Maritime GPS-Denied UAV Autonomy & Alt-PNT, No-Emit, ASPN & Mod Payload Compliant Solution: Rhoman Aerospace, Naval Postgraduate School, Firestorm Labs, InSitu

The proposed effort between Rhoman Aerospace and the Naval Postgraduate School enhances Air Force, Navy, and SOCOM funded GPS-Denied UAV navigation, targeting, and autonomy technology for maritime use and deploys the solution through a software download or modular payload that is compliant with Mod Payload AD1167779 and ASPN/pntOS standards. The proposed effort demonstrates the solution on a Firestorm Labs UAV and an InSitu UAV in a maritime environment, characterizes the performance and portability to other UAV, and delivers a hardware prototype. Adaptation of the current baseline Air Force, Navy, and SOCOM funded GPS-Denied navigation system to maritime operations is feasible because machine-vision algorithms can be updated & optimized for maritime visual features, prior AFRL R&D shows success with UAV heading estimation using wave-direction tracking, visual features used in navigation remain still relative to the UAV speed over a period of seconds, and the machine-learning drift-correction systems to be deployed are ideal for the structurally based, albeit non-linear format of wave dynamics.Adapt algorithms to optimally detect & track horizontal blobs, to enable wave/swell tracking over timeUpdate object-detection algorithms to optimally detect the sun and horizon, duplicate algorithm and enact two tuning optimizations, for high-contrast highly-spherical bright object and for linear, long-distance horizon objectAdd a gain-optimization layer, to increase contrast of similarly colored environments of open water, a gain-optimizer increases contrast to help horizontal blob-cluster detector and wave trackerCreate wave-direction model, current algorithms detect and track objects over time, this algorithm will be updated to instead track multiple horizontal cluster/blobs over time, to ascertain wave direction for use in heading estimateDeploy variable frame-rate filter layer for wave-detection and other-algorithm optimization, there is a critical time scale threshold above which the waves dissipate and become difficult to track Tune machine-learning drift correction model for various wave height, size, wavelength, and lighting conditionsUse vision aided nav to augment IMU estimates and switch out higher cost IMUs from systemProposed Demonstrations & Performance Characterization: Flight demonstration and performance characterization tests are planned with partner UAV firms Firestorm Labs and InSitu, a Boeing Company. Performance characterization includes testing the solution at varying altitudes, flight speeds, and with various maneuver profiles in maritime, coastal, and open-water environments.. The proposed platforms include the performance needed for the proposed effort, and are already  ModPayload AD1167779 compliant, and so are able to easily deploy the  proposed payload. Partner UAV are able to cover 100km-300km in range, which is adequate for planned 50km-150km planned flight tests.