Hackathon on Permafrost

28th – 30th November 2019 @ ETH Zurich

The mountains are changing: permafrost is receding, rockslides can be the result. ETH Zurich scientists Jan Beutel and Matthias Meyer want to understand this in more detail. They have been researching and monitoring permafrost around the clock for many years. They have collected a series of images and data that is unique in the world, and would like to analyze them with you at a hackathon – supported by technology partner Microsoft. Are you in?


Anyone who enjoys programming and for whom data science, cognitive services and image recognition are no unknowns can take part, for example, students from ETH, universities or technical colleges as well as IT apprentices. Partners from the ecosystems of ETH and Microsoft will also participate.

The project by Jan Beutel and Matthias Meyer is called “Alpine Laboratory”. It investigates the changes in permafrost in order to be able to predict rock and mountain falls more precisely, enabling those affected being warned in due time and preventive measures being taken early enough.

Understanding instabilities in rock better thanks to artificial intelligence

The basis of the project is a high-quality amount of image data as well as a data set that is including seismic signals, which have been registered at 29 locations on the Matterhorn over the last 10 years, unique in the world. With the help of artificial intelligence, recognizable phenomena such as thunderstorms or mountain hikers are to be correlated with the blips in the seismic data. The aim is to eliminate existing disturbing factors in order to identify true instabilities in the rock material more clearly and reliably. In addition, a mobile application is to be developed through which mountain hikers can feed observed rock shifts or falls directly into a database. This crowdsourcing of additional data points is intended to further improve the quality of the model in the future.

Procedure at the hackathon:

1.     Interactive introduction to the project and the data set. Analysis and preparation of the data in order to use learning-based methods.

2.     Individual development ranging from learning-based procedures to the automated analysis of data

3.     Interpretation and presentation of results

The thawing permafrost and the accumulation of rockslides will lead to changes in the mountain world in high alpine regions. The “Alpine Laboratory” wants to show how artificial intelligence can be used to find solutions that are important for people and the living and economic environment.