Tech's Touchy Feely Side
A novel dual-mode e-skin mimics human touch, offering simultaneous tactile sensing and reproduction, enabling new applications in robotics.
Electronic skin, or e-skin, is an emerging technology designed to mimic the functionality of human skin, enabling machines to perceive tactile sensations in a way that is similar to humans. It typically consists of flexible and stretchable electronic components integrated with sensors that detect various stimuli such as pressure and temperature. E-skin holds immense potential in a number of fields, ranging from healthcare to robotics, due to its ability to enhance human-machine interactions and provide valuable sensory feedback.
In human-robot interaction, e-skin plays a pivotal role in improving the communication and collaboration between humans and robots. By equipping robots with e-skin, they can better understand and respond to human touch, gestures, and environmental cues, facilitating more natural and intuitive interactions. For instance, in caregiving robots, e-skin enables them to perceive gentle touches or detect changes in pressure, allowing them to respond appropriately and ensure safe interaction with humans, particularly in delicate tasks such as assisting the elderly or individuals with disabilities.
However, a significant limitation of current e-skin technology is that most existing systems can either sense tactile information or reproduce it, but not do both simultaneously. This constraint hinders the integration of e-skin into robotics applications, as multiple systems are often required to achieve bidirectional tactile communication. Moreover, the additional bulkiness of a multi-system approach can compromise the aesthetics and practicality of the e-skin, detracting from the immersive experience and naturalness of human-robot interactions.
As technology continues to improve and shrink in size, new opportunities are emerging to change this present paradigm. A team led by researchers at Tsinghua University has seized on these opportunities to create a lightweight and inexpensive e-skin that can both sense tactile interactions and reproduce them. This dual-mode e-skin weighs in at 29 grams and costs under $26 to produce.
The device is composed of a soft and flexible silicon elastomer that is biocompatible and can be worn either over human skin or an artificial structure, like a robot’s end effector. A thin magnetic film and a Hall sensor array work in conjunction to detect deformations caused when the e-skin comes into contact with other objects. An array of tiny vibration motors are also integrated into the device to provide tactile feedback to the wearer. These sensors and actuators are connected to a microcontroller (ESP32-PICO-D4) to provide the necessary processing power for operation.
Since robots often run into issues with computer vision-based approaches to perception when fine operations are required, or due to an obstructed view, the researchers tested their sensing system as an alternative in several different scenarios. In one experiment, the robot was tasked with picking up objects, then classifying them based on the way that they “feel.” A convolutional neural network was trained for this purpose, and it was found to be capable of identifying the correct object, from a set of 12 options, in almost 99 percent of cases.
In another trial, the e-skin was attached to both a human arm and a robotic arm. The participants were then asked to control the robot, in real-time, using feedback received from the robot to perform some high-precision weighing operations. The sensing capabilities of the device worn by the humans were utilized to send commands to control the robot arm.
These experiments showed that the device can successfully sense and reproduce tactile sensations, and may have utility in a wide variety of applications. But to make the system more practical for those applications, the team plans to put some further work into miniaturizing the components. This would make it easier to integrate it into robotic systems and prosthetic devices in the future.
R&D, creativity, and building the next big thing you never knew you wanted are my specialties.