Session
Technical Poster Session III
Location
Utah State University, Logan, UT
Abstract
UltraNavis a low size, weight, power, and cost (SWaP-C) compute-enabled camera system capable of implementing industry standard computer-vision algorithms such as object recognition and tracking, state-estimation, and relative navigation. This system simplifies advanced missions that require capabilities such as Terrain Relative Navigation (TRN), visual Simultaneous Localization and Mapping (SLAM), 3D reconstruction, or fiducial targeting. Algorithms such as these have been limited to large spacecraft on high budget missions due to processing requirements and the resulting increase in SWaP-C. The sensor’s flexibility allows it to be used for low level image capture, image processing, or as a host for a full GN&C solution.
Current large-scale missions are demonstrating the importance of high-performance vision-based algorithms for accomplishing cutting edge, high priority science objectives. OSIRIS-REx and Hayabusa2 have performed precise operations at previously unexplored asteroids, and Mars Perseverance is expected to be the most precise automated landing on a non-terrestrial planetary body to date. Vision applications such as TRN and 3D modeling will open up higher-level autonomy capabilities for small satellites missions that have not had these available to date. Prior visual navigation tools available to small spacecraft have tended to be application specific (e.g., star trackers) or have required extensive engineering effort (e.g., mission-specific ad hoc integration and algorithm development). This limited the operation of small spacecraft in GPS-uncertain environments. UltraNav promises high capability while maintaining low cost, and ease-of-integration. Additionally, UltraNav’s reconfigurability enables it to be used for multiple applications in the same mission, such as switching from localization to star tracking.
Ultra-Low SWaP Relative Navigation
Utah State University, Logan, UT
UltraNavis a low size, weight, power, and cost (SWaP-C) compute-enabled camera system capable of implementing industry standard computer-vision algorithms such as object recognition and tracking, state-estimation, and relative navigation. This system simplifies advanced missions that require capabilities such as Terrain Relative Navigation (TRN), visual Simultaneous Localization and Mapping (SLAM), 3D reconstruction, or fiducial targeting. Algorithms such as these have been limited to large spacecraft on high budget missions due to processing requirements and the resulting increase in SWaP-C. The sensor’s flexibility allows it to be used for low level image capture, image processing, or as a host for a full GN&C solution.
Current large-scale missions are demonstrating the importance of high-performance vision-based algorithms for accomplishing cutting edge, high priority science objectives. OSIRIS-REx and Hayabusa2 have performed precise operations at previously unexplored asteroids, and Mars Perseverance is expected to be the most precise automated landing on a non-terrestrial planetary body to date. Vision applications such as TRN and 3D modeling will open up higher-level autonomy capabilities for small satellites missions that have not had these available to date. Prior visual navigation tools available to small spacecraft have tended to be application specific (e.g., star trackers) or have required extensive engineering effort (e.g., mission-specific ad hoc integration and algorithm development). This limited the operation of small spacecraft in GPS-uncertain environments. UltraNav promises high capability while maintaining low cost, and ease-of-integration. Additionally, UltraNav’s reconfigurability enables it to be used for multiple applications in the same mission, such as switching from localization to star tracking.
