China3D printingNet August 4th, researchers from the Delft University of Technology (TU Delft) in the Netherlands have created multicolor3D printingSensors make soft robots self-aware and adaptable.
As stated by Rob Scharff, the first author of the study, soft robots can bend, stretch and twist at the same time, making existing sensors inappropriate.Through flexible embedded3D printingThe sensing method can detect various deformations and increase the interaction between the robot and the object. “These are important features of robots that interact closely with humans or fragile objects, such as nursing robots or tongs that handle fruits and vegetables of various sizes.” Scharff commented. “Because the interactions happening here are very unpredictable, the robot needs to be able to perceive the position of its own fingers and body.”
3D printingSoft robot
The soft robot is made of highly flexible materials and can perform natural movements similar to living creatures.Currently, NASA is using3D printingSilicone mold actuators to build soft robots designed for space exploration. Other applications have also been studied in medicine and zoology.
TU Delft researchers implemented Stratasys PolyJet 3D printingAnd Agilus Black, VeroCyan and VeroMagenta materials to create a curved actuator that holds air with an expandable bellows shape on the top. According to Scharff, “inflating the air chamber causes the bellows at the top to expand while maintaining the same length at the bottom, creating a bending motion.”
In order to increase the “sensing” capability, Scharff added, “We print color patterns in these top bellows, and observe these color patterns with a color sensor on the inextensible bottom of the actuator.
When the actuator is inflated, the color that appeared from the sensor will gradually disappear. We use this color change and intensity change to predict the shape of the performer. “
Actuator and color sensor. Picture from TU Delft.
Sturdy gripper
The sensor uses a feedforward neural network (FNN) to operate. The 1000 sensor value samples collected in the FNN correspond to the actuator shape. The shape of the actuator is represented by six marks on the non-stretchable layer, which are tracked by the camera. The input of the network is the reading of 4 color sensors, with 4 channels (red, green, blue, white).
“Our method is able to predict the position of each marker, the error
Usually between 0.025 and 0.075mm. “Scharff added.
“Unlike existing sensors in soft robots, we are able to measure the exact shape of the fixture bending around the object. Therefore, this work is a big step towards being able to accurately move and grasp objects with soft robots.”
3D printing
Actuators and color sensors” alt=”
3D printing
Actuators and color sensors” />
3D printingActuators and color sensors. Picture from TU Delft.
China3D printingOriginal article on the net, please give back to the link for reprinting!
(Editor in charge: admin)
0 Comments for “TU DELFT engineers develop 3D printed soft robots with self-awareness”