Reaching Through Latent Space: From Joint Statistics to Path Planning in Manipulation

Abstract--We present a novel approach to path planning for robotic manipulators, in which paths are produced via iterative optimisation in the latent space of a generative model of robot poses. Constraints are incorporated through the use of constraint satisfaction classifiers operating on the same space. Optimisation leverages gradients through our learned models that provide a simple way to combine goal reaching objectives with constraint satisfaction, even in the presence of otherwise non-differentiable constraints. Our models are trained in a task-agnostic manner on randomly sampled robot poses. In baseline comparisons against a number of widely used planners, we achieve commensurate performance in terms of task success, planning time and path length, performing successful path planning with obstacle avoidance on a real 7-DoF robot arm. Once trained, gradients through the VAE decoder and collision I. INTRODUCTION ATH planning is a cornerstone of robotics. This requires that path to the target that satisfies the collision constraint. Due to its importance, path planning is a richly planning and control (e.g.