Autonomous Navigation in Lunar Lava Tubes: Sensing SLAM Trade-Offs and a Mission-Oriented GNC Architecture

Lunar lava tubes are subsurface cavities generated by volcanic activity and are regarded as promising targets for exploration because they can offer natural shielding and potentially support future lunar infrastructures as protected shelters and scientific laboratories. Autonomous navigation in these environments remains challenging due to the absence of illumination, sparse or repetitive geometric features, uneven terrain, and intermittent communications that limit teleoperation. In this framework, the Italian Space Agency (ASI) is pursuing a dedicated mission, and OHB Italia has been appointed the prime contractor to perform a candidate system-architecture study for lava tube exploration. This paper presents the activities and results related to the definition of the subsurface Guidance, Navigation, and Control (GNC) algorithm for a rover/hopper system. To address the above constraints, this study investigates the requirements for autonomous onboard navigation, focusing on sensor selection for Simultaneous Localization and Mapping (SLAM) as a fundamental prerequisite for mission success. A weighted-criteria evaluation framework is developed to assess various sensing modalities, considering mission-specific constraints. Based on this analysis, a sensor configuration optimized for GPS-denied and unilluminated environments is proposed. The effectiveness of the selected sensing architecture is validated through a simulation campaign conducted in simulation environments (CoppeliaSim v4.10.0/MATLAB 2025a), using two representative SLAM pipelines (ICP and LOAM) in LiDAR-only and LiDAR + IMU configurations. Finally, a modular Guidance, Navigation, and Control (GNC) architecture incorporating frontier-based exploration is proposed.