Home > News > Techscience

Even Robots Can Fall in Love with Marshmallows Using This Interface

ChenHuanHuan Tue, Mar 26 2024 11:25 AM EST

Can a rigid robot experience the softness of pillows, plush toys, and marshmallows like humans do? Actually making robots understand and replicate the sensation of softness is incredibly challenging, involving numerous complex sensory and cognitive processes. Swiss scientists have developed an interface for robots to perceive softness, enabling them to feel the softness of real materials. The related findings were published on March 11th in the Proceedings of the National Academy of Sciences of the United States of America. 65fd1222e4b03b5da6d0b9bc.png The Softness Rendering Interface (SORI) for robot tactile perception has been developed by researchers at the Swiss Federal Institute of Technology in Lausanne (EPFL), aiming to replicate human-like softness perception in robots. This interface distinguishes between tactile clues (sensory feedback from fingertip skin) and kinesthetic clues (force feedback from finger joints), addressing a previous limitation in robot touch sensors.

Mustafa Mete, a doctoral student at EPFL's Reconfigurable Robotics Lab, explains, "When you press a marshmallow with your fingertip, it's easy to tell it's soft. But if you put a hard biscuit on top of the marshmallow and press again, even though your fingertip touches a hard surface, humans can still perceive the softness underneath."

"We wanted to see if we could create a robot platform that could achieve this level of perception," says Mete.

The researchers developed SORI, which can discern between tactile and kinesthetic sensory clues, replicating the softness of various materials. Mete notes, "We realized that because everyone's fingertip shape is different, people may perceive softness differently. Therefore, in our research, we first had to develop geometric parameters for the fingertip and its contact area to estimate the softness of the fingertip." Then, the researchers extracted softness parameters from a range of materials and input them into the SORI device.

Experiments were conducted using a lab-developed origami robot as a research platform. This robot is equipped with motor-driven joints topped with silicone membranes that can expand or contract under airflow, adjusting softness.

By simultaneously perceiving tactile and kinesthetic cues, SORI successfully replicated the softness of various materials such as beef, salmon, and marshmallows in several experiments. It also mimicked some complex materials, like marshmallows covered with biscuits or books bound in leather. In a virtual experiment, SORI even replicated the sensation of a beating heart, demonstrating its ability to perceive softness during motion.

Mete summarizes, "This is not about serving as a softness sensor for robots but about digitally conveying the sense of 'touch'."

The lab believes this research "fills a gap in robot studies," enabling robots to utilize softness perception for applications such as deep-sea exploration, robot-assisted surgery, malignant tumor detection, and harvesting fruits and vegetables.

For more details, the related paper can be found at: https://doi.org/10.1073/pnas.231490112.