We have built a robot to venture into a glacier’s ice sheet to provide glaciologists with critical data.
Debugging our robot at -5 degrees Celcius/ 23 degrees Fahrenheit
The Mer de Glace glacier is with 7.5km the second largest glacier in the alps and with up to 200m depth the Mont Blanc’s deepest glacier. Glaciologists are interested in better understanding glacial dynamics under the influence of climate change. In order to do so, they are interested in extracting measurements and samples at the interface between the glacier’s ice sheet and its rockbed, where subglacial caves, streams and tunnels can be found.
Mer de Glace (sea of ice) glacier
This is a difficult and dangerous environment to access. One access point is through so called moulins. Moulins are formed when meltwater from the glacier’s surface is draining through crevasses and leaves behind near vertical chimney-like structures on its way to the glacier’s bed.
Approaching moulins on the Mer de Glace
View from the top into the moulin (our robot is rappelling on the yellow tether)
We have built a robot to venture into the ice sheet through moulins to provide glaciologists with critical data. Our robot is equipped with two propellers and rappells into the ice. Further, it caries a stereo camera, a light, a Lidar sensor, multiple IMUs, and motor and flight controllers. Finally, a Jetson Xavier NX runs all our control and mapping algorithms and enables efficient data capturing and processing and communication.
Robot rappelling during its exploration of the moulin
Here is a link to the preprint of our paper where we describe the motivation, methods and results of our robot in more detail: Into the Ice: Exploration and Data-Capturing in Glacial Moulins by a Tethered Robot - Authorea
We have started this project as a collaboration with the Swiss Polar Institute and our goal would be to eventually deploy our robot to even bigger glaciers like in the Himalayas or on the Greenlandic ice sheet, Project Detail – Swiss Polar Institute
And finally, this project could not be realized without many contributions from the open source community and the availability of outstanding compute hardware. I would like to point towards this blog post which I have written together with Foxglove that contains additional details and photos: Spotlight: Using Foxglove Studio to Map Glaciers in the French Alps - Foxglove