Researchers from Massachusetts Institute of Technology (MIT) have developed long, curved, human finger-shaped, camera-based touch sensors to improve the grasping mechanism of robotic grippers.

MIT’s GelSight Svelte sensor uses cameras to cull information about objects to be grasped — much like the sensory receptors in human skin, which send information to the brain about what is in its grasp.

Source: MITSource: MIT

To accomplish this, the sensor features two strategically placed mirrors that both reflect and refract light, enabling one camera that is located at the base of the sensor to “see” along the entire length of the finger-shaped sensor. In addition to mirrors and cameras, the sensor also features LEDs and everything is attached to a flexible plastic backbone encased in flexible skin made of silicone gel.

This design reportedly enables three-finger robotic hands to grasp heavy objects using the entire sensing area of all of the robotic fingers that make up the robotic gripper.

According to the developers of the GelSight Svelte, the sensor’s shape enables robotic grippers to use different grasps for different tasks instead of just the pinching grasps common among most robotic grippers.

An article detailing the sensor, GelSight Svelte: A Human Finger-shaped Single-camera Tactile Robot Finger with Large Sensing Coverage and Proprioceptive Sensing, appears in the journal arXiv.

For more on the GelSight Svelte sensor, watch the accompanying video that appears courtesy of MIT.

To contact the author of this article, email