Extended Navigation Capabilities for a Future Mars Science Helicopter Concept
Jeff Delaune, Roland Brockers, David S. Bayard, Harel Dor, Robert A. Hewitt, Jacek Sawoniewicz, Gerik Kubiak, Theodore Tzanetos, Larry Matthies, J. Balaram
Abstract
This paper introduces an autonomous navigation system suitable for supporting a future Mars Science Helicopter concept. This mission concept requires low-drift localization to reach science targets far apart from each other on the surface of Mars. Our modular state estimator achieves this through range, solar and Visual-Inertial Odometry (VIO). We propose a novel range update model to constrain visual-inertial scale drift in the absence of motion excitation using a single-point static laser range finder, that is designed to work over unknown terrain topography. We also develop a sun sensor measurement model to constrain VIO yaw drift. Solar VIO performance is evaluated in a simulation environment in a Monte Carlo analysis. Range-VIO is demonstrated in flight in real time on 1 core of a Qualcomm Snapdragon 820 processor, which is the successor of the NASA's Mars Helicopter flight processor.