Increasing the autonomy of planetary exploration robots

Mobile robots designed for planetary exploration are essential for advancing our understanding of celestial bodies, enabling scientists to collect valuable data and explore areas beyond the reach of human astronauts. These robots are equipped with advanced sensors, cameras, scientific instruments, and communication systems, allowing them to navigate autonomously, analyze the environment, and transmit data back to Earth. This thesis focuses on increasing the degree of autonomy of planetary exploration robots. The research objectives are divided into three main topics: • Terrain Awareness: This research leverages machine learning algorithms to classify terrain types using proprioceptive data modulated by roverenvironment interactions. By selecting informative subsets of this data, the aim is to improve terrain classification for both generalization and extrapolation, crucial for long-range navigation and safety • Innovative Suspension Systems: a novel o↵-road tracked robot is introduced in this study, and its suspension system is evaluated through analytical and multibody models. By analyzing its performance on challenging terrains this research contributes to improving the design of robotic systems for planetary exploration. • Path Planning: This work delves into reactive computing path planning, focusing on the use of Harmonic Artificial Potential Fields (HAPF). The aim is to enhance local and global planning while addressing typical reactive computing limitations, such as local minima and sub-optimality. By addressing these research objectives, this thesis seeks to advance the capabilities of mobile robots for planetary exploration, contributing to safer and more ecient exploration of celestial bodies, such as Mars and the Moon.