A Human-Like Triple-Layer Sense of Touch Lets This Trash-Sorting Robot Feel Its Way to Success
Using three different touch-based senses, this robot can quickly determine the type of trash it's grasping in its hand.
Researchers from Tsinghua University are aiming to boost recycling rates by building a trash-sorting robot with a human-like sense of touch — allowing it to feel its way through mixed waste.
"We propose utilizing spatiotemporal tactile sensing during hand grasping to extend the robotic function and ability to simultaneously perceive multi-attributes of the grasped object," corresponding author Rong Zhu explains of the team's work, "including thermal conductivity, thermal diffusivity, surface roughness, contact pressure, and temperature."
The sensor developed by the team is based around a three-layer design, in which an upper layer can feel the texture of a surface, the lower layer tracks pressure, and the middle layer attempts to mimic the human ability to determine changes in temperature. Coupled with a cascade classification algorithm, which works through possible classifications in order from easiest to hardest in order to boost efficiency, that's enough to separate a range of materials.
To test the sensor, the team had the robot sort out ten different object types into four categories: recyclable waste, made up of paper, a plastic bag, a plastic bottle, a drink carton, and cloth; kitchen waste, made up of bread and an orange peel; hazardous waste, in the form of expired drugs; and other waste, comprised of a napkin and a sponge. The resulting trash-sorting robot design was able to deliver a classification accuracy of 98.85 percent, the team says — which could mean a big boost not only to recycling capabilities but to robot sensing in general.
"In addition," Zhu says of the team's plans for future extensions to the project, "by combining this sensor with brain-computer interface technology, tactile information collected by the sensor could be converted into neural signals acceptable to the human brain, re-empowering tactile perception capabilities for people with hand disabilities."
The team's work has been published in the journal Applied Physics Reviews under closed-access terms.