Hopping towards a better soft robot

Traditional industrial robots are rigid — mostly metal — and are fast, precise and powerful. Their speed and precision comes at the cost of complexity and can often pose a danger to humans who get too close. Soft robots are adaptable and resilient but slow, difficult to fabricate, and challenging to make autonomous because most motors, pumps, batteries, sensors, and microcontrollers are rigid. 

But what if you could combine the autonomy and speed of a rigid robot with the adaptability and resiliency of a soft robot — and do it relatively cheap and fast?

Harvard engineers have done just that, developing one of the first 3-D printed, soft robots that moves autonomously. Described in the July 10, 2015 issue of the journal Science, the design offers a new solution to an engineering challenge that has plagued soft robotics: the integration of rigid and soft materials.  

The robot’s body transitions from soft to hard, reducing the stress where the rigid electronic components join the body and increasing the robot’s resiliency. The body’s monolithic design — created in one continuous print job, using several different materials — increases its strength and robustness.  With no sliding parts or traditional joints, the robot isn’t victim to dirt or debris like its more intricate cousins, making it a good candidate for use in harsh terrains.

“The vision for the field of soft robotics is to create robots that are entirely soft,” said senior author Robert J. Wood, Charles River Professor of Engineering and Applied Sciences at the Harvard John A. Paulson School of Engineering and Applied Sciences (SEAS) and Core Faculty Member at the Wyss Institute for Biologically Inspired Engineering at Harvard. “But for practical reasons, our soft robots typically have some rigid components — things like batteries and control electronics. This robot is a demonstration of a method to integrate the rigid components with the body of the soft robot through a gradient of material properties, eliminating an abrupt hard-to-soft transition that is often a failure point.”

The combustion-powered robot — reminiscent of a toy rubber popper — is comprised of two main parts: a soft plunger-like body with three pneumatic legs and the rigid core module, containing power and control components and protected by a semi-soft 3-D printed shield. The design builds from previous work of co-author and chemist George Whitesides, the Woodford L. and Ann A. Flowers University Professor.

To initiate movement, the robot inflates its pneumatic legs to tilt its body in the direction it wants to go. Then butane and oxygen are mixed and ignited, catapulting the robot into the air. It’s a powerful jumper, reaching up to six times its body height in vertical leaps and half its body width in lateral jumps. In the field, the hopping motion could be an effective way to move quickly and easily around obstacles.

photo/Harvard Microrobotics Lab

“The wonderful thing about soft robots is that they lend themselves nicely to abuse,” said Nicholas Bartlett, first author of the paper and a graduate student at SEAS. “The robot’s stiffness gradient allows it to withstand the impact of dozens of landings and to survive the combustion event required for jumping. Consequently, the robot not only shows improved overall robustness but can locomote much more quickly than traditional soft robots.”

The robot’s jumping ability and soft body would come in handy in harsh and unpredictable environments or disaster situations, allowing it to survive large falls and other unexpected situations. 

This new design demonstrates the possibilities of 3-D printing in soft robotics. Traditional methods of fabrication — custom molds and multi-step assembly — are costly and slow. The ever-increasing variety of materials compatible with 3-D printers is allowing engineers to prototype new designs faster, and increased complexity does not necessarily lead to increased cost. 

“Soft robotics is a relativity nascent subfield and 3-D printing is adding to the repertoire of things we can do in a really practical way,” said Wood. 

The co-first author of the paper is Michael Tolley, of University of California, San Diego. Coauthors include Johannes T.B. Overvelde and Katia Bertoldi of SEAS; James Weaver of the Wyss Institute; Bobak Mosadegh of Weill Cornell Medical College; and George Whitesides of Harvard and the Wyss Institute. 

This research was supported by the National Science Foundation and Wyss Institute at Harvard University. Harvard’s Office of Technology Development has filed for a provisional patent on the technology and is actively pursuing commercialization opportunities.