Naval Postgraduate School
Calhoun, Institutional Archive of the Naval Postgraduate SchoolNot a member yet
71232 research outputs found
Sort by
GENERATION OF DOMAIN SPECIFIC DOCUMENTS USING LARGE LANGUAGE MODEL TECHNIQUES
No full textIn recent years, the development and proliferation of large language models (LLMs) has significantly impacted multiple sectors of society. It is critical that the U.S. military immediately adopt and experiment with this burgeoning technology to meet increasing operational requirements through more efficient workflows to maintain a technical advantage over near-peer adversaries and to find limitations and vulnerabilities. This dissertation develops an end-to-end framework to generate domain specific documents from LLM selection to document evaluation. Our developed domain specific document generation framework incorporates multiple techniques including prompt engineering (PE), retrieval augmented generation (RAG), an agentic approach, and an LLM-as-evaluator approach. For our test use-case to validate the framework, we selected a road-to-war document, which is commonly used as a starting point for scenario development in the Department of Defense (DoD) wargaming, operational, training, and analysis communities. Using our framework, we were able to generate plausible domain specific documents with LLMs that were validated by experts as having utility, and which were not meaningfully distinguishable from a human-generated exemplar document. Our research demonstrates that LLMs can augment domain specific workflows for text-generation tasks and that significant time savings can be achieved through leveraging this rapidly evolving technology.Distribution Statement A. Approved for public release: Distribution is unlimited.Lieutenant Colonel, United States Arm
Faces of NPS: Lt. Col. Scotty Black, USMC
Get PDFFaces of NPS features interviews spotlighting the students, faculty, staff and alumni of our Nation's premier defense education and research institution
Faces of NPS: Christian Fitzpatrick
No full textFaces of NPS features interviews spotlighting the students, faculty, staff and alumni of our Nation's premier defense education and research institution
Faces of NPS: Lt. Cmdr. Katherine Chlebo, USN
No full textFaces of NPS features Interviews spotlighting the students, faculty, staff and alumni of our Nation’s premier defense education and research institution
ENHANCING ORGANIZATIONAL EFFICIENCY THROUGH AN INTEGRATED DIGITAL DATA AND EVENT MANAGEMENT SYSTEM
No full textThis thesis examines inefficiencies in digital data and event management within U.S. Marine Corps (USMC) units and proposes a unified system that integrates Microsoft tools to enhance operational coordination, reduce redundancy, and improve readiness. The central premise is that an integrated solution utilizing existing Microsoft resources can replace fragmented manual processes without incurring additional costs or security risks. A task analysis identified recurring issues, such as duplicate data entry, outdated information, and delayed communication. A working prototype was developed using Microsoft PowerApps, Power Automate, and Lists to automate training updates, event coordination, and personnel notifications; the system supported the real-world task flow and dynamic updates typical of USMC operations and training. Performance testing with dynamic data and scenario-based use cases demonstrated responsiveness, accuracy, and workload reduction improvements. The resulting system provided a centralized, role-specific interface that enhanced information sharing, streamlined execution, and reduced administrative burdens. This research effort presents a scalable, cost-neutral solution ready for unit-level adoption. It supports a more connected and efficient force by delivering timely information to stakeholders using tools already available to every Marine.Distribution Statement A. Approved for public release: Distribution is unlimited.Captain, United States Marine Corp
Faces of NPS: Lt. Joshua Mayberry
No full textFaces of NPS features interviews spotlighting the students, faculty, staff and alumni of our Nation's premier defense education and research institution
THREE-DIMENSIONAL MULTISTATIC SONAR OPTIMIZATION WITH ENHANCED PROPAGATION MODELING
