Showing 1–11 of 11 results for author: Gillespie, K

Skeletal Density Measurements for Adsorbent Nanomaterials

Authors: Andrew K. Gillespie, Zeke A. Piskulich, Ernest Knight, Matthew Prosniewski, Peter Pfeifer

Abstract: Due to their tunable material properties, sorptive materials have a wide range of applications in energy storage, water treatment, carbon capture, analytical chemistry, and more. One crucial factor in determining the effectiveness of such materials is their skeletal density, or "true density" because it is often used to calculate key metrics, such as storage capacities. In this paper, we present s… ▽ More Due to their tunable material properties, sorptive materials have a wide range of applications in energy storage, water treatment, carbon capture, analytical chemistry, and more. One crucial factor in determining the effectiveness of such materials is their skeletal density, or "true density" because it is often used to calculate key metrics, such as storage capacities. In this paper, we present skeletal density measurements through helium pycnometry for several types of adsorbent carbon materials derived from either corncob, sawdust, coffee grounds, polyvinylidene chloride (PVDC), graphitic carbon nitride (GCN), or metal organic frameworks (MOFs). The measured skeletal density of sawdust-based activated carbon was 2.02 +/- 0.05 g/$cm^{ 3}$. The measured skeletal density of coffee-based activated carbon was 2.23 +/- 0.06 $cm^{ 3}$. We also expound upon the impact that skeletal density has upon hydrogen excess adsorption measurements and other calculated engineering quantities. If a skeletal density is underestimated by 10%, it can affect the room temperature excess adsorption by at least 5% at 100 bar and by 7% at 200 bar, depending on the material type. △ Less

Submitted 4 June, 2024; originally announced June 2024.

Comments: 12 pages, 4 figures, 2 tables

YORI: Autonomous Cooking System Utilizing a Modular Robotic Kitchen and a Dual-Arm Proprioceptive Manipulator

Authors: Donghun Noh, Hyunwoo Nam, Kyle Gillespie, Yeting Liu, Dennis Hong

Abstract: This article introduces the development and implementation of the Yummy Operations Robot Initiative (YORI), an innovative, autonomous robotic cooking system. YORI marks a major advancement in culinary automation, adept at handling a diverse range of cooking tasks, capable of preparing multiple dishes simultaneously, and offering the flexibility to adapt to an extensive array of culinary activities… ▽ More This article introduces the development and implementation of the Yummy Operations Robot Initiative (YORI), an innovative, autonomous robotic cooking system. YORI marks a major advancement in culinary automation, adept at handling a diverse range of cooking tasks, capable of preparing multiple dishes simultaneously, and offering the flexibility to adapt to an extensive array of culinary activities. This versatility is achieved through the use of custom tools and appliances operated by a dual arm manipulator utilizing proprioceptive actuators. The use of proprioceptive actuators enables fast yet precise movements, while allowing for accurate force control and effectively mitigating the inevitable impacts encountered in cooking. These factors underscore this technology's boundless potential. A key to YORI's adaptability is its modular kitchen design, which allows for easy adaptations to accommodate a continuously increasing range of culinary tasks. This article provides a comprehensive look at YORI's design process, and highlights its role in revolutionizing the culinary world by enhancing efficiency, consistency, and versatility in food preparation. △ Less

Submitted 17 May, 2024; originally announced May 2024.

Comments: This manuscript is 13 pages long, includes 10 figures, and cites 20 references. It is to be submitted

Helium Detection in Technical Materials

Authors: Andrew K. Gillespie, Cuikun Lin, Django Jones, R. V. Duncan

Abstract: Materials used to study nuclear fusion can retain atmospheric helium unless pretreated before an experiment. Understanding helium outgassing is important for accurate diagnostics in experiments surrounding nuclear fusion. The presence of helium is often cited as the primary evidence that a nuclear reaction has occurred, so it is imperative that known sources of helium are mitigated prior to procee… ▽ More Materials used to study nuclear fusion can retain atmospheric helium unless pretreated before an experiment. Understanding helium outgassing is important for accurate diagnostics in experiments surrounding nuclear fusion. The presence of helium is often cited as the primary evidence that a nuclear reaction has occurred, so it is imperative that known sources of helium are mitigated prior to proceeding with novel nuclear experiments. It is also necessary to ensure hermiticity when transferring gas aliquots from an experiment to a mass spectrometer. In this article, we present studies of detecting helium leak rates in systems used in novel nuclear experiments. We also present studies of helium retention in materials subjected to various heating profiles and atmospheric concentrations. Without pretreatment, stainless-steel 316 retains between 15 $\unicode{x2013}$ 240 pmol of $^{ 4}$He or an areal outgassing amount of 0.07 $\unicode{x2013}$ 1.20 pmol/$cm^{ 2}$. It also may reabsorb $^{ 4}$He from the atmosphere in time. These studies also demonstrate that it is necessary to pretreat most materials prior to performing experiments where the presence of $^{ 4}$He is being used as an indicator for novel nuclear reactions. △ Less

Submitted 25 March, 2024; originally announced May 2024.

