Towel cloth manufacturers explain that towels can also be creative like this

I believe that every single man hopes to have a robot that can help make beds and clothes. And recently a Berkeley research group has taken a solid step toward that goal: their autonomous robot can fold a towel. Suzhou terry cloth

"Robots have been able to assemble cars for decades, but folding towels is only very recently." Jeremy Maitin-Shepard, a Ph.D. student in Berkeley's Department of Electrical Engineering and Computer Science, leads the terry towel manufacturer's development team. Teaching assistant Pieter Abbeel said.

Most of the robots on the automobile assembly line work in a very stable structural environment, which can achieve work precision that is difficult for humans to achieve, but this equipment is only suitable for repetitive work in strict environments, and has limited capabilities in other environments.

While little things like packing up and making a bed may seem inconspicuous, the real challenge for robots is to perceive and manipulate "morphing objects." Towel cloth wholesale

"Our work started with some towel shapes that the robot could recognize, and the robot would fold the towels according to a standard algorithm," Maitin-Shepard said.

But the difficulty is that when the robot picks up one of the towels from a pile, it has no guarantee that the shape of the towel is recognizable by itself. It must first transform the shape of the towel into one that it can recognize. This is because existing image recognition techniques are more focused on rigid objects. The Berkeley researchers had to determine the actual shape of the towel based on the captured image of the towel and the text patterns on it. Solving this problem will go a long way toward enabling robots to fold towels, and of course it's a breakthrough in robotics.

The technical innovation of the team lies in the development of the key point recognition technology on the cloth surface on the basis of the existing computer vision, which is a very effective algorithm because it can rely on some reliable geometric clues to infer the deformation of the cloth surface. actual shape. The method has proven so reliable that the Berkeley team will present its study in May at the International Conference on Robotics and Automation in Anchorage.