Inductive sensors and shape reconstruction for intraluminal flexible and soft robots

Soft and flexible robots are transforming medical interventions, thanks to their flexibility, miniaturization, and multidirectional movement abilities. In recent days, to increase the performance of the controlling action in unstructured and dynamic environments, shape reconstruction methods have been developed enormously. However, the methods presented in the literature still lack to address the issue of soft and flexible robots. In this thesis, a method to reconstruct the shape of soft and flexible robots using a stretchable sensor is presented. This thesis is divided into two parts. In the first part, a stretchable sensor is selected by investigating hysteresis, stretchability, and feasibility of integration. It started with resistive sensors which shows that dealing with high hysteresis and non-linearities needs a complicated and advanced fabrication process. Given the effort needed to reduce the hysteresis and non-linearities, a stretchable sensor based on the electrical inductive principle is considered. During the investigation, three contributions are developed. First, a simple voltage measurement across a conventional spring acting as a variable inductor is presented for strain estimation. This simplifies the measuring equipment and also allows for the optimization of the measurement conditions. Second, an experimental investigation shows that a single calibration curve generated outside of the robot is enough to estimate the linear or bending strain of the robot, given the system type. Third, a full characterization is conducted that shows the sensor is capable of detecting large deformation while showing high precision and accuracy. To show the feasibility of the sensor integration process, the sensor is integrated into a soft and a flexible robot and closed loop control are performed. The second part focuses on the shape reconstruction of soft and flexible robots employed in minimally invasive surgery using the inductive sensor. The shape reconstruction process is performed in 2D and 3D workspace, and the corresponding accuracy is estimated by comparing with ground truth values. Ultimately, the thesis is concluded by providing remarks on the use of the inductive sensor for the shape reconstruction of soft and flexible robots while highlighting the limitations and perspective of future work.