No abstract

No abstract

No abstract

Human robot cooperation is a challenging task of future generation home robots and is relevant also for industrial applications, where robots are supposed to act together with humans in non-structured environments. The paper focuses on on-line adaptation of robot trajectories, where robots and humans are autonomous agents coupled only through the manipulated object. Within the proposed approach, the robot adapts to the human intentions through the sensory feedback, where safety is one of the most important issues. The algorithm is based on representation of trajectories with the Dynamic Movement Primitives, where the adaptation of the corresponding robot trajectories relies on the Iterative Learning Controller framework. In order to demonstrate the effectiveness of the proposed approach, we applied two KUKA LWR robots in a bimanual human-robot collaborative task.all acquired at times t j , j = 0, . . . , T . Positions and orientations are represented as a frame Q = {p, q}, where p = [x, y, z] is a 3-dimensional vector and q = (v, u) is a unit quaternion, respectively. Positions and orientations are represented as a frame Q = {p, q}, where p ∈ R 3 and 978-1-4799-7397-2/14/$31.00

A new method to recognize the intention of a human worker while performing a collaborative task with a robot is proposed. For this purpose, two recurrent neural network (RNN) architectures capable of predicting the worker's target were developed. The first uses marker-based tracking of hand positions and the second RGB-D videos of human motion. The system was implemented to perform a collaborative assembly task. The results show high intention prediction accuracy for both networks, with accuracy increasing once a larger portion of human motion has been observed, making the proposed method viable for efficient and dynamic humanrobot collaboration. Furthermore, we developed a framework that enables online adaptation of robot trajectories based on estimated human intentions.