GelLink: A Compact Multi-phalanx Finger with Vision-based Tactile Sensing and Proprioception

TL;DR

GelLink is a compact, underactuated, linkage-driven robotic finger with low-cost, high-resolution vision-based tactile sensing and proprioceptive sensing capabilities and the integration of vision- based tactile sensors can significantly enhance the capabilities of underactuated fingers and potentially broaden their future usage.

Abstract

Compared to fully-actuated robotic end-effectors, underactuated ones are generally more adaptive, robust, and cost-effective. However, state estimation for underactuated hands is usually more challenging. Vision-based tactile sensors, like Gelsight, can mitigate this issue by providing high-resolution tactile sensing and accurate proprioceptive sensing. As such, we present GelLink, a compact, underactuated, linkage-driven robotic finger with low-cost, high-resolution vision-based tactile sensing and proprioceptive sensing capabilities. In order to reduce the amount of embedded hardware, i.e. the cameras and motors, we optimize the linkage transmission with a planar linkage mechanism simulator and develop a planar reflection simulator to simplify the tactile sensing hardware. As a result, GelLink only requires one motor to actuate the three phalanges, and one camera to capture tactile signals along the entire finger. Overall, GelLink is a compact robotic finger that shows adaptability and robustness when performing grasping tasks. The integration of vision-based tactile sensors can significantly enhance the capabilities of underactuated fingers and potentially broaden their future usage.

Figures (9)

• Figure 1: From left to right: GelLink touching a plastic strawberry, the corresponding raw tactile images, unwarped images, and difference images
• Figure 2: Left: Linkage model of GelLink. Solid lines represent the actual linkage system. Dashed lines represent the outlines of the phalanges, where red, blue and green lines correspond to distal, intermediate, and proximal phalanges respectively. The gray lines represent aluminum bars. Right: Side view of the finger, where nodes are labeled in a similar manner to the linkage model.
• Figure 3: Left: 2D reflection simulation of GelLink. Joint 1 is the proximal interphalangeal joint (PIP), and Joint 2 is the distal interphalangeal joint (DIP). Right: Section view of the finger.
• Figure 4: Exploded view of GelLink. Main bodies of distal, intermediate, and proximal phalanges are rendered in red, blue, and green, respectively.
• Figure 5: Workflow of image data interpretation. 1) Global thresholding is applied to the captured raw image to generate masks for distorted tactile images. 2) Contours and convex hulls can be easily found from the mask image, which are highlighted with red and blue lines in the figure. 3) Convex hulls are fed to polygon approximation algorithm ramer1972iterative, which adapts contours with polygons and outputs the feature corners highlighted by green dots. 4) Unwarped tactile images and difference images (calculated as an after-contact tactile image subtracting an initial non-contact tactile image) can be computed. 5) A lookup table between joint angles and polygon vertices can be established for proprioceptive sensing.