Deep probabilistic traversability with test-time adaptation for uncertainty-aware planetary rover navigation

Traversability assessment of deformable terrain is vital for safe rover navigation on planetary surfaces. Machine learning (ML) is a powerful tool for traversability prediction but its inherent predictive uncertainty increases the risk of wheel slips and permanent rover immobilization. To address this issue, we integrate principal approaches to uncertainty handling—quantification, exploitation, and adaptation—into a single learning and planning framework for rover navigation. The key concept is deep probabilistic traversability, an end-to-end probabilistic ML model that predicts slip distributions directly from terrain appearance and geometry. This probabilistic model quantifies uncertainties in slip prediction and exploits them as uncertainty-aware traversability costs in path planning. Its end-to-end nature also allows adaptation of pre-trained models with in-situ traverse experience to reduce prediction errors. We perform extensive simulations in synthetic environments posing representative uncertainties in planetary analog terrains. Simulation results show that our method achieves more robust path planning under novel environmental conditions than existing methods.