USAFA Project C

Title: AI-informed Relative Navigation for GPS-Denied Environments
Advisor: Ioan Feier, Michael Anderson​
Synopsis of Presentation: Developed a relative localization framework for GPS-denied autonomous agents, utilizing the semantic reasoning capabilities of a Vision-Language Model (VLM) for high-level mission guidance. By anchoring these agents to a single teammate equipped with Magnetic-based Navigation (MagNav) for global positioning, the system can interpret flexible mission prompts to identify and globally locate critical points of interest. Validated through both simulation and hardware experiments, this approach demonstrates how context-aware AI can adapt decision-making and ensure mission success even in environments without GPS.