Thanks to a new layer touch sensor that can identify and measure distinct forces, the sensory gap between human and robotic touch could be narrowing.
A new device consisting of a robotic gripping arm and a fingertip-mounted sensor could perform complex tasks like holding fragile and delicate objects and even guiding thread through the eye of a needle.
The novel tactile sensor with several characteristics matching those of natural skin has been developed by a team comprised of researchers from the City University of Hong Kong (CityU)and the University of Hong Kong (HKU).
Whilst robot devices that replicate human grip and touch have come on leaps and bounds, current tactile sensors for robotic applications simply can't match the sensitivity of our skin. Sensing subtle changes in force and stimuli is a skill human skin possesses that isn't currently replicable in even the most advanced robotics.
The team's soft tactile sensor contains a multi-layer composition similar to that of human skin. On top of these layers is a sophisticated specially magnetized film, just half a millimeter thick, that deforms when an external force is applied to it. This deformation results in a change in magnetic fields that the sensor can measure to assess the magnitude of the applied force.
Their results are published in a paper featured in the latest edition of the journal Science Robotics¹.
Mimicking Human Skin Skills
The sensor has another skill that really sets it aside from other alternatives. It can separate forces and identify them as either force applied perpendicularly or as a force applied from the side that causes objects to slide across each other. These separate forces — the normal force and shear force respectively — can also be distinctly measured once determined.
Identifying shear force and distinguishing it from normal force is a vital part of the skin's sensory capability. Think about holding a wet glass on a hot summer's day. It is the ability to detect the shear force that alerts you to the glass slipping in your hand and allows you to adjust your grip and avoid an embarrassing incident.
The researchers believe that these qualities could help robotic sensitivity close the gap on human touch.
It is important to decouple the external force because each force component has its own influence on the object. And it is necessary to know the accurate value of each force component to analyze or control the stationary or moving state of the object.
Yan Youcan, the paper's first author, Ph.D. student at the Department of Biomedical Engineering (BME)
The team's layered sensor has another ability that brings its sensory capability closer to that of human skin. Something the team is describing as 'super-resolution.'
Super-resolution is the ability of the new sensor to accurately pinpoint the location at which stimuli are occurring, just like human skin does naturally. For the sensor, this is achieved with an advanced deep-learning algorithm.
We have developed an efficient tactile super-resolution algorithm using deep learning and achieved a 60-fold improvement of the localization accuracy for contact position, which is the best among super-resolution methods reported so far. To the best of our knowledge, this is the first tactile sensor that achieved self-decoupling and super-resolution abilities simultaneously.
Dr. Shen , BME Associate Professor and project co-leader
Mounting this impressive sensor upon the 'fingertip' of a robotic gripping arm allowed the team to use it to grasp fragile objects such as an egg. They found that their robot arm could do this even as another force was applied in an attempt to drag the egg away.
"The super-resolution of our sensor helps the robotic hand to adjust the contact position when it grasps an object," Shen continues. "And the robotic arm can adjust force magnitude based on the force decoupling ability of the tactile sensor."
Shen believes that in the future, these sensors could be combined in an array that could form a 'skin' used to coat an entire robot — an e-skin. The sensitivity of this e-skin could be altered by adjusting the primary layer's magnetic field strength, resulting in a covering with areas of high sensitivity and varying measurement ranges — again, just like human skin.
"This proposed sensor could be beneficial to various applications in the robotics field, such as adaptive grasping, dextrous manipulation, texture recognition, smart prosthetics, and human-robot interaction," concludes Shen. "The advancement of soft artificial tactile sensors with skin-comparable characteristics can make domestic robots become part of our daily life."
1. Yan. Y., Hu. Z., Yang. Z., et al, , 'Soft magnetic skin for super-resolution tactile sensing with force self-decoupling,' Science Robotics, [DOI: 10.1126/scirobotics.abc8801]