Vision and Sensing for Lunar Rovers

Summary of work on a Lunar Rover feasibility study, exploring appropriate contemporary perception technologies.

Part of my work over the Dec 2020 to Feb 2021 summer break was working on sensing and perception capabilities of lunar rovers. In particular, for rovers undertaking exploration at the lunar poles.

The lunar poles have harsh, unattenuated illumination from the sun that illuminates the surface at shallow angles. This casts shadows which can be up to several kilometers in length, and where at the furthest extents can move extemely fast (on the order of centimeters per second). Lunar regolith also experiences the coherent backscatter effect, leading to retroreflection which renders images flat and featureless. These are all phenomenon that are not encountered in traditional terrestrial field robotics applications.

I worked under a broader project to characterise state of the art sensing methods for use in rover navigation and local path planning in this environment. You can see a 1-minute summary of these results below.

My work is hopefully going to inform sensing and perception decisions for an upcoming mission which is incredibly exciting. I’m very interested to see where this project will go, and in particular to see if there would be any upcoming opportunities to see this work through to an actual rover system.