Designing an Adaptive Velocity Obstacle Avoidance System for Autonomous Mars Rover Navigation in Dynamic Terrains

Over the past decade, various velocity obstacle-based methodologies have been developed to address collision avoidance in dynamic environments. Despite their advancements, these methods often face limitations when managing multiple obstacles, sequential encounters, or maintaining safety in complex, unstructured terrains. This paper presents an adaptive collision avoidance strategy based on the velocity obstacle method, tailored to enable autonomous Mars rovers to navigate safely in dynamic and uncertain terrains while effectively avoiding multiple obstacles.

The proposed strategy introduces an adaptive velocity cone framework, which dynamically accounts for moving obstacles and terrain features. This approach ensures continuous safety and seamless navigation toward designated waypoints. By integrating risk metrics, such as distance and time to the closest point of approach, the system dynamically switches between target-reaching and collision-avoidance modes, maintaining both efficiency and robustness.

We validate the proposed methodology through simulations of Mars exploration scenarios, encompassing challenging multi-obstacle environments. The results demonstrate significant performance improvements, including increased safety distances and enhanced adaptability in navigating unstructured terrains. These findings highlight the suitability of the approach for autonomous planetary exploration, where collision avoidance is critical to safeguarding sensitive equipment, optimizing energy consumption, and ensuring mission success.

The adaptive strategy represents a step forward in ensuring reliable navigation in extraterrestrial settings, paving the way for safer and more efficient Mars rover missions