Soft Robots That Bend and Twist Like Human Fingers

Researchers devise a mathematical model to help design soft robots that move in a specified trajectory.

One of two soft robots designed with the new mathematical modelling technique, this robot mimics the bending of an index finger. (Image courtesy of Harvard School of Engineering and Applied Sciences.)

One of two soft robots designed with the new mathematical modelling technique, this robot mimics the bending of an index finger. (Image courtesy of Harvard School of Engineering and Applied Sciences.)

Researchers have developed a new strategy that aims to automate the design of soft robots. The approach uses mathematical modelling to design for a specified trajectory of movement, such as a robot that mimics the bending of a human finger.

The actuators used to create the soft robots are fluid-powered, fiber-reinforced actuators that can replicate a wide range of movements. Although many types of movements are possible, designing the actuators to output these movements isn’t a trivial task.

“The design is so complicated because one actuator type is not enough to produce complex motions,” said researcher Fionnuala Connolly of Harvard School of Engineering and Applied Sciences. “You need a sequence of actuator segments, each performing a different motion and you want to actuate them using a single input.”

There are four different actuator motions Connolly is referring to: extension, expansion, twisting, and bending. By combining different types of actuator segments, you can achieve a much more complex movement than any individual segment could offer on its own.

The researchers created analytical models of the four different motion types. Using these, they created a design tool that takes as input the kinematics of the desired motion, and outputs the optimal segment lengths and fiber angles to replicate the motion.

A) Analytical models were created for extension, expansion, twisting, and bending of the fluid-powered actuators. B) The desired motion is specified as input. C) The model outputs the fiber angles and segment lengths needed to replicate the motion. (Image courtesy of The Proceedings of the National Academy of Sciences/Harvard.)

A) Analytical models were created for extension, expansion, twisting, and bending of the fluid-powered actuators. B) The desired motion is specified as input. C) The model outputs the fiber angles and segment lengths needed to replicate the motion. (Image courtesy of The Proceedings of the National Academy of Sciences/Harvard.)

Using their model, the researchers designed two soft robots: one that bends like an index finger, and another that twists like a thumb. Both robots are powered by a single pressure source.

“This research streamlines the process of designing soft robots that can perform complex movements,” said researcher Conor Walsh. “It can be used to design a robot arm that moves along a certain path or a wearable robot that assists with motion of a limb.”

(Image courtesy of Harvard SEAS.)

(Image courtesy of Harvard SEAS.)

You can read the team’s research paper here. The new methodology will be included in the Soft Robotic Toolkit, an open-source online resource developed to assist researchers, educators and hobbyists to design, fabricate, model, characterize and control their own soft robots. 

For more soft robotics news, read How Robots Can Feel with Optical Waveguides.

Written by

Michael Alba

Michael is a senior editor at engineering.com. He covers computer hardware, design software, electronics, and more. Michael holds a degree in Engineering Physics from the University of Alberta.