11115 research outputs found
Sort by
On Intrinsic Dimensionality of Data Sets and Neural Networks
The concept of Intrinsic Dimensionality (ID) is of special interest in the field of Neural Networks (NNs) since it promotes both (a) a deeper understanding of the underlying mechanisms, and (b) embraces parsimonious modeling (that is, building the right-sized model for the task) with associated benefits to processing speed and storage requirements. This thesis explores the concept of ID via two separate, but related, questions. First, we study the potential of NN ID prediction by exploiting easily obtained quantities measured on the data. We then explore NN ID as an independent concept by comparing the results of different methods for NN ID estimation
Measuring the Comparative Performance of USAF Behavioral Health Clinic Operations: Emphasis on Capacity
This research examines the operations of behavioral health clinics in Air Force Continental United States facilities, with a focus on operational efficiency and capacity through Data Envelopment Analysis. Most facilities operate near capacity, demonstrating efficient resource usage, but concerns arise with the potential of increased demand. Days-to-care and leakage are identified as major sources of inefficiency, suggesting that addressing these issues can enhance operations and reduce costs. The study also explores the impact of COVID-19 on care delivery
An Assignment Model for Lateral Transfers Matching Base Repair Facilities to XF3 Coded NSNs Requiring Repair
This research addresses challenges in efficiently distributing reparable National Stock Numbers (NSNs) to base-level repair facilities, aiming to ease strain on depot resources. It establishes a network integrating bases with similar repair capabilities for NSNs and allocates NSNs to balance repair capacities. Drawing from USAF reparable inventory modeling and inventory management theory, it generates data mimicking historical data to maximize total expected part repairs by assigning NSNs based on the best percentage of base repair (PBR)
Evaluating United States Air Force Installation Water Security
Water shortage at Mountain Home Air Force Base in the last decade ignited the need for an assessment of relative water security risk across United States Air Force installations. Current evaluation methods often differ from actual conditions suggesting a need to categorize security across the enterprise for multiple discrete locations. Indicator-based approaches for characterizing water security are common in academic literature but have little methodological consensus and have rarely been applied to discrete areas spread across geographically separated locations with diverse climates. Using a two-stage analytical hierarchy process and normalized utility functions, water security at 30 installations across the continental United States was calculated using 14 indicators spanning climate, demand, and infrastructure categories obtained from the Air Force Civil Engineering Water Dashboard and external sources. Findings of this research suggest bases in the western half of the United States have lower relative water security and should be prioritized for future climate resilience investment. This thesis supports National Defense Authorization Act Section 2827 and climate-informed decision making under the Defense Climate Adaptation Plan, while demonstrating one way to leverage available data for strategic asset management and resilience planning
Retrospective Cohort Study of Pure Tone Audiometry Hearing Changes from Ototoxic Metalsand Solvents, Continuous Noise, and Impulse Noise Exposures at Hill Air Force Base From 2005 to 2019
This retrospective cohort epidemiological study investigated the relative risks of hearing loss associated with ototoxicants in combination with noise exposure. Utilizing Department of Defense (DoD) industrial hygiene and hearing conservation data, this research expanded on a 2020 thesis conducted on Tinker Air Force Base (AFB), applying a similar methodology to Hill AFB, adding 893 evaluated individuals. Grouped into twelve exposure combinations with a minimum of three years of exposure duration, the study assessed various hearing loss indicators, including DoD Standard Threshold Shift (STS) and NIOSH STS. Ototoxic substances consistently elevated relative risk (RR) compared to noise-only exposure groups, but none reached significance at the 95% confidence level. Incorporating Hill AFB to findings from Tinker AFB (n=2,372) revealed exposure groups with a RR greater than one for developing a NIOSH STS were significant at the 95% confidence level, with the greatest RR coming from the metal, solvent, continuous noise exposure group in the left ear at 2,000 Hz (RR=2.25; 1.96-2.57). Logistic regression modeling identified age and audiogram duration as significant independent variables for hearing loss indicator development prediction. Results further suggest potential limitations in the DoD Hearing Conservation Program\u27s ability to capture hearing risk from ototoxic substance exposure