No full textThis thesis develops a three-dimensional optimization model for multistatic sonar sensor networks to enhance underwater surveillance and detection capabilities. While current research simulates multistatic sonar placement in two dimensions, these approaches fail to account for the critical depth dimension and complex underwater acoustic propagation. This study extends an existing optimization model to incorporate variable deployment depths, realistic bathymetric data, and target strength considerations. The Acoustic Research Laboratory Python Tool is integrated to model underwater sound propagation under environmental conditions. The mathematical formulation addresses both cost reduction and coverage maximization objectives. The model is implemented by using mixed-integer programming with linearization techniques. This research provides naval forces worldwide with an enhanced methodology for optimizing multistatic sonar system deployments to monitor strategic waterways, protect naval assets, and guard maritime borders to increase national security.Distribution Statement A. Approved for public release: Distribution is unlimited.Outstanding ThesisMajor, German Arm
EVALUATING THE INCORPORATION OF ARTIFICIAL INTELLIGENCE INTO NAVAL FLIGHT EDUCATION
No full textArtificial intelligence (AI) has taken the world by storm. As the technology advances, the United States Military and its branches are eager to find ways in which the incorporation of this evolving technology can increase mission effectiveness. In the U.S. Navy, the aviation community is curious about the ways in which AI would impact naval flight education. This study aims to identify the positive and negative implications of incorporating AI into naval flight education, with a focus on the first phase of flight training, Naval Introductory Flight Evaluation (NIFE). This study uses library research as the primary method of research. The result analysis of this research identifies limitations of the current NIFE curriculum, explains the potential benefits of AI-based tools, discusses challenges that need to be addressed before adopting this technology, and makes suggestions for areas of future research on this topic. Using published research on the effectiveness of AI tools in education and aviation, this study concludes that tools such as Intelligent Tutor Systems (ITS) and AI-based virtual instructors have significant potential to make knowledge acquisition and skill development more efficient and available for Student Naval Aviators (SNAs). This study also concludes that there is significant need for future research in development and validation before these tools can be most effectively implemented in the Naval Aviation Training Curriculum.Distribution Statement A. Approved for public release: Distribution is unlimited.Ensign, United States NavyEnsign, United States NavyEnsign, United States Nav
OPTIMIZING PREDICTIVE MAINTENANCE AND LOGISTICS WITH DIGITAL TWINS IN CONTESTED ENVIRONMENTS
No full textThis study explored the integration of Digital Twin (DT) technology into U.S. Marine Corps (USMC) maintenance and logistics systems, with a focus on artillery platoons operating in contested environments. Through a qualitative approach, the research identifies critical gaps in current maintenance practices, including insufficient training, fragmented data systems, and a lack of real-time visibility. Despite limited awareness of DT concepts among participants, there is strong demand for predictive maintenance capabilities, core functions that DTs enable. The analysis shows that DT technology can enhance operational readiness by enabling predictive maintenance, minimizing equipment downtime, and improving resource allocation. However, successful implementation will require addressing challenges related to data integration, system interoperability, user training, and organizational culture. The study recommends targeted policy updates, pilot programs, infrastructure investment, and PME curriculum reform to prepare the force for DT adoption. Ultimately, DTs offer a scalable solution for transforming the Marine Corps’ sustainment model into a more agile, data-driven, and mission aligned system.Distribution Statement A. Approved for public release: Distribution is unlimited.Captain, United States Marine CorpsCaptain, United States Marine Corp
EFFECT OF TWS NOISE INTERFERENCE ON SAR IMAGE FORMATION WITH PHASE CODED WAVEFORM
No full textSynthetic Aperture Radar (SAR) is an important imaging tool for applications ranging fromgeographical surveys to intelligence gathering, offering the ability to propagate through the atmospheremore effectively than optical sensors. Phase-coded waveforms are favored for their excellent autocorrelation and/or cross-correlation properties. This research investigates the impact of transmit waveform shaped noise jamming (TWS-NJ) on SAR images employing m-sequence and Gold sequence coding. Antenna directivityon phase-coded SAR imaging is also investigated via simulations. The findings reveal that TWS-NJ produces more noisy pixels and distortion to SAR images compared to traditional spot noise jamming (SNJ).Distribution Statement A. Approved for public release: Distribution is unlimited.Lieutenant, United States Nav