Self-driving cars require an understanding of important edge case scenarios that can occur in driving scenes. SuperDepth is a neural network which addresses this challenge by processesing raw images and predicting the true per-pixel depth of all pixels in the input image, allowing for a complete 3D reconstruction of the driving scene. The outputs of SuperDepth can then be used to detect the object-ness of every pixel through methods such as 3D Voxel Analysis, to detect all static, movable, and moving objects in the scene, irrespective of what that object is. This type of classification-agnostic obstacle detection is an essential component in robustly dealing with 'long-tail' edge case scenarios. SuperDepth is part of the AutoSeg Foundation Model which forms the core of the vision-pipeline of the Autoware Autonomous Highway Pilot System