Optimizing the Shape of Planetary Rover Wheels using the Discrete Element Method and Bayesian Optimization

== Abstract ==

The MMX Rover is a contribution by CNES and DLR to JAXA's Martian Moons eXploration (MMX) mission and will explore the surface of the Mars moon Phobos. As there haven’t been any successful landers on Phobos, little is known about the conditions on the moon’s surface. Additionally, the gravity on Phobos is only about 1/2000 of Earth’s gravity and the behavior of regolith in such low gravity is still a topic of active research. In order to design a suitable wheel for the MMX Rover, we made worst-case assumptions about the soil conditions and implemented them in a simulation model using the Discrete Element Method. This simulation is then used within an optimization loop that automatically tests different wheel shapes on their suitability for reliable locomotion on Phobos. The result is an optimized wheel shape as well as a dataset showing how different wheel shapes affect wheel performance. These results are now being used to design the wheels for the MMX Rover.
                                                                                                
== Video ==
{{#evt:service=cloudfront|id=276723|alignment=center|filename=1036.mp4}}
                                                
== Abstract ==
<pdf>Media:Stubbig_Lichtenheldt_2022a_5293_1036_abstract.pdf</pdf>

== Full paper ==
<pdf>Media:Draft_Content_9399065881036_paper.pdf</pdf>