Stanford engineers have created a robot glove with sensors that allows robotic hands to feel physical sensation, thereby increasing their dexterity.
The sensors work well enough to let robot hands handle delicate objects. Chemical engineer Zhenen Bao and her team have demonstrated that wearing their glove, robot hands can hold berries and ping pong balls without squashing them.
Sensors in the glove’s fingertips are able to measure the intensity and direction of pressure, which is essential to achieving manual dexterity. The technology that automatically controls these sensors is yet to be perfected. Once that is done, a robot wearing the glove could have the dexterity needed to hold an egg without crushing it or letting it slip from it’s grasp.
“This technology puts us on a path to one day giving robots the sort of sensing capabilities found in human skin,” Bao said.
In humans, the outer layer of skin has sensors that detects pressure, heat and other stimuli. The palms and fingertips are particularly rich in these touch sensors. They work in conjunction with a sublayer of skin called the spinosum, a microscopic terrain of bumps and valleys.
This bumpiness is of high importance. When we touch an object with our finger, the outer layer of skin moves closer to the spinosum. A lighter touch is felt by sensors close to the hilltops. When pressure is more intense, it forces the outer layer of skin into the valleys of the spinosum, triggering more intense touch sensations.
The spinosum also helps reveal the direction of pressure or shear force. For example, a finger pressing north will send strong signals to the southern slopes of the spinosum’s hills. The ability to detect shear force is what allows humans to grip fragile objects like eggs gently but firmly.
In the fingertips on the glove, each sensor is made of three layers that work together. The top and bottom layer are electrically active. Researchers laid a grid of electrical lines on both surfaces, and turned them perpendicular to each other to create a dense array of sensing pixels. The bottom layer is also bumpy like a spinosum. The middle layer acts as an insulator. This is important because electrodes that are in close proximity without touching can store energy. As a robotic finger presses down, energy is released, while the hills and valleys of the bumpy bottom layer are able to provide a way to map the intensity and direction of pressure to specific points on the perpendicular grid.
According to Bao, with proper programming, a robotic hand wearing her glove could perform repetitive tasks like picking up eggs and placing them into cartons. It could also be used in more complex tasks like surgery, where precision is of the highest importance. However, for now Bao remains focused on developing an advanced version of the glove that automatically applies just the right amount of force to handle an object safely without prior programming.
“We can program a robotic hand to touch a raspberry without crushing it, but we’re a long way from being able to touch and detect that it is raspberry and enable the robot to pick it up,” she said.