Turning a “Flaw” into a Superpower: Sheffield Researchers Redefine How Robots Move

The team of engineers from the School of Electronic and Electrical Engineering have turned a robotic ‘flaw’ into an advantage by using hysteresis to enable soft robots to move, morph, and grow with minimal actuation.

Turning a “Flaw” into a Superpower: Sheffield Researchers Redefine How Robots Move
Caption: The robot has intrinsic shape hysteresis and can be reconfigured into billions of different shapes with sequential commands of only two motors (A and B).

A research team led by Dr Lin Cao from the University of Sheffield’s School of Electrical and Electronic Engineering has reimagined one of robotics’ long-standing “flaws” as a breakthrough feature — unveiling a new way for soft robots to move, morph, and even “grow” with unprecedented dexterity.

 The study, published in Science Advances, introduces Hysteresis-Assisted Shape Morphing (HasMorph) — a concept that could change how engineers design flexible robots for medicine, industry, and disaster response.

From “More Motors” to “Smarter Motion”

Traditionally, roboticists have believed that to achieve more dexterous motion, robots need more actuators — like adding more strings to a puppet. But this makes robots bulky, expensive, and hard to control.
 Dr Cao’s team took a radically different view: what if fewer actuators could do more, by taking advantage of a natural mechanical behaviour that engineers usually try to eliminate — hysteresis.

Hysteresis occurs when a system’s motion doesn’t exactly retrace its path when forces are reversed — for instance, the small delay between gears switching direction because of clearance between gear teeth. “Instead of fighting this effect, we decided to use it,” said Dr Cao. “Hysteresis can actually be harnessed to make robots remember their previous shapes and perform complex movements with minimal actuation.”

Three Breakthroughs in One Concept

  1. Flipping the mindset – The team turned hysteresis from a system flaw into a design advantage, using it to create controllable, stable shape changes in soft robots.
     
  2. The HasMorph actuation paradigm – With only two tendons, the robot can control multiple bending sections independently, achieving billions of possible shapes. This represents a major shift from the conventional “more motors for more dexterity” approach.
     
  3. Reversible shape morphing for growing robots – By combining HasMorph with a tip-everting soft growing robot that extends at the tip like a plant, the team enabled dexterous control of both shape and growth direction. The robot can grow forward, steer around obstacles, follow the exact path of its tip (“follow-the-leader”), and even shorten from the tip — a long-sought capability in the field.

Why It Matters

This combination of simplicity and intelligence in motion opens new possibilities in several fields:

  • Minimally invasive surgery – A thin robotic endoscope could navigate frictionlessly inside the body, avoiding healthy tissue to reach target organs safely and precisely.
     
  • Search and rescue – Robots could move through collapsed structures or rubble to locate survivors.
     
  • Pipeline and structural inspection – Robots could explore confined, winding spaces without needing bulky mechanisms or multiple motors.
     

“For patients, this could mean safer, less traumatic procedures,” said Dr Cao. “For roboticists, HasMorph is a paradigm shift — it shows that more dexterous motion doesn’t always mean more motors. It means designing smarter.”

About the Study

The paper, “Hysteresis-assisted shape morphing for soft continuum robots,” was published in Science Advances in 2025.
 Authors: Zhezyuan Bi, Tianchen Ji, Sanja Dogramadzi, Soo Jay Louis Phee, Jiajun. Liu, Wenjie Lai, Diyang Wu, Bing Zhang, and Lin Cao.
 DOI: 10.1126/sciadv.adx3024