In the challenging context of human-robot collaboration, the ability to capture critical human-grasp-related features in a short period is essential for the transfer of skills/experience and the effective imitation learning of human grasping strategies to robotic systems. The paper at hand introduces a novel system for pose optimization of an active (moving) sensor, reducing uncertainty in grasp perception and enabling a more accurate observation of the human grasp pose. The proposed framework, coined as "GRAsPAD" (Generalized fRAmework for optimal graSp key-Points Active Detection), leverages a keypoint detection model to identify grasp-related features in a tomato fruit. The camera pose is iteratively adjusted through the Covariance Matrix Adaptation with Margin - Evolution Strategy (CMAwM-ES) to minimize uncertainty in detected keypoints, enhancing the accuracy of grasp perception and ensuring a clearer, more stable grasp representation. To mitigate sensor noise and variability, a Kalman filter is applied to refine keypoint detections over time. The experimental results demonstrate the significant improvement achieved by the pose optimization of the observer, in terms of grasp perception quality, which is critical for robotic skill learning and effective transfer of human grasping strategies to robotic systems.

Human-human learning is usually conducted utilizing multiple modalities (verbal, visual, kinesthetic etc), which is proven to significantly improve the quality of communication between the teacher and the learner. Towards extending this idea to human-robot learning, we propose here a multi-modal scheme for robot Learning by Demonstration (LbD), that combines visual observation of the human hand from a depthless RGB camera, with enhanced direct kinesthetic teaching of the robot. The proposed method, exploits the visually perceived action of the human hand, to provide haptic guidance during kinesthetic teaching, facilitating corrections along the axes with the maximum uncertainty. The method is experimentally evaluated using an RGB camera and a UR10e robot, and compared against single-modal approaches, i.e. visually trained behavior and trained through kinesthetic teaching. Our method is shown to outperform the latter, in terms of both the quality of skill transfer, and the cognitive/physical load required for the training.

Bio-inspired undulatory fin propulsion holds considerable potential for endowing robotic underwater vehicles with low-speed manoeuvrability and stable station-keeping. Robotic fins typically comprise a number of serially arranged and individually actuated “fin rays”, interconnected by a membrane-like flexible surface. Propulsive forces are generated by the propagation of a traveling wave along the mechanism, via appropriately timed ray oscillations. The present paper describes a dynamic model for an elementary two-ray fin system, analyzed as a standard robot mechanism with additional contributions arising from the elastic deformation of the flexible membrane and from the hydrodynamic forces. The model's main aspects, particularly with regard to the hydrodynamic effects, are explored via simulation studies, as well as via experiments with a robotic prototype. The developed model can serve a number of purposes towards optimizing the mechanical design, the control strategies, and the propulsive efficacy of robotic undulatory fins.

This work refers to a planar system of two spherical elastic fingertips grasping a rigid polygonal object under rolling contact constraints at equilibrium. A contact motion model is proposed for elastic-rolling fingertips based on previously reported experimental findings on the rolling distance for a variety of soft materials. The equilibrium conditions are derived from the system dynamics and depend on the deformation and on the kind of elastic material with regard to the fingertip's rolling distance characteristics. Equilibrium conditions of the three contacted bodies (elastic fingertips-object) show that in general no-collinear interaction forces act on the object's surface and they correspond to the equilibrium of an "internal object" hold by two point contacts with friction.

The control of object stable grasping and pose regulation by two robotic fingers with soft rolling contacts is considered. The motion of the dual fingers is confined to the horizontal plane and is not affected by the gravity force. A simple feedback control law is proposed with the task to achieve a desired normal contact force and appropriate tangential forces to ensure a dynamically stable grasp. The nonlinear asymptotic stability of the closed loop system is proved. Simulation results for two 3-DOF fingers manipulating a rectangular object from an initial state to a final stable grasp configuration demonstrate the effectiveness of the controller.

In this paper dealing with the problem of stable grasping of a rigid and rectangular object with two robotic fingers with soft tips, we consider the case of rolling fingertips under a soft contact motion model which leads to non-holonomic constraints even for the planar case while taking into consideration the gravity effect. The analytical hand-object system kinematics and dynamics are derived and a feedback controller is proposed. The controller compensates finger gravity and superimposes control signals that realize a stable grasp configuration and compensate for the object gravity forces while driving the object at the upright position. The controller is shown to achieve asymptotic convergence to the desired upright object position at a stable grasp configuration. Simulation results are presented confirming the theoretical findings.