Concentric-tube robots (CTR) consist of a series of pre-curved flexible tubes that make up the robot structure and provide the high dexterity required for performing surgical tasks in constrained environments. This special design introduces new challenges in shape sensing as large twisting is experienced by the torsionally compliant structure. In the literature, fiber Bragg grating (FBG) sensors are attached to needle-sized continuum robots for curvature sensing, but they are limited to obtaining bending curvatures since a straight sensor layout is utilized. For a CTR, in addition to bending curvatures, the torsion along the robots shaft should be determined to calculate the shape and pose of the robot accurately. To solve this problem, in our earlier work, we proposed embedding FBG sensors in a helical pattern into the tube wall. The strain readings are converted to bending curvatures and torsion by a strain-curvature model. In this paper, a modified strain-curvature model is proposed that can be used in conjunction with standard shape reconstruction algorithms for shape and pose calculation. This sensing technology is evaluated for its accuracy and resolution using three FBG sensors with 1 mm sensing segments that are bonded into the helical grooves of a pre-curved Nitinol tube. The results show that this sensorized robot can obtain accurate measurements: resolutions of 0.02 rad/m with a 100 Hz sampling rate. Further, the repeatability of the obtained measurements during loading and unloading conditions are presented and analyzed.
This paper presents the development and application of an approach for sensorizing a surgical robotic instrument for two degree-of-freedom (DOF) lateral force sensing. The sensorized instrument is compatible with the da Vinci® Surgical System and can be used for skills assessment and force control in specific surgical tasks. The sensing technology utilizes a novel layout of four fiber Bragg grating (FBG) sensors attached to the shaft of a da Vinci® surgical instrument. The two cross-section layout is insensitive to error caused by combined force and torque loads, and the orientation of the sensors minimizes the condition number of the instrument’s compliance matrix. To evaluate the instrument’s sensing capabilities, its performance was tested using a commercially available force-torque sensor, and showed a resolution of 0.05N at 1 kHz sampling rate. The performance of the sensorized instrument was evaluated by performing three surgical tasks on phantom tissue using the da Vinci® system with the da Vinci Research Kit (dVRK): tissue palpation, knot tightening during suturing and Hem-O-Lok® tightening during knotless suturing. The tasks were designed to demonstrate the robustness of the sensorized force measurement approach. The paper reports the results of further evaluation by a group of expert and novice surgeons performing the three tasks mentioned above.