Comments: 10 pages, 4 figures, 5 tables

Simulation and Experimental Analysis of Aerogel Attenuation for High Energy Alpha Particles in Fission Fusion Fragment Rocket Applications

Authors: Sandeep Puri, Andrew Gillespie, Ian Jones, Cuikun Lin, Ryan Weed, Robert V. Duncan

Abstract: Emerging studies are geared toward exploring new methods of nuclear rocket propulsion to provide more efficient space transit beyond Earth's orbit. One method is to employ a Fission Fragment Rocket Engine utilizing fissionable layers embedded in a low-density aerogel. A quantitative understanding of particle attenuation is essential for develo** a functional prototype that permits fission fragme… ▽ More Emerging studies are geared toward exploring new methods of nuclear rocket propulsion to provide more efficient space transit beyond Earth's orbit. One method is to employ a Fission Fragment Rocket Engine utilizing fissionable layers embedded in a low-density aerogel. A quantitative understanding of particle attenuation is essential for develo** a functional prototype that permits fission fragments to escape the layers and contribute to specific impulse rather than being attenuated and generating waste heat. In this investigation, the MCNP code was used to theoretically analyze the attenuation of alpha particles from Am-241 sources within aerogel materials. Simulations were conducted on aerogels with various densities and compositions. These simulations aimed to predict the expected intensity of alpha particles reaching a detector. CR-39 was employed as a Plastic Nuclear Track Detector to assess particle attenuation by the aerogels. The threshold areal density of atoms was found to be around $10^{ 20}$ atoms/$cm^{ 2}$ for the three materials studied in this project. Using a 0.22 mm thick SiO$_2$ aerogel with a density of 90 mg/$cm^{ 3}$, which exceeds the threshold, nearly all alpha particles were attenuated. Conversely, employing a 1.6 mm thick graphene aerogel with a density of 12.5 mg/$cm^{ 3}$ resulted in an average attenuation of 32.3%. △ Less

Submitted 28 May, 2024; v1 submitted 31 January, 2024; originally announced February 2024.

Comments: 14 pages, 5 figures, 1 table

Integrating Summarization and Retrieval for Enhanced Personalization via Large Language Models

Authors: Chris Richardson, Yao Zhang, Kellen Gillespie, Sudipta Kar, Arshdeep Singh, Zeynab Raeesy, Omar Zia Khan, Abhinav Sethy

Abstract: Personalization, the ability to tailor a system to individual users, is an essential factor in user experience with natural language processing (NLP) systems. With the emergence of Large Language Models (LLMs), a key question is how to leverage these models to better personalize user experiences. To personalize a language model's output, a straightforward approach is to incorporate past user data… ▽ More Personalization, the ability to tailor a system to individual users, is an essential factor in user experience with natural language processing (NLP) systems. With the emergence of Large Language Models (LLMs), a key question is how to leverage these models to better personalize user experiences. To personalize a language model's output, a straightforward approach is to incorporate past user data into the language model prompt, but this approach can result in lengthy inputs exceeding limitations on input length and incurring latency and cost issues. Existing approaches tackle such challenges by selectively extracting relevant user data (i.e. selective retrieval) to construct a prompt for downstream tasks. However, retrieval-based methods are limited by potential information loss, lack of more profound user understanding, and cold-start challenges. To overcome these limitations, we propose a novel summary-augmented approach by extending retrieval-augmented personalization with task-aware user summaries generated by LLMs. The summaries can be generated and stored offline, enabling real-world systems with runtime constraints like voice assistants to leverage the power of LLMs. Experiments show our method with 75% less of retrieved user data is on-par or outperforms retrieval augmentation on most tasks in the LaMP personalization benchmark. We demonstrate that offline summarization via LLMs and runtime retrieval enables better performance for personalization on a range of tasks under practical constraints. △ Less

Submitted 30 October, 2023; originally announced October 2023.

