Faye Wu wears the supernumerary robotic fingers (Melanie Gonick, MIT)
18 July 2014. Engineers at Massachusetts Institute of Technology designed a glove-like robotic device that adds two more fingers and coordinates with a person’s hand to help with manual activities. Mechanical engineering professor Harry Asada and graduate student Faye Wu discussed the device earlier this week at the Robotics Science and Systems conference in Berkeley, California.
Asada and Wu are seeking to build a device that can help people with limited hand functions or only one hand perform day-to-day activities, such as lifting objects or opening a letter. Rather than building a device that requires separate commands, the researchers instead are building a system that adjusts to and coordinates with an individual’s natural gripping patterns.
The device designed by Asada and Wu, called supernumerary robotic fingers, has two larger fingers on either side of a glove with sensors and actuators worn on the wrist. The two larger fingers make it possible for the wearer to hold and lift heavier objects.
The researchers devised an algorithm to coordinate the two extra fingers with the motions of the wearer’s natural hand and five fingers. In first learning the physiology of hand movements, Asada and Wu discovered the muscles in a person’s hands and fingers are highly coordinated. And while grasping various objects requires some differences in muscle movements, they discovered in grasping any object, the hand uses the same basic two actions: bringing the fingers together and closing them in toward the palm.
When adding the two robotic fingers, Wu — who conducted tests of the device — discovered a similar pattern. She grabbed various objects, from a cookie to a football, multiple times and from various angles, with the hand assisted by the robotic device, recording the movements and actions each time. The tests revealed two or three basic grasping actions, when using the robotic device.
The algorithm then reads the posture of the wearer’s hand and coordinates the movements of the two extra fingers to enhance a person’s grip when performing manual activities. In further development of the device, Wu seeks to better understand the amount of force needed to assist the human grasp. “With an object that looks small but is heavy, or is slippery,” says Wu in a university statement, “the posture would be the same, but the force would be different, so how would it adapt to that?”
The researchers hope to compile a collection of posture and force patterns for the algorithm in next versions of the device. Because of subtle differences in grasping behavior between individuals, future versions may need to learn a person’s grasp, much like voice command systems today learn a person’s vocal patterns.
In the following video Asada and Wu discuss and demonstrate the supernumerary robotic fingers.
- FDA Approves Motorized Exoskeleton for Spinal Cord Injury
- University, Spin-Off Partner on Enviro Flying Robotic Device
- Mobile Low-Power Gesture-Recognition System Developed
- Bio-Inspired Cooperative Robotic Control System Developed
- Computer Learning Process Developed for Robotic Arm Movement
* * *
RSS - Posts
You must be logged in to post a comment.