ESA GNC Conference Papers Repository

Autonomous Lunar Landing Based on Bio-inspired Visual Motion sensors tested in-flight
Sabiron, G.; Burlion, L.; Kervendal, E.; Bornschlegl, E.; Raharijaona, T.; Ruffier, F.
Presented at:
Porto 2014
Full paper:

The challenging task of autonomous lunar landing has regain interest in the last few decades. New means of ensuring safe descent with strong final conditions and aerospace-related constraints in terms of mass, cost and computational resources are required to ensure soft landing on celestial bodies. Visual based systems seem to be well suited to achieve Entry Descent and Landing or autonomous aerial vehicle navigation tasks and are therefore widely considered in the literature. Bio-inspired visual sensors could provide the miniaturization aspects required in such applications. In this paper, a two-phase approach is presented: first a biomimetic strategy inspired from flying insects is presented as a solution to perform safe lunar landing. In order to design a Guidance Navigation and Control solution relying solely on optic flow and inertial measurements, a multiple-sensors setup is introduced. The relevancy of these visual motion sensors applied in EDL applications has already been showed with pertinent results in the literature. These multiple sensors might be used either to perform navigation tasks for instance estimating the flight path angle or to control directly the dynamical system as previously seen with flying insects. Secondly, low-speed Visual Motion Sensors (VMS) inspired by insects’ visual systems performing angular speed measurements in the [1.5 degrees/s; 25 degrees/s] range and weighing only 2.8 g are presented. The entire processing stages have been updated with respect to the original VMS’s design to cope with the optic flow range experienced during lunar landing (considering an Apollo-like scenario). Tests onboard an 80 kg unmanned helicopter under free- flying outdoor conditions were realized and followed by an offline comparison to ground-truth optic flow computed thanks to various sensors mounted onboard the UAV such as GPS, IMU and LIDAR. These results show that the optic flow measured despite the natural disturbances encountered closely matched the ground-truth optic flow.