Artificial Intelligence Space Robot Successfully Tested

Der hybride Schreit- und Fahrrover Sherpa TT durchquerte dank der neuen Software autonom die Wüste und legte dabei eine Strecke von über 1,3 km zurück.
Thanks to the new software, the hybrid walking and driving rover Sherpa TT crossed the desert autonomously and covered a distance of over 1.3 km.

The exploration of Mars and the moon is a great dream for mankind. The EU also wants to contribute to space exploration. An exciting project has now been successfully implemented in the Moroccan desert.

At the end of 2018, some 30 scientists from eleven countries visited the Moroccan desert. Researchers from the Robotics Innovation Center of the German Research Center for Artificial Intelligence (DFKI) also took part. The researchers collected findings for the EU "Peraspera" project of the European Space Agency ESA, the German Aerospace Center (DLR), and the national space agencies of France, Spain, Italy, and Great Britain. In this project, a research roadmap for the Strategic Research Cluster on Space Robotics Technologies will be developed and monitored in several sub-projects. Important components of the cluster are extensive tests and evaluation measures for the developed technologies.

Field Tests in the Desert

The tests took place in the Erfoud region near the Algerian border, famous among tourists for its impressive sand dunes. With almost vegetationless stone deserts and reddish mountain ridges, the landscape comes very close to that of Mars. This makes it ideal for field tests under real-life conditions. The research team was able to put robots, sensors, and software through their paces. "Only through such field tests can we be sure that the software and hardware we have developed will work as planned under the harsh environmental conditions on Mars and the Moon," stresses Gianfranco Visentin, coordinator of the Peraspera project and head of the Automation and Robotics Department at the European Space Agency ESA.

The team used the first two weeks to prepare hardware and software systems for field use in a workshop. With the help of a flying drone, a detailed digital elevation map of the test area could be produced. At the end of November, the researchers moved to a base camp in the desert near Rissani. There, the new software technologies for autonomy and data fusion were tested in the "Ergo" and "InFuse" subprojects.

Rover Masters Task

In the subproject "Facilitators", coordinated by DFKI and managed by the Spanish company GMV, the developed software was tested. The hybrid walking and driving rover Sherpa TT of DFKI, which represents a new generation of space robots, served as test platform for the software. In addition, the pair of robots Minni and Manna of the CNRS-LAAS from Toulouse and a handheld Central Rover Unit of DLR Oberpfaffenhofen were used. Using the new software, the Sherpa TT was able to complete an autonomous long-distance mission. The driving rover covered a distance of 1.3 km over steep slopes and gorges before successfully taking a soil sample at its destination.

Together with the elevation map, the data collected by the rover represents an extensive set of test data. Thus the research partners were able to show that the developed software modules function under Mars-like conditions, but also collect a broad database for further research work. The tested techniques will now serve as core components for more complex robotics applications. Their development is planned from 2019 to 2021 as part of another series of EU-funded projects.

Bilder: 5

driving rover Sherpa TT

Test of the new software for the space robot Sherpa TT in the Moroccan desert.