Enhancing Safety under Milligravity Condition: Robust Attitude Estimation for the MMX Rover IDEFIX

Kurzfassung

During planetary exploration, the safety of rovers plays a crucial role, as hazardous movements may result in irreparable damage. Ensuring long-term operational capability of such vehicles requires minimizing risks and preserving integrity. Critical aspects of risk reduction involve tracking the rover attitude to prevent unstable configuration. There, unmonitored commands can lead to catastrophic outcomes such as tumbling. By predicting the rover's behavior, autonomous systems can abort movements to maintain safe operational boundaries. This paper addresses the challenge in the context of IDEFIX, a non-steerable 4-wheeled rover designed by CNES and DLR. IDEFIX will operate on Phobos, where gravity is roughly 1/2000 of Earth's. To overcome sensing limitations, we propose a linear Kalman Filter, integrating gyroscope data to estimate the rover attitude robustly. A tailored sensor calibration technique is adjointly developed to enhance reliability and robustness. Results demonstrate that the method improves real-time attitude awareness and prediction, enabling safer locomotion in extreme extraterrestrial conditions.