State lattice generation and nonholonomic path planning for a planetary exploration rover

International audienceESA's Exomars mission will deploy on Mars a 300kg class rover , which is able to perform subsurface drilling, automatic sample collection and distribution to the onboard instruments for in-situ scientific measurements. Due to limited communication to Earth, the use of autonomous navigation in driving the rover towards a scientific target will increase the mission return by minimizing human intervention. The autonomous navigation on-board software should be designed with respect to the mission-specific constraints like energy consumption, memory and computation power, and time costs. This paper proposes an effective method of nonholonomic path planning using a pre-computed state lattice to generate trajectories which meets vehicle locomotion constraints. The objective is to allow the rover to autonomously reach the selected mission target while minimizing in-place-turn manoeuvres which have high impact on the total travelled distance and execution time during a mission. The proposed approach was implemented in the CNES EDRES environment and tested on real data acquired during navigation campaigns on the SEROM Mars-like test site and on simulated data using a digital elevation model constructed from HiRISE stereo images acquired onboard the MRO