MSGID: 1:317/3 642ba7e5   
   PID: hpt/lnx 1.9.0-cur 2019-01-08   
   TID: hpt/lnx 1.9.0-cur 2019-01-08   
    Robotic hand can identify objects with just one grasp    
    The three-fingered robotic gripper can 'feel' with great sensitivity   
   along the full length of each finger -- not just at the tips    
      
     Date:   
         April 3, 2023   
     Source:   
         Massachusetts Institute of Technology   
     Summary:   
         Newly created soft-rigid robotic fingers incorporate powerful   
         sensors along their entire length, enabling them to produce a   
         robotic hand that could accurately identify objects after only   
         one grasp.   
      
      
         Facebook Twitter Pinterest LinkedIN Email   
   FULL STORY   
   ==========================================================================   
   Inspired by the human finger, MIT researchers have developed a robotic   
   hand that uses high-resolution touch sensing to accurately identify an   
   object after grasping it just one time.   
      
      
   ==========================================================================   
   Many robotic hands pack all their powerful sensors into the fingertips, so   
   an object must be in full contact with those fingertips to be identified,   
   which can take multiple grasps. Other designs use lower-resolution sensors   
   spread along the entire finger, but these don't capture as much detail,   
   so multiple regrasps are often required.   
      
   Instead, the MIT team built a robotic finger with a rigid skeleton   
   encased in a soft outer layer that has multiple high-resolution sensors   
   incorporated under its transparent "skin." The sensors, which use a   
   camera and LEDs to gather visual information about an object's shape,   
   provide continuous sensing along the finger's entire length. Each finger   
   captures rich data on many parts of an object simultaneously.   
      
   Using this design, the researchers built a three-fingered robotic hand   
   that could identify objects after only one grasp, with about 85 percent   
   accuracy.   
      
   The rigid skeleton makes the fingers strong enough to pick up a heavy   
   item, such as a drill, while the soft skin enables them to securely grasp   
   a pliable item, like an empty plastic water bottle, without crushing it.   
      
   These soft-rigid fingers could be especially useful in an at-home-care   
   robot designed to interact with an elderly individual. The robot could   
   lift a heavy item off a shelf with the same hand it uses to help the   
   individual take a bath.   
      
   "Having both soft and rigid elements is very important in any hand,   
   but so is being able to perform great sensing over a really large area,   
   especially if we want to consider doing very complicated manipulation   
   tasks like what our own hands can do. Our goal with this work was to   
   combine all the things that make our human hands so good into a robotic   
   finger that can do tasks other robotic fingers can't currently do,"   
   says mechanical engineering graduate student Sandra Liu, co-lead author   
   of a research paper on the robotic finger.   
      
   Liu wrote the paper with co-lead author and mechanical engineering   
   undergraduate student Leonardo Zamora Yan~ez and her advisor, Edward   
   Adelson, the John and Dorothy Wilson Professor of Vision Science in the   
   Department of Brain and Cognitive Sciences and a member of the Computer   
   Science and Artificial Intelligence Laboratory (CSAIL). The research   
   will be presented at the RoboSoft Conference.   
      
   A human-inspired finger The robotic finger is comprised of a rigid,   
   3D-printed endoskeleton that is placed in a mold and encased in a   
   transparent silicone "skin." Making the finger in a mold removes the   
   need for fasteners or adhesives to hold the silicone in place.   
      
   The researchers designed the mold with a curved shape so the robotic   
   fingers are slightly curved when at rest, just like human fingers.   
      
   "Silicone will wrinkle when it bends, so we thought that if we have the   
   finger molded in this curved position, when you curve it more to grasp   
   an object, you won't induce as many wrinkles. Wrinkles are good in some   
   ways -- they can help the finger slide along surfaces very smoothly and   
   easily -- but we didn't want wrinkles that we couldn't control," Liu says.   
      
   The endoskeleton of each finger contains a pair of detailed touch sensors,   
   known as GelSight sensors, embedded into the top and middle sections,   
   underneath the transparent skin. The sensors are placed so the range   
   of the cameras overlaps slightly, giving the finger continuous sensing   
   along its entire length.   
      
