11115 research outputs found
Sort by
Opportunities and Limitations: Integrating Narrative AI into Game-Based Assessment Creation and Evaluating Student Impacts
Get PDFThe Department of Defense (DoD) has identified the need for a technically proficient workforce in the areas of science, technology, engineering, and mathematics. To meet this need, the DoD is actively seeking innovative technology capable of creating workforce development opportunities that are both accessible and effective. Educational research indicates serious games provide a potential avenue to achieve this goal. Unfortunately, a limited number of tools that simplify the game development process and leverage artificial intelligence are available. Content-generating artificial intelligence might help reduce instructor and game-based assessment designers\u27 workloads while promoting individualized learning in students. This research presents a novel tool that utilizes ChatGPT\u27s content generation ability to create narrative game-based assessments and evaluate student impacts. Four instructors used the tool and generated game-based assessments for class use. Data were collected via an instructor survey, student pre/post-assessment surveys, a traditional assessment, and a game-based assessment. Although findings show that ChatGPT narratives result in less engaging and fun experiences, evidence suggests the generated game-based assessments do not produce significantly different scores than traditional assessments. The lack of a significant score difference suggests that game-based assessments are no worse than traditional tests in terms of assessment effectiveness. The generated game-based assessments may also reduce student test anxiety when taking an assessment of appropriate difficulty. The GBA Creation Tool, data collected during this research, and ChatGPT Integration Framework appear beneficial to the game-based assessment research community and the DoD
Modeling Lightning Flashes in Dissimilar Non-Homogeneous Clouds
Get PDFA new methodology for near-infrared lightning radiative transfer through clouds is presented. High-resolution weather modeling is used to generate realistic three-dimensional non-homogeneous clouds, which are then used as environments for a Monte Carlo simulation of photon multiple scattering from approximations of lightning flashes. Resultant emissions from the cloud tops are in line with previous studies and satellite observations
Artificial Intelligence and Perception: An Empirical Study
Get PDFThis thesis investigates the impact of adjusting artificial intelligence explainability levels’ outputs on user perception. The overarching study extends within the Explainable Artificial Intelligence (XAI) domain. It examines a spectrum of variables, including performance, cognizance, familiarity, transparency, system bias, and the overall impact of AI, to understand their collective and individual effects that enable effective professional use in an organization. The study aims to illuminate the relationship between the degree of explainability provided by large language models such as ChatGPT, Bard, and Bing AI and the performance of these models when tasked with XAI adjustments
Digital Airworthiness: Development of a Sysml Based Framework for the USAF Airworthiness Process
Get PDFIn 2018 USAF decided to move towards a digital transition for programs and processes. To maximize the impact of initial transition efforts the airworthiness process was identified as a candidate for digital implementation. The airworthiness process is mandated for nearly every system or modification that flies. The airworthiness process identifies specific tasks that must be performed and artifacts that must be produced. This research worked to develop a framework that utilizes SysML to pattern these processes and the required artifacts. A system agnostic, importable library was developed using custom stereotypes and relationships for the purpose of application to digital system models. The airworthiness requirements from MIL-HDBK-516 were digitized; risks, artifacts and relationships were patterned; self-scoping tables were developed to enable airworthiness evaluation; and a base level of error checking was built in. To enable further development of the library, example logic for expanded functionality was patterned but has not yet been implemented. The library was applied to a nominal system with existing airworthiness to document the process of using the library and determine the level of effort required to do so. Finally, to continue development and integration, the library was passed on to the Digital Transformation Office
Attitude Control Analysis of Space-Based Mirror Satellites for Illuminating Dimly-Lit Resident Space Objects
No full textThis research focuses on the application of an attitude controller and the assessment of its precision in directing light onto a Resident Space Object (RSO) within a proximity of 22 meters during a natural motion circumnavigation (NMC) orbit. Specifically, the study utilizes the well-known Hill-Clohessy-Wiltshire (HCW) equations to position a servicer satellite around an RSO located in geosynchronous Earth orbit (GEO). An underlying assumption is that the spacecraft is equipped with a mirror initially facing the RSO. In summary, this research aims to develop an attitude controller that enables a servicer satellite to illuminate all six sides of a non-compliant RSO, thereby enhancing the effectiveness of satellite mission operations
Machine Visual Perception for Autonomous Docking Maneuvers