A Framework for Dynamic Combat Support Exercises
Future armed conflict will be characterized by the unknown as adversaries innovate and evolve—surprise is inevitable. Current exercises provide inadequate opportunities for combat support forces to develop improvisational skills. This research proposes an extensible framework for human-in-the-loop control of exercises using a graph network for modeling combined with topological analysis and the resource-constrained project scheduling problem. This framework is assessed using the United States Air Force Silver Flag exercise as a case study with promising results
A Model-Based and System-Theoretic Approach to Design Advanced Autonomy for Air Dominance Missions: A Loyal Wingman Case Study
In contested air environments, safe coordination between decision-makers is paramount. Although the Department of Defense (DoD) prioritizes the development of Artificially Intelligent (AI) wingmen for air combat, a lack of methodology exists to design safe, holistic coordination between human and autonomous wingmen in the same environment. This thesis delivers a framework using Systems Theoretic Process Analysis Extended for Coordination (STPA-Coord) to analyze and design holistic coordination for the Loyal Wingman concept in an Air Dominance mission. STPA-Coord is a safety and hazard analysis process that uses Systems Theory to analyze and design coordination between decisionmakers in a system-of-systems architecture. Using the proposed framework, this thesis delivered 290 loss scenarios and causal factors, ultimately resulting in 83 design considerations to apply to a Loyal Wingman architecture. Additionally, this thesis used a Model-Based Systems Engineering (MBSE) approach to conduct the STPA-Coord using Risk Analysis and Assessment Modeling Language (RAAML). This thesis proposes modifications to SysML’s RAAML guidance to optimize STPA documentation in SysML and to account for STPA’s coordination extension. Results also describe the observed benefits of using model-based methodology over traditional document-based methodologies to conduct STPA-Coord analyses and describe the time required to conduct an STPA-Coord in SysML
An Analysis of Military Construction Cost Escalation and How It Is Accounted for in Cost Estimating Practices
Construction escalation is a significant economic phenomenon that plays an important role in the construction industry. It refers to the steady increase in the costs associated with facility and infrastructure projects over time. Accurately capturing escalation is a key component in the development of construction cost estimates, and failure to adequately account for escalation can have far-reaching consequences. This is especially true for Military Construction (MILCON) projects, which account for billions of dollars of defense spending annually. The Department of Defense (DoD) utilizes a unit cost approach to develop rough order of magnitude (ROM) estimates for MILCON projects. The DoD methodology accounts for escalation by using the average of three common construction cost indices and calls the resulting index the Selling Price Index (SPI). The methodology has two inherent assumptions that are explored in this thesis: (1) escalation occurs at the same rate across all facility types, and (2) construction cost escalation is comparable between the civilian sector and the DoD. The results in this paper suggest that both assumptions are invalid, with the second assumption being the most consequential to MILCON cost-estimating practices. An adjustment factor to the SPI is proposed to mitigate the apparent disparity between the civilian sector and MILCON escalation
Synergistic Effects of Neutron and Gamma Radiations on the Electrical Properties of MOSFETs
In many radiation environments, electronic assets are exposed to a wide variety of species of radiation. This thesis studies the effect of sequential exposure of MOSFETs to neutron and gamma radiation. 27 Goford Semiconductor GT060N04T MOSFETs with gate thickness ∼500 nm were exposed to gamma radiation from a Cesium-137 source and mixed gamma and neutron radiation from a Californium-252 source with total amounts of ionizing radiation dose, neutron fluence, and order of irradiation varied to determine if synergistic effects of neutron and gamma irradiation can be measured. Neutron fluence values as low as 3.45x109 and as high as 3.40x1010 neutrons/cm2 and Cesium-137 gamma dose of up to 200 rad were used. Comparison of the effect of neutron irradiation in fresh MOSFETs versus MOSFETs previously irradiated by gamma rays from the Cesium-137 source shows strong evidence of synergistic effects, with the threshold voltage shift after Californium-252 irradiation being on average 24 ± 28 fV/(neutron/cm2) in the fresh MOSFETs and 71 ± 18 fV/(neutron/cm2) in the pre-irradiated MOSFETs