   The GelSight sensor, based on technology pioneered in the Adelson group,   
   is composed of a camera and three colored LEDs. When the finger grasps   
   an object, the camera captures images as the colored LEDs illuminate   
   the skin from the inside.   
      
   Using the illuminated contours that appear in the soft skin, an algorithm   
   performs backward calculations to map the contours on the grasped object's   
   surface. The researchers trained a machine-learning model to identify   
   objects using raw camera image data.   
      
   As they fine-tuned the finger fabrication process, the researchers ran   
   into several obstacles.   
      
   First, silicone has a tendency to peel off surfaces over time. Liu and   
   her collaborators found they could limit this peeling by adding small   
   curves along the hinges between the joints in the endoskeleton.   
      
   When the finger bends, the bending of the silicone is distributed along   
   the tiny curves, which reduces stress and prevents peeling. They also   
   added creases to the joints so the silicone is not squashed as much when   
   the finger bends.   
      
   While troubleshooting their design, the researchers realized wrinkles   
   in the silicone prevent the skin from ripping.   
      
   "The usefulness of the wrinkles was an accidental discovery on our   
   part. When we synthesized them on the surface, we found that they actually   
   made the finger more durable than we expected," she says.   
      
   Getting a good grasp Once they had perfected the design, the researchers   
   built a robotic hand using two fingers arranged in a Y pattern with a   
   third finger as an opposing thumb.   
      
   The hand captures six images when it grasps an object (two from each   
   finger) and sends those images to a machine-learning algorithm which   
   uses them as inputs to identify the object.   
      
   Because the hand has tactile sensing covering all of its fingers, it   
   can gather rich tactile data from a single grasp.   
      
   "Although we have a lot of sensing in the fingers, maybe adding a palm   
   with sensing would help it make tactile distinctions even better,"   
   Liu says.   
      
   In the future, the researchers also want to improve the hardware to   
   reduce the amount of wear and tear in the silicone over time and add   
   more actuation to the thumb so it can perform a wider variety of tasks.   
      
   This work was supported, in part, by the Toyota Research Institute,   
   the Office of Naval Research, and the SINTEF BIFROST project.   
      
       * RELATED_TOPICS   
             o Matter_&_Energy   
                   # Robotics_Research # Detectors # Civil_Engineering #   
                   Engineering_and_Construction   
             o Computers_&_Math   
                   # Robotics # Neural_Interfaces # Artificial_Intelligence   
                   # Computers_and_Internet   
       * RELATED_TERMS   
             o Robotic_surgery o Robot o Trigonometry o Aircraft o Radar   
             o Denaturation_(biochemistry) o Vanadium o Carpal_tunnel   
      
   ==========================================================================   
   Story Source: Materials provided by   
   Massachusetts_Institute_of_Technology. Original written by Adam   
   Zewe. Note: Content may be edited for style and length.   
      
      
   ==========================================================================   
   Related Multimedia:   
       * Soft-rigid_robotic_finger   
   ==========================================================================   
   Journal Reference:   
      1. Sandra Q. Liu, Leonardo Zamora Yan~ez, Edward H. Adelson. GelSight   
         EndoFlex: A Soft Endoskeleton Hand with Continuous   
         High-Resolution Tactile Sensing. Submitted to arXiv, 2023 DOI:   
         10.48550/arXiv.2303.17935   
   ==========================================================================   
      
   Link to news story:   
   https://www.sciencedaily.com/releases/2023/04/230403133515.htm   
      
   --- up 1 year, 5 weeks, 10 hours, 50 minutes   
    * Origin: -=> Castle Rock BBS <=- Now Husky HPT Powered! (1:317/3)   
   SEEN-BY: 15/0 106/201 114/705 123/120 153/7715 226/30 227/114 229/110   
   SEEN-BY: 229/111 112 113 307 317 400 426 428 470 664 700 292/854 298/25   
   SEEN-BY: 305/3 317/3 320/219 396/45   
   PATH: 317/3 229/426