overlook the impact of religion on the marriage choices of Black Americans. Likewise, extant literature on religion and marriage focuses primarily on Christian communities, ignoring the connections between religion and marriage for Muslims, a significant religious minority among Black Americans. To fill these gaps, I interview 44 never-married Black Muslims living in Houston or Atlanta. I ask: How do religion and race shape marriage processes among Black Muslims? And to what extent do Muslim marriage processes shape Black Muslim identity? I outline the multiple ways that Black Muslims seek out marriage and consider how race, gender, and immigrant status intersect with religion to shape Muslim marriage processes. Although Black Muslims draw on Islamic frameworks to give meaning to marriage, they actively negotiate these frameworks to decide how to navigate romantic relationships as practicing Muslims in the United States. Given the sociohistorical legacies of anti-Black racism in the United States, Black Muslims identify significant challenges to marriage within Black communities and look to Islam as a potential tool to address these broader community concerns. While Islam frames their perceptions of marriage, Black Muslims frequently encounter racialized barriers to marriage within the multiethnic Muslim American community. I consider how Black Muslims navigate these racialized barriers, and I argue that through this process, marriage becomes a site of both religious and racial identity formation. I examine how gender shapes experiences with racialized barriers to marriage in the Muslim American community and illustrate how such experiences can push Black women to reinterpret Islamic frameworks of marriage. Additionally, I analyze the marriage processes of immigrant-origin Black Muslims and show the extent to which family-origin and ethnicity shape marriage opportunities and choice. My findings show how U.S. Black Muslims use Muslim marriage to highlight their place as Muslims in American society, while simultaneously articulating their unique position as Black people. I illustrate how the formation of Black Muslim identity compels Black Muslims to see themselves as distinctive in both Muslim American community and in American national contexts
Model-Based Approaches for Understanding Touch Perception in Complex & Multisensory Domains
No full textTwo major research directions are driving the field of wearable haptic systems: providing multisensory haptic feedback to enhance realism in Extended Reality (XR) and enabling haptic communication through wearable displays. Multisensory haptics leverage our perceptual ability to combine or integrate multiple senses, with the goal being to either emulate multisensory touch experiences encountered in the real-world or create touch sensations for interactions with virtual objects that do not physically exist. Haptic communication depends on our cognitive ability to learn cross-modal associations, with the goal of encoding auditory or visual information into haptic forms. Although engineering advances in wearable displays have contributed to each of these research directions, considerations of the interdisciplinary nature of the field of wearable haptics is necessary to achieve greater impact. Currently, there are several limitations in our perceptual and cognitive understanding of touch interactions that require individuals either to combine percepts acquired across multiple senses or to learn to map information across senses. This thesis aims to provide insight into touch perception in such complex and multisensory domains. We aim to gain deeper fundamental understanding of multisensory touch interactions, inform the development of new forms of wearable haptic displays, as well as better assess the overall feasibility of haptic technology intended for complicated use-cases. Three experimental studies involving human subjects employ analysis techniques including perceptual assessments, non-invasive neuroimaging, and computational frameworks to address these aims. First, using functional neuroimaging, we investigate what computations may support multisensory processing of auditory and tactile frequency signals in the brain, which is relevant to surface and texture perception in both the real-world and virtual environments. Second, we characterize perceptual responses to assess a new form of haptics that stimulate the skin via traveling waveforms. Finally, we consider Representational Similarity Analysis of electroencephalography (EEG) to track perceptual shifts that occur in the brain after an individual learns cross-modal associations, which is proposed as a method for tracking haptic learning. Critically, in each investigation we employ modeling techniques to help make sense of these complicated touch interactions. Findings from each study contribute to deeper perceptual understanding of complex and multisensory touch interactions, which will help to inform future goals and the development of haptic displays intended for strategic XR interactions and modes of haptic communication
Mind the gap: Schwartzman groups of dynamically defined operators and related problems
No full textThe spectral properties of Schrödinger operators with ergodic potentials lie at the intersection of dynamical systems and spectral theory. These operators are fundamental in understanding quantum systems in disordered or aperiodic media.
This thesis focuses on gap labeling, a method for identifying admissible spectral gaps based solely on the underlying dynamics of the operator. We establish the sets of possible gap labels predicted by gap labeling theorems for various dynamical systems, including affine automorphisms of compact connected abelian groups, codings of rotations, and quasi-Sturmian subshifts.
For Schrödinger operators with potentials generated by the full shift, we analyze whether all labels allowed by the gap labeling theorem correspond to actual spectral gaps and identify potential obstructions to their realization. In contrast, for codings of rotations, we prove that every predicted gap is realized, extending known results for Sturmian subshifts.
This thesis contains joint work with David Damanik and Jake Fillman