A smarter way for robots to grip

PHOTO © Nay | ADOBE STOCK

A breakthrough slip-prevention method has been shown to improve how robots grip and handle fragile, slippery, or asymmetric objects, according to a University of Surrey-led study published in Nature Machine Intelligence. The innovation could pave the way for safer, more reliable automation across industries ranging from manufacturing to healthcare.

In the study, researchers from Surrey’s School of Computer Science and Electronic Engineering demonstrated how their innovative approach allows robots to predict when an object might slip – and adapt their movements in real time to prevent it. Similar to how humans naturally adjust their motions, this bio-inspired method outperforms traditional grip-force strategies by allowing robots to move more intelligently and maintain a secure hold without simply squeezing harder.

“If you imagine carrying a plate that starts to slip, most people don’t simply squeeze harder – they instinctively adjust their hand’s motion by slowing down, tilting, or repositioning to stop it from falling,” says Dr. Amir Ghalamzan, associate professor in Robotics and lead author of the study from the University of Surrey. “Traditionally, robots have been trained to rely solely on grip force, which can be ineffective or even damaging to delicate items.

“We’ve taught our robots to take a more human-like approach, sensing when something might slip and automatically adjusting their movements to keep objects secure. This could be a game changer for future automation, from handling surgical tools in healthcare and assembling delicate parts in manufacturing to sorting awkward packages in logistics or assisting people in their homes.”

Working in collaboration with the University of Lincoln, Arizona State University, Korea Advanced Institute of Science and Technology (KAIST), and Toshiba Europe’s Cambridge Research Laboratory, the study is the first to demonstrate and quantify the effectiveness of trajectory modulation for slip prevention in humans and robots.

The findings show a predictive control system powered by a learned ‘tactile forward model’ allows robots to anticipate when an object is likely to slip, continuously analyzing its planned movements.

Researchers also demonstrated the system works on objects and movement paths it wasn’t trained on, highlighting its potential to generalize effectively to real-world settings.

“We believe our approach has notable potential in a variety of industrial and service robotic applications, and our work opens up new opportunities to bring robots into our daily life,” Dr. Ghalamzan adds. “We hope our findings will inspire future research in this area and further advance the field of robotics.”

University of Surrey
https://www.surrey.ac.uk