How Does It Feel? Self-Supervised Costmap Learning for Off-Road Vehicle Traversability

Abstract-- Estimating terrain traversability in off-road environments requires reasoning about complex interaction dynamics between the robot and these terrains. However, it is challenging to create informative labels to learn a model in a supervised manner for these interactions. We propose a method that learns to predict traversability costmaps by combining exteroceptive environmental information with proprioceptive terrain interaction feedback in a self-supervised manner. Additionally, we propose a novel way of incorporating robot velocity into the costmap prediction pipeline. Yet, this abstracts away all the nuance of Outdoor, unstructured environments are challenging for the interactions between the robot and different terrain types. Rough interactions with terrain can result Under an occupancy-based paradigm, concrete, sand, and in a number of undesirable effects, such as rider discomfort, mud would be equally traversable, whereas tall rocks, grass, error in state estimation, or even failure of robot components. In reality, Unfortunately, it can be challenging to predict these interactions specific instances of a class may have varying degrees of a priori from exteroceptive information alone. Yet, what we are compliance of the objects on the ground, affect the dynamics really interested in capturing is roughness as the vehicle of the robot as it traverses over these features.