VaMEx - Vipe: Exploration in schwer zugaenglichem Terrain anhand visueller und propriozeptiver Daten im Valles Marineris
Daniel Kuehn
series DFKI Documents, volume 15-03, pages 6, Sep/2015. DFKI GmbH.

Zusammenfassung (Abstract) :

The presentation introduced the Vipe project, which is a part of the ”VaMEx - Valles Marineris Explorer” Initiative, started by the DLR Space Administration. The Initiatives aim to explore craters on Mars up to 7 km deep fully autonomous by a heterogeneous swarm of robots, including the hominid robot Charlie. The Valles Marineris, a jagged rift valley, places high demands on robotic exploration mission. This environment appears due to the earlier volcanic activity as well as the references to water resources extremely promising for a variety of scientific issues. To have a comprehensive picture of Valles Marineris and thus potential niches for extraterrestrial life, areas which are difficult to access have to be included in the exploration in particular. Within the first VaMEx project, a swarm of heterogeneous robots (rovers and aerial robots) already allowed a significant application expansion of the exploration mission. Still, caves, steep slopes, and rugged rock formations continue to be a major challenge for the use of mobile robots. The aim of VIPE is to fill this gap within the newest swarm member Charlie, to increase the overall swarm locomotion and navigation abilities. Due to Charlies lightweight and highly integrated design, its agility, and integrated tactile sensors ideally suited to deal with difficult terrain. Furthermore, a novel visual positioning and mapping approach will be developed, featuring a 360 panoramic camera which allows a positioning with very low drift despite the above-mentioned, demanding conditions. This visual positioning is to be supplemented by a complementary proprioceptive approach based on tactile sensors to improve self-localization. This is a prerequisite for movement planning and reactive motion control to allow the robot to overcome obstacles autonomously.


© DFKI GmbH
zuletzt geändert am 27.02.2023