Dynamic Optimization Fabrics for Motion Generation

Abstract--Optimization fabrics are a geometric approach to realtime local motion generation, where motions are designed by the composition of several differential equations that exhibit a desired motion behavior. We generalize this framework to dynamic scenarios and non-holonomic robots and prove that fundamental properties can be conserved. We show that convergence to desired trajectories and avoidance of moving obstacles can be guaranteed using simple construction rules of the components. The open-source implementation can be found at https://github. Imagine physical limits and obstacle avoidance. It applications of such optimization-based approaches to mobile is requested to perform different tasks, such as cleaning the robots, the computational costs limit applicability when dealing floor or picking a wide range of products. Datadriven manipulation tasks may vary in their dimension and accuracy approaches to speed up the optimization process usually requirements, e.g. Thus, it is important for motion planning algorithms to Moreover, due to the scalar objective function, the user must support various goal definitions. Further, the robot is operating carefully weigh up different parts of the objective function. As alongside humans, it has to constantly react to the changing a consequence, optimization-based approaches are challenging environment and consequently update an initial plan. As to tune and inflexible to generic motion planning problems customers move fast, the adaptations must be computed in real with variable goal objectives [6, 7]. Therefore, motion planning is often divided into global motion planning [1] and local motion planning, which we will In the field of geometric control, namely Riemannian motion refer to as motion generation in this paper.