Comments: 4 pages, International Workshop on Personalized Generative AI (@CIKM 2023)

ACM Class: I.2.7; H.3.3

Simulating Radiation Shielding Effectiveness Against Three Neutron Sources

Authors: Andrew K. Gillespie, Cuikun Lin, Matthew Looney, R. V. Duncan

Abstract: Laboratories and Universities regularly apply for approval from the United States Nuclear Regulatory Commission (NRC) to use neutron generators for experimental research. To comply with the regulations set by the NRC, adequate shielding is necessary to ensure that the radiation dose rates experienced by an operator, and outside the walls of any containment buildings, are below the prescribed level… ▽ More Laboratories and Universities regularly apply for approval from the United States Nuclear Regulatory Commission (NRC) to use neutron generators for experimental research. To comply with the regulations set by the NRC, adequate shielding is necessary to ensure that the radiation dose rates experienced by an operator, and outside the walls of any containment buildings, are below the prescribed levels. Typically, the neutron source needs to be shielded such that the radiation dose rate experienced by any user is less than 0.25 mRem/hr (500mrem/2000-hour work-year). To address this requirement, we investigate the effectiveness of boronated concrete, boronated water, and light water shielding materials and their applicability to three neutron sources. We present our findings on the radiation shielding design and calculations for three neutron sources situated inside shielding layers. Our modeling utilized the Monte Carlo n-Particle transport codes (MCNP6.2) to simulate neutron attenuation of the shielding. The simulation results reveal that a light water shielding can sufficiently reduce the dose rate for an individual located as close as two meters from the source. Therefore, this shield design can effectively decrease the radiation dose below the maximum recommended limit. △ Less

Submitted 31 January, 2024; v1 submitted 19 September, 2023; originally announced September 2023.

Comments: 9 pages, 6 figures, 2 tables

Photoneutron Yield for an Electron Beam on Tantalum and Erbium Deuteride

Authors: Andrew K. Gillespie, Cuikun Lin, R. V. Duncan

Abstract: An electron beam may be used to generate bremsstrahlung photons that go on to create photoneutrons within metals. This serves as a low-energy neutron source for irradiation experiments [1-3]. In this article, we present simulation results for optimizing photoneutron yield for a 10-MeV electron beam on tantalum foil and erbium deuteride (ErD$_3$). The thickness of the metal layers was varied. A tan… ▽ More An electron beam may be used to generate bremsstrahlung photons that go on to create photoneutrons within metals. This serves as a low-energy neutron source for irradiation experiments [1-3]. In this article, we present simulation results for optimizing photoneutron yield for a 10-MeV electron beam on tantalum foil and erbium deuteride (ErD$_3$). The thickness of the metal layers was varied. A tantalum foil thickness of 1.5 mm resulted in the most photons reaching the second metal layer. When a second metal layer of ErD$_3$ was included, the photoneutron yield increased with the thickness of the secondary layer. When the electron beam was directly incident upon a layer of ErD$_3$, the photoneutron yield did not differ significantly from the yield when a layer of tantalum was included. The directional photoneutron yield reached a maximum level when the thickness of the ErD$_3$ layer was around 12 cm. About 1 neutron was generated per $10^4$ source electrons. When using a 2-mA beam current, it is possible to generate up to $10^{ 12}$ neutrons per second, making this combination a relatively-inexpensive neutron generator. △ Less

Submitted 4 August, 2023; originally announced August 2023.

Comments: 15 pages, 9 figures, 4 tables

Journal ref: Nuclear Engineering and Technology, 2024

The Fission Fragment Rocket Engine for Mars Fast Transit

Authors: John Gahl, Andrew K. Gillespie, Cuikun Lin, R. V. Duncan

Abstract: In this paper we discuss the advantages and challenges of utilizing Fission Fragment Rocket Engines (FFREs) to dramatically reduce transit time in space travel, for example, traveling to Mars. We discuss methods to decrease the size and weight of FFREs. These include utilizing metallic deuterides as moderators, driving the engines with electron beam bremsstrahlung, and operating the FFREs as subcr… ▽ More In this paper we discuss the advantages and challenges of utilizing Fission Fragment Rocket Engines (FFREs) to dramatically reduce transit time in space travel, for example, traveling to Mars. We discuss methods to decrease the size and weight of FFREs. These include utilizing metallic deuterides as moderators, driving the engines with electron beam bremsstrahlung, and operating the FFREs as subcritical assemblies, not as nuclear reactors. We discuss these and other new innovations based upon improved materials and technology that may be integrated into a revolutionary nuclear rocket technology. △ Less

Submitted 2 August, 2023; originally announced August 2023.

