Robotics are not in place. Engineers taught robots to understand human language, and walk on their own. But there is still a problem of tactile feedback, when a clever mechanism can’t only move the object, but also feel it. Scientists have created a new sensor that works according to the principle of human skin, differing fragile, soft, hard and other properties of objects.
Created by researchers from the Swiss higher Technical School of Zurich, the sensor consists of an elastic silicone skin with colored plastic microballs, and a conventional camera fixed on its underside. Thanks to machine vision, the robot “sees” what part of the skin enters into contact with the subject. The fisheye lens captures the size and depth of the dent on the silicone skin, calculating the required strength to hold the object, depending on its characteristics.
Silicone skin can distinguish between several types of exposure to the sensor surface, calculating the degree of effort required to hold it with high precision. The robot will also be able to determine the specific direction in which more or less effort is required. After the robot thus studied the properties of various objects, machine learning allows it to identify the pattern on the skin with its size, weight and density.
CURRENTLY, a team of researchers is working on new sensors that will teach robots to recognize more complex objects, and also consider the possibility of miniaturization.
Robots taught to “feel” objects by sight and by touch are great!