Localization and Sensing for Hopping Robots

Abstract

This paper describes the state of the art of hopping robots for space exploration and new conceptual advances in algorithms for localization and sensing of hopping robots. We first review the state of the art of hopping robot systems and describe their characteristics with reference to past and current space missions. Then, we briefly summarize the operational principles of minimally actuated jumping robots that are suitable for exploring the unstructured terrain of celestial bodies. These principles are supported by the description of some of the key mechanisms implemented, and by laboratory tests. Then, a new localization algorithm for hopping localization that fuses camera, inertial and acceleration data to compute the most likely landing location of the robot is proposed. To support localization with geo-referenced measures, we also briefly propose a new stereo system, based on omnidirectional cameras, whose physical characteristics are well suited for a hopping robot. Computer simulations of the localization algorithm, together with preliminary geometrical analysis of the proposed stereo system, support our claims of their applicability to planetary exploration. Conclusions and plans for future development end the paper.

Spacecraft/rover hybrids for the exploration of small Solar System bodies

“…A featurebased temporal matching technique, known as visual odometry, has been used for estimating the pose on Mars rovers [12]. While, as discussed before, such techniques cannot be readily applied to the hybrids, the idea of matching visual features to estimate pose can be adapted to this unique platform (indeed, some preliminary work to adapt visual odometry to hopping platforms is already available [13], [14], [15]). In our case, using onboard stereo imaging would likely not fit within the envisioned mass and volume of the hybrids.…”

Gravity-Independent Locomotion: Dynamics and Position-Based Control of Robots on Asteroid Surfaces

“…Determining, updating and maintaining knowledge of rover position and orientation on an asteroid surface can be important for recording spatial context for surface science measurements and for certain mission concepts of operation. Localization approaches for hopping robots have been proposed with some reliance on range measurements to an orbiting or station-keeping mother spacecraft (Yoshimitsu et al, 2001) and via use of more general approaches such as particle filters (Martinez, 2004), Kalman Filters with landmark geo-referencing (Fiorini et al, 2005), and optical flow as well as visual odometry without continuous terrain feature tracking while tumbling (So et al, 2008;. During local navigation across the terrain, existing localization approaches for rolling or walking robots may apply.…”

Visual Odometry for a Hopping Rover on an Asteroid Surface using Multiple Monocular Cameras

“…On aerial vehicles and spacecraft, inertial sensors such as accelerometers and gyroscopes are used. Fiorini perform localization on hopping rovers [9]. However, these methods are susceptible to unbounded accumulation error.…”