Reactive Motion Generation on Learned Riemannian Manifolds
This work addresses motion generation for robots, enabling adaptation to unseen conditions like obstacle avoidance, but it appears incremental as it builds on existing motion learning and Riemannian manifold concepts.
The paper tackles robot motion learning by proposing a method to learn Riemannian manifolds from human demonstrations, where geodesics serve as motion skills, and introduces obstacle avoidance by reshaping the manifold. The result includes successful testing on a 7-DoF manipulator for learning complex patterns and generating trajectories in multiple-mode settings.
In recent decades, advancements in motion learning have enabled robots to acquire new skills and adapt to unseen conditions in both structured and unstructured environments. In practice, motion learning methods capture relevant patterns and adjust them to new conditions such as dynamic obstacle avoidance or variable targets. In this paper, we investigate the robot motion learning paradigm from a Riemannian manifold perspective. We argue that Riemannian manifolds may be learned via human demonstrations in which geodesics are natural motion skills. The geodesics are generated using a learned Riemannian metric produced by our novel variational autoencoder (VAE), which is especially intended to recover full-pose end-effector states and joint space configurations. In addition, we propose a technique for facilitating on-the-fly end-effector/multiple-limb obstacle avoidance by reshaping the learned manifold using an obstacle-aware ambient metric. The motion generated using these geodesics may naturally result in multiple-solution tasks that have not been explicitly demonstrated previously. We extensively tested our approach in task space and joint space scenarios using a 7-DoF robotic manipulator. We demonstrate that our method is capable of learning and generating motion skills based on complicated motion patterns demonstrated by a human operator. Additionally, we assess several obstacle avoidance strategies and generate trajectories in multiple-mode settings.