Semi-Autonomous Robot Support for A Mars Analog Exploration Cascade

In this article we present a semi-autonomous robotics framework developed to support the exploration cascade within the Mars analog simulation AMADEE 2024. The exploration cascade defines the process of refining the situation awareness during the scientific exploration of parts of a planet like Mars. Starting from remote sensing using orbiters over the use of UAVs down to in situ investigations using rovers or astronauts the understanding of an area or phenomena is improved. In particular the final step of inspecting, taking measurements, and collecting samples at the point of interest is of importance and the direct involvement of a human is a plus as the human is able to immediately provide a comprehensive intuitive interpretation. Because EVAs pose a risk and are demanding, having a remote inspection tool like a rover to eliminate less interesting or promising areas is of interest. Thus, we developed a semi-autonomous robot system able to travers challenging Mars-like terrain and to collect data such as images, 3D models, and spectroscopy measurements. The astronaut and the rover work as a team where the astronaut defines inspection goals and may take over if the rover gets into a unwanted or dangerous while most of the time the rover acts autonomously. The rover is integrated into a mission software architecture which stores all relevant information (e.g. missions, trajectories, goals, data) in a common GIS system in order to support the task execution by the astronaut but also the mission planning by the flight plan team as well as the interpretation of the data by the science team. We will discuss our design decisions derived from interaction with mission experts and will present preliminary results from a field deployment during the AMADEE 2024 Mars analog simulation.