Comments: 10 pages, 2 figures, 2 tables

Journal ref: Front. Space Technol. 4 (2023)

Prospects For A New Light Nuclei, Fission Fusion Energy Cycle

Authors: R. V. Duncan, Cuikun Lin, Andrew K. Gillespie, John Gahl

Abstract: Future advanced nuclear rocket propulsion, and the availability of new nuclear power cycle designs, will benefit substantially from the large current investment in alternative nuclear energy that is underway today. We propose a new nuclear cycle which includes the primary fission of lithium-6, followed by secondary fusion of deuterium and tritium, and a secondary fission of lithium-7 by tritium. T… ▽ More Future advanced nuclear rocket propulsion, and the availability of new nuclear power cycle designs, will benefit substantially from the large current investment in alternative nuclear energy that is underway today. We propose a new nuclear cycle which includes the primary fission of lithium-6, followed by secondary fusion of deuterium and tritium, and a secondary fission of lithium-7 by tritium. This cycle does not produce nuclear waste from its nuclear fuel, since all byproducts of these cascade reactions are stable, provided that the triton production during the primary reaction is fully consumed in the secondary reactions. This cycle may, however, activate surrounding technical materials from its neutron flux. This light-element nuclear fuel is readily obtained through the ongoing expansion of the lithium mining industry and electric vehicle (EV) battery recycling industries. △ Less

Submitted 22 March, 2024; v1 submitted 2 August, 2023; originally announced August 2023.

Comments: 10 pages, 2 figures

Improving Device Directedness Classification of Utterances with Semantic Lexical Features

Authors: Kellen Gillespie, Ioannis C. Konstantakopoulos, Xingzhi Guo, Vishal Thanvantri Vasudevan, Abhinav Sethy

Abstract: User interactions with personal assistants like Alexa, Google Home and Siri are typically initiated by a wake term or wakeword. Several personal assistants feature "follow-up" modes that allow users to make additional interactions without the need of a wakeword. For the system to only respond when appropriate, and to ignore speech not intended for it, utterances must be classified as device-direct… ▽ More User interactions with personal assistants like Alexa, Google Home and Siri are typically initiated by a wake term or wakeword. Several personal assistants feature "follow-up" modes that allow users to make additional interactions without the need of a wakeword. For the system to only respond when appropriate, and to ignore speech not intended for it, utterances must be classified as device-directed or non-device-directed. State-of-the-art systems have largely used acoustic features for this task, while others have used only lexical features or have added LM-based lexical features. We propose a directedness classifier that combines semantic lexical features with a lightweight acoustic feature and show it is effective in classifying directedness. The mixed-domain lexical and acoustic feature model is able to achieve 14% relative reduction of EER over a state-of-the-art acoustic-only baseline model. Finally, we successfully apply transfer learning and semi-supervised learning to the model to improve accuracy even further. △ Less

Submitted 29 September, 2020; originally announced October 2020.

Comments: Accepted and Published at ICASSP 2020

Journal ref: 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Barcelona, Spain, 2020, pp. 7859-7863

LSTM-based Whisper Detection

Authors: Zeynab Raeesy, Kellen Gillespie, Zhenpei Yang, Chengyuan Ma, Thomas Drugman, Jiacheng Gu, Roland Maas, Ariya Rastrow, Björn Hoffmeister

Abstract: This article presents a whisper speech detector in the far-field domain. The proposed system consists of a long-short term memory (LSTM) neural network trained on log-filterbank energy (LFBE) acoustic features. This model is trained and evaluated on recordings of human interactions with voice-controlled, far-field devices in whisper and normal phonation modes. We compare multiple inference approac… ▽ More This article presents a whisper speech detector in the far-field domain. The proposed system consists of a long-short term memory (LSTM) neural network trained on log-filterbank energy (LFBE) acoustic features. This model is trained and evaluated on recordings of human interactions with voice-controlled, far-field devices in whisper and normal phonation modes. We compare multiple inference approaches for utterance-level classification by examining trajectories of the LSTM posteriors. In addition, we engineer a set of features based on the signal characteristics inherent to whisper speech, and evaluate their effectiveness in further separating whisper from normal speech. A benchmarking of these features using multilayer perceptrons (MLP) and LSTMs suggests that the proposed features, in combination with LFBE features, can help us further improve our classifiers. We prove that, with enough data, the LSTM model is indeed as capable of learning whisper characteristics from LFBE features alone compared to a simpler MLP model that uses both LFBE and features engineered for separating whisper and normal speech. In addition, we prove that the LSTM classifiers accuracy can be further improved with the incorporation of the proposed engineered features. △ Less

Submitted 5 April, 2020; v1 submitted 20 September, 2018; originally announced September 2018.