Sometimes our best ideas come to us when we are least looking for them. A number of otherwise seemingly buttoned-up professionals have dreamed up their best inventions while playing with toys, of all things. For instance, Stanford bioengineers created a human-powered centrifuge, capable of spinning at over 20,000 RPM, that was inspired by a button whirligig. NASA engineers also spend some time playing with toys apparently, as a stacking ring toy led them to the perfect solution for a compact, collapsible heat shield that may one day be used by a Mars lander.
Now, the world of soft robotics is being moved forward by fidget popper toys. These toys come in all sorts of shapes and sizes, but their defining features are little circles that remain stable in one of two states, storing energy that can be released in a powerful “pop” with just a little nudge. A group of engineers at Purdue University realized that this same mechanism could be useful in designing soft robots. So they designed and developed bistable components that can be used as grippers or walkers for soft robots that produce big movements with small energy inputs.
This gripper can take a lickin’ and keep on tickin’ (📷: Purdue University)
Soft robots are known for their flexibility, adaptability, and inherent safety, allowing them to perform tasks that rigid robots struggle with. They can bend, twist, and squeeze into places traditional machines cannot reach, making them valuable for applications like delicate medical procedures or interacting safely with humans. But this very flexibility also makes them difficult to control. Unlike rigid robots, which can be precisely modeled with straightforward equations, soft robots have nearly infinite degrees of freedom and nonlinear material responses that make predicting their behavior a big challenge.
The researchers approached this challenge by drawing inspiration from the bistability of fidget poppers, which naturally settle into two stable states. By incorporating similar structures — dome-shaped units that can “snap” between different configurations — they created what they call Dome Phalanx Fingers. These building blocks can be combined into robotic hands, grippers or walking machines. This design discretizes the robot’s otherwise continuous and unpredictable movements into a manageable set of possible configurations.
This discretization opens the door to a new form of control that does not rely on heavy computing power or complex sensors. Instead, the control is embodied in the robot’s structure itself. Each dome unit can be tuned to respond to specific pneumatic inputs, producing predictable motions such as grasping, releasing, or stepping. This approach allows for simpler open-loop control, where the robot executes tasks without constant feedback or adjustment.
Demonstrations of this approach included a soft gripper capable of distinguishing between objects of different sizes and weights, as well as a six-legged walker that could move and steer with nothing more than simple pressure modulations. In each case, the use of multistability provided built-in reliability. The robots snapped into predictable configurations without requiring continuous sensing or precise pressure control. This not only simplified the control problem but also made the robots more tolerant to imperfections and damage.
I don’t know about you, but this work has inspired me to get out some old toys for… err… research purposes. LEGO bins, here I come!