Get PDFThis dissertation presents a novel approach to autonomous docking using machine learning for visual perception, particularly during probe and drogue aerial refueling. Autonomous vehicles have become pervasive in both civilian and defense sectors, and their ability to interact with their surroundings and each other autonomously is critical for future operations. Traditional methods relying on signals or inertial sensors face significant limitations such as interference, jamming, and drift. This research focuses on developing a computer vision-based solution to overcome these limitations. A novel pipeline, termed relative vectoring, is introduced, which utilizes dual object detection and machine learning to estimate relative positions between the receiver and the drogue during aerial refueling using imagery alone. The proposed solution leverages object detection models to detect and match 2D image points to 3D object points, enabling accurate pose estimation and vector computation without relying on extrinsic camera calibrations. The pipeline was validated through extensive simulation using the AftrBurner graphics engine, which provided realistic imagery and dynamic scenarios. Simulation results demonstrated the pipeline’s accuracy, reliability, real-time performance, and resilience to occlusions. Additionally, relative vectoring was tested in real-world scenarios using transfer learning techniques. The findings showed that scene augmentation significantly enhances model generalization and accuracy, bridging the sim-to-real gap. Real-world results confirmed that this method is reliable, accurate (within 3 cm at contact), and fast (56 fps) even under varying environmental conditions. Overall, results indicate that the proposed solution can effectively perform autonomous docking in complex and dynamic environments, thus advancing the capabilities of autonomous aerial refueling and related applications
Engaging the Acquisition Work Force: A Study of Antecedents in a Program Office Context
Get PDFIn an era marked by international conflict and heightened geo-political tensions, the Air Force has necessarily begun preparing for a near-peer fight. Air Force senior leaders have defined improving acquisitions career field performance as a necessary objective to keep pace with adversaries. The linkage between employee engagement and enhanced performance outcomes, as widely supported throughout the literature, suggests that fostering a more engaged acquisition work force may be the key to delivering technology to the field quickly and affordably. As such, this research effort surveyed an acquisitions program office to characterize the key drivers of engagement. Stepwise regression showed that openness (R2 = 0.29, p \u3c 0.001), the ability to develop and be oneself (R2 = 0.24, p \u3c 0.05), and service to others (R2 = 0.62, p \u3c 0.001) were the most influential factors. Practical recommendations based on these insights were made for implementation by acquisition leaders to enhance the engagement of their teams in pursuit of improved performance
Optimal Placement of Artificial Hair-Cell Airflow Sensors for Bioinspired Flight-by-Feel
Get PDFThe Sparse Sensor Placement Optimization for Prediction (SSPOP) algorithm reduces highdimensional airflow data to a low-dimensional sparse approximation to identify an optimal placement for any number of sensors on airfoil or wing models of arbitrary shape and size, and outperforms conventional optimization techniques in accuracy and speed. For 2D flow this algorithm found a sensor placement solution (design point, or DP) which predicts AoA to within 0.10 degrees and ranks within the top 1 percent of the design space. On 3D wing models SSPOP found four-sensor DPs ranked well within the top 0.10 percent. Experimental validation with velocity and pressure sensors confirmed the relative and absolute performance of five DPs: Best Possible (true optimum), SSPOP, Expert Opinion, and two Random DPs
Editorial: Observations and Simulations of Layering Phenomena in the Middle/upper Atmosphere and Ionosphere
Get PDFThe middle/upper atmosphere and ionosphere are the transition between neutral and ionized components of the Earth’s atmosphere, including stratosphere, mesosphere, thermosphere, ionospheric E region and ionospheric F region (Laštovička et al., 2006; Xu, et al., 2007; Smith, 2012). The atmospheric thermal structure and composition are significantly affected by dynamical processes through coupling. The layering phenomena such as mesospheric metal layers, sporadic E layers, and noctilucent clouds are important tracers to study mechanisms of the vertical coupling from the lower to the upper atmosphere (Dou et al., 2010; Plane, 2012; Xue et al., 2013)
Effects of RF Signal Eventization Encoding on Device Classification Performance
Get PDFThe results of first-step research activity are presented for realizing an envisioned “event radio” capability that mimics neuromorphic event-based camera processing. The energy efficiency of neuromorphic processing is orders of magnitude higher than traditional von Neumann-based processing and is realized through synergistic design of brain-inspired software and hardware computing elements. Relative to event-based cameras, the development of event-based hardware devices supporting Radio Frequency (RF) applications is severely lagging and considerable interest remains in obtaining neuromorphic efficiency through event-based RF signal processing. In the Operational Technology (OT) protection arena, this includes efficient software computing capability to provide reliable device classification. A Random Forest (RndF) classifier is considered here as a reliable precursor to obtaining Spiking Neural Network (SNN) benefits. Both 1D and 2D eventized RF fingerprints are generated for bursts from NDev = 8 WirelessHART devices. Average correct classification (%C) results show that 2D fingerprinting is best overall using detected events in burst Gabor transform responses. This includes %C ≥ 90% under multiple access interference conditions using an average of NEPB ≥ 400 detected events per burst. This is sufficiently promising to motivate next-step activity aimed at (1) reducing fingerprint dimensionality and minimizing the required computational resources, and (2) transitioning to a neuromorphic-friendly SNN classifier—two significant steps toward developing the necessary computing elements to achieve the full benefits of neuromorphic processing in the envisioned RF event radio