The robot is capable of deciphering 117 different materials using its finger
Researchers from the University of Southern California have created a robot that is capable of touch, and can also feel the difference between various textures.
Jeremy Fishel, a recent doctoral student at University of Southern California's Viterbi School of Engineering, along with biomedical engineering Professor Gerald Loeb, have built the specialized robot that is capable of identifying 117 different materials through advanced touch technology.
The robot uses BioTac sensors, which imitate the human fingertip and use an algorithm that allow them to use human methods for deciding which material is which. In addition, the sensors can tell which direction forces are applied as well as the temperature of the objects they're touching.
The robot has what looks like a human finger, where the BioTac sensors are protected by a liquid filling and skin-like covering. It has a fingertip that looks just like a human fingertip, complete with a fingerprint that helps the robot to feel vibrations. These vibrations are felt thanks to a hydrophone inside the finger.
By using the same exploratory movements that humans use to determine a texture, the robot is capable of deciphering 117 different materials using its finger. The researchers found that the robot is right 95 percent of the time when confronted with different textures.
Fishel and Loeb hope to use the sensors in human prosthetics in order to make the limbs more life-like. The sensors could also be used for companies that study the feel of consumer products.
Fishel and Loeb are now selling their BioTac sensors through SynTouch LLC, which was founded by the University of California's Medical Device Development Facility. SynTouch LLC develops tactile sensors for systems that imitate the human hand.
Robotic hands have come a long way. Just last year, Virginia Tech researchers had just created robotic hands capable of performing human tasks like opening doors and typing.
Check out the robotic hand in the video below: