Abstract-Motivated by the need of a robust and practical Inverse Kinematics (IK) algorithm for the WAM robot arm, we reviewed the most used Closed-Loop IK (CLIK) methods for redundant robots, analysing their main points of concern: convergence, numerical error, singularity handling, joint limit avoidance, and the capability of reaching secondary goals. As a result of the experimental comparison, we propose two enhancements. The first is a new filter for the singular values of the Jacobian matrix that guarantees that its conditioning remains stable, while none of the filters found in literature is successful at doing so. The second is to combine a continuous task priority strategy with selective damping to generate smoother trajectories. Experimentation on the WAM robot arm shows that these two enhancements yield an IK algorithm that improves on the reviewed state-of-the-art ones, in terms of the good compromise it achieves between time step length, Jacobian conditioning, multiple task performance, and computational time, thus constituting a very solid option in practice. This proposal is general and applicable to other redundant robots.

Dynamic Movement Primitives (DMPs) are nowadays widely used as movement parametrization for learning robot trajectories, because of their linearity in the parameters, rescaling robustness and continuity. However, when learning a movement with DMPs, a very large number of Gaussian approximations needs to be performed. Adding them up for all joints yields too many parameters to be explored when using Reinforcement Learning (RL), thus requiring a prohibitive number of experiments/simulations to converge to a solution with a (locally or globally) optimal reward. In this paper we address the process of simultaneously learning a DMPcharacterized robot motion and its underlying joint couplings through linear Dimensionality Reduction (DR), which will provide valuable qualitative information leading to a reduced and intuitive algebraic description of such motion. The results in the experimental section show that not only can we effectively perform DR on DMPs while learning, but we can also obtain better learning curves, as well as additional information about each motion: linear mappings relating joint values and some latent variables.

Abstract-Humans as well as humanoid robots can use a large number of degrees of freedom to solve very complex motor tasks. The high-dimensionality of these motor tasks adds difficulties to the control problem and machine learning algorithms. However, it is well known that the intrinsic dimensionality of many human movements is small in comparison to the number of employed DoFs, and hence, the movements can be represented by a small number of synergies encoding the couplings between DoFs. In this paper, we want to apply Dimensionality Reduction (DR) to a recent movement representation used in robotics, called Probabilistic Movement Primitives (ProMP). While ProMP have been shown to have many benefits, they suffer with the high-dimensionality of a robotic system as the number of parameters of a ProMP scales quadratically with the dimensionality. We use probablistic dimensionality reduction techniques based on expectation maximization to extract the unknown synergies from a given set of demonstrations. The ProMP representation is now estimated in the low-dimensional space of the synergies. We show that our dimensionality reduction is more efficient both for encoding a trajectory from data and for applying Reinforcement Learning with Relative Entropy Policy Search (REPS).

Abstract-Learning motion tasks in a real environment with deformable objects requires not only a Reinforcement Learning (RL) algorithm, but also a good motion characterization, a preferably compliant robot controller, and an agent giving feedback for the rewards/costs in the RL algorithm. In this paper, we unify all these parts in a simple but effective way to properly learn safety-critical robotic tasks such as wrapping a scarf around the neck (so far, of a mannequin).We found that a suitable compliant controller ought to have a good Inverse Dynamic Model (IDM) of the robot. However, most approaches to build such a model do not consider the possibility of having hystheresis of the friction, which is the case for robots such as the Barrett WAM. For this reason, in order to improve the available IDM, we derived an analytical model of friction in the seven robot joints, whose parameters can be automatically tuned for each particular robot. This permits compliantly tracking diverse trajectories in the whole workspace.By using such friction-aware controller, Dynamic Movement Primitives (DMP) as motion characterization and visual/force feedback within the RL algorithm, experimental results demonstrate that the robot is consistently capable of learning tasks that could not be learned otherwise.

Abstract-This paper presents a method to estimate external forces exerted on a manipulator during motion, avoiding the use of a sensor. The method is based on task-oriented dynamics model learning and a robust disturbance state observer. The combination of both leads to an efficient torque observer that can be incorporated to any control scheme. The use of a learning-based approach avoids the need of analytical models of joints' friction or Coriolis dynamics effects.