<p>En este trabajo se presenta un análisis cinemático aplicado a dos manipuladores serie redundantes: el Kuka LWR 4+ y el ABB Yumi. En particular se deriva la cinemática directa para ambos manipuladores y se resuelve el problema de la cinemática inversa. Para el Kuka LWR 4+ dicha solución se obtiene en forma analítica, mientras que para el ABB Yumi se sigue un enfoque analítico y numérico. Además, se calculan simbólicamente tanto las singularidades del Kuka LWR 4+ como las direcciones singulares asociadas a estas. Este estudio contribuye al conocimiento cinemático de dos manipuladores redundantes de gran actualidad e interés para la comunidad robótica, y proporciona información útil para el diseño de diferentes algoritmos y leyes de control.</p>

In this paper, a new teleoperation system consisting on the integration of a mobile manipulator, an UAV, and a haptic is presented. The camera of the UAV is used to give visual feedback to the operator. An algorithm is presented to allow the operator to command both the UAV and the mobile manipulator while keeping the point of view pointing towards the robot by only using a single haptic device. The presented algorithm combines a position-position and a position-velocity workspace mapping from the haptic to the mobile manipulator and the UAV, both in position and orientation. Further, a solution is presented that takes advantage of the null space of the mobile manipulator to keep the body and the arm of the mobile manipulator from occluding its own TCP and the object it is carrying, thus easing the teleoperation task. Experimentation has been carried on the system, both in a virtual and a real scenario, showing its potential in teleoperation scenarios. Overall, a novel teleoperation system and the ongoing progress towards its implementation in real situations is presented in this work.

The paper presents a novel teleoperation system that allows the simultaneous command of a mobile manipulator and a free flying camera, implemented using a UAV, from which the operator can monitor the task execution in real-time. A novel use of the kinematic redundancy is presented to prevent the robot parts from occluding the end-effector to the operator view. Following an obstacle avoidance approach, an input is fed to the null space that keeps the robot links away from the end-effector in the image plane as a secondary task. Simulations and the implementation in a real setup show the goodness of the proposed approach.