Today’s cutting-edge robots are very intelligent, and technologies like computer vision and deep learning make them very adaptable. A fulfillment warehouse pick-and-place robot in the past, for example, would have been explicitly programmed with the exact locations of items. Now, such a robot can understand what it’s seeing and find those items itself. But, it still needs to be able to pick them up, which is a task that MIT’s new soft robot gripper excels at.
This new soft robot gripper was created by Daniela Rus, director of MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL), and her team. It was designed specifically to address problems associated with traditional rigid robot grippers, which tend to have difficulty picking up delicate or irregularly-shaped objects. This new “Magic-Ball” gripper uses vacuum to fold like origami, and can grasp those same objects with ease.
The key to the design is the cone-shaped folding structure. That easily fits around items with a wide range of shapes and sizes, and then contracts and firmly grips them when vacuum is applied. The vacuum isn’t applied directly to the item, as with many pick-and-place grippers, but rather within the airtight body of the gripper. That causes the origami structure to fold in on itself, and around any object that fits within the gripper.
The design can lift several pounds, is affordable to build, and doesn’t require any advanced materials or fabrication techniques. The internal origami structure can be 3D-printed or made with traditional molds, and the “skin” can be anything airtight and flexible. That means that this gripper would be very practical to implement in current warehouse settings, or anywhere a versatile robot gripper is needed.