Recently, researchers from Carnegie Mellon University in the U.S. have been working on a groundbreaking project involving a robot named “Baxter.†This robot is being trained to improve its ability to grasp objects effectively, using a combination of visual and tactile feedback through continuous experimentation and refinement. Their goal is to develop the next generation of industrial robots that can seamlessly integrate both visual and tactile processing capabilities.
In the realm of artificial intelligence, enabling robots to handle objects as skillfully as humans remains a significant challenge. One of the critical aspects of achieving this lies in the tactile feedback systems of robots. The Carnegie Mellon team has equipped Baxter with a unique gripper called Fingervision, which is essentially a 3D-printed gadget installed at the end of its arm. This gripper features a transparent silicone sleeve embedded with black dots. As the gripper interacts with objects, the deformation of these dots is captured by a tiny camera within the device. Based on this visual input, Fingervision processes the data and adjusts its grip accordingly.
Moreover, Baxter leverages advanced AI learning techniques. It sends the sensory data collected by Fingervision to a deep neural network modeled after the human brain. By comparing this data with images in ImageNet, the world’s largest image database, the robot enhances its recognition accuracy by up to 10% compared to robots relying solely on visual data.
Currently, Fingervision is capable of controlling the strength of its grip based on whether an object is slipping. This allows it to perform intricate tasks like peeling a banana. When encountering familiar objects, Fingervision applies a firm grip, but it cautiously withdraws its arm when faced with unknown items.
Other institutions are also making strides in tactile robotics. Cornell University, for instance, has developed a tactile manipulator that uses optical signals to perceive the shape and texture of objects. Its flexible, inflatable fingers contain optical materials that change their conductivity when deformed, allowing the robot to gather data about the objects it touches.
The University of Glasgow in the UK is exploring bionic skin technology. Made from graphene, this innovative material serves as a highly sensitive tactile sensor. Thanks to its rapid response to minute changes, the skin can provide precise feedback to robots. Additionally, the skin’s low power consumption—only 20 nanowatts per square centimeter—is offset by an integrated solar panel beneath the graphene layer, ensuring a sustainable energy supply.
If the advancements in tactile technology from Carnegie Mellon, Cornell, and Glasgow continue to mature, robots could soon excel in broader applications. As the creators of Fingervision envision, these developments may lead to robots working alongside humans in increasingly safe and efficient ways, transforming industries and daily life.
Led Tv Led Display,Led Tv Screens,Crystal Led Display,Lcd Led
ShenZhen Megagem Tech Co.,Ltd , https://www.megleddisplay.com