Eye-in-Hand Visual Servoing of Concentric Tube Robots

International audienceThis paper deals with the development of a vision-based controller for a continuum robot architecture. More precisely, the controlled robotic structure is based on three-tube concentric tube robot (CTR), an emerging paradigm to design accurate, miniaturized, and flexible endoscopic robots. This approach has grown considerably in the recent years finding applications in numerous surgical disciplines. In contrast to conventional robotic structures, CTR kinematics arise many challenges for an optimal control such as friction, torsion, shear, and non-linear constitutive behavior. In fact, in order to ensure efficient and reliable control, in addition to computing an analytical and complete kinematic model, it is also important to close the control loop. To do this, we developed an eye-in-hand visual servoing scheme using a millimeter-sized camera embedded at the robot's tip. Both the kinematic model and the visual servoing controller were successfully validated in simulation with ViSP (Visual Servoing Platform) and using an experimental setup. The obtained results showed satisfactory performances for 3-degrees of freedom positioning and path following tasks with adaptive gain control