Gecko-inspired farmHand combines grasping strength with a light touch
When it comes to robotic hands, there are ones that can grasp tough items firmly, and ones that can grasp delicate items lightly. The experimental new gecko-inspired farmHand, however, is capable of doing both.
Created with the handling of a wide range of objects in mind, the device was developed at Stanford University by engineers Wilson Ruotolo, Dane Brouwer and Mark Cutkosky.
It has four articulated fingers, which are able to close around objects via the tensioning of internal tendons. The elastomer pad of each finger segment is modeled after the foot pads of the gecko, allowing it to securely grip the surface of a fragile item – like an heirloom tomato, for instance – without applying much force.
Gecko's feet are able to stick to surfaces thanks to millions of microscopic hair-like projections called setae. These temporarily bond with surfaces at a molecular level, due to what are known as Van der Waals forces. When the reptiles pull their feet forward, the bond is broken.
Instead of setae, the farmHand finger pads incorporate similarly performing microscopic flaps. Like setae, those flaps not only molecularly bond with surfaces, but they also leave no adhesive residue behind once that bond is broken.
In order to help the gecko-inspired pads conform to curved surfaces, they also incorporate a collapsible rib-like structure that buckles with even a small amount of force. This arrangement ensures that as the pads come into contact with a non-flat surface, an equal amount of force is distributed throughout the contact area – this both minimizes the chances of damage to the item, and reduces the likelihood of it slipping out of the hand's grip.
In tests conducted so far, the farmHand has safely grasped delicate items such as grapes and raw eggs, yet has also securely grasped and lifted heavier, more robust objects like a basketball and an angle grinder. Plans now call for a force feedback system to be added.
The farmHand is described in a paper that was recently published in the journal Science Robotics. You can see it in action, in the following video.
Source: Stanford University