Vision-Aided Navigation for a Free-Flying Unmanned Robotic System to Support Interplanetary Bodies Prospecting and Characterization Missions

This paper investigates vision-aided navigation strategies for an autonomous free-flying robotic vehicle designed to explore interplanetary bodies, such as moons or asteroids, for the purposes of In Situ Resource Utilization (ISRU). ISRU has the potential to facilitate planetary exploration by drawing needed resources, such as water, from the local environment. The realization of ISRU requires the development of advanced unmanned space systems integrated with sample-capture devices and guidance, navigation, and control systems capable of supporting autonomous exploration of challenging environments such as craters and lava tubes.   Navigation on interplanetary bodies is challenging due to the unavailability of traditional navigation sesnors such as GPS and magnetometers. This paper focuses on the use of one or more vision sensors to augment an onboard inertial measurement unit, which is composed of accelerometers and rate gyros, in order to provide vision-aided navigation solutions for the free-flyer robotic system. In this study, a vision-aided navigation strategy is considered that entails using object detection and tracking algorithms to identify known landmarks in the scene, which provides information that can be used to estimate the position of the vehicle within the environment. Vision-aided navigation filters are developed for a stereo implementation of landmark detection and tracking, and the algorithms are implemented on video obtained from flights of a quadrotor UAV at the Hazard Field at the NASA Kennedy Space Center.