And the use of cumbersome multiple motion-capture cameras to provide precise information about the robot’s 3D movement and positions somehow defeats the purpose of designing a soft robot. Now, researchers from MIT have leveraged Artificial Intelligence (AI) algorithms to analyse the output of flexible kirigami-shaped sensors integrated to the skin of a soft robot trunk, and enable proprioception, the ability for the soft robot to “feel” how it is twisted or bent and understand its own position in space.
Described in a paper published in the journal IEEE Robotics and Automation Letters, the skin sensors consist of sheets of conductive materials used for electromagnetic interference shielding, which the researchers hollowed out or cut into precise kirigami patterns that make the sheets much more flexible and stretchable.
Because of their piezoresistive properties (varying their electrical resistance when strained), these materials turn out to make effective soft sensors as they deform in response to the trunk’s stretching and compressing. Electrical resistance of the sensors is converted to a specific output voltage, which is then fed into a novel deep-learning model that sifts through the noise and captures clear signals to estimate the robot’s 3D configuration, correlated to real movement data captured with a motion-capture system for ground truth data.
The researchers validated their system on a soft robotic arm resembling an elephant trunk, that can predict its own position as it autonomously swings around and extends.
“We’re sensorizing soft robots to get feedback for control from sensors, not vision systems, using a very easy, rapid method for fabrication,” explains Ryan Truby, a postdoc in the MIT Computer Science and Artificial Laboratory (CSAIL) who is co-first author on the paper along with CSAIL postdoc Cosimo Della Santina.
“We want to use these soft robotic trunks, for instance, to orient and control themselves automatically, to pick things up and interact with the world. This is a first step toward that type of more sophisticated automated control.”
One future aim is to help make artificial limbs that can more dexterously handle and manipulate objects in the environment.
“Think of your own body: You can close your eyes and reconstruct the world based on feedback from your skin,” says co-author Daniela Rus, director of CSAIL and the Andrew and Erna Viterbi Professor of Electrical Engineering and Computer Science. “We want to design those same capabilities for soft robots.”
The researchers’ robotic trunk comprises three segments, each with four fluidic actuators used to move the arm. They fused one sensor over each segment, with each sensor covering and gathering data from one embedded actuator in the soft robot.
To estimate the soft robot’s configuration using only the sensors, the researchers built a deep neural network to do most of the heavy lifting. They also developed a new model to kinematically describe the soft robot’s shape that vastly reduces the number of variables needed for their model to process.
In training, the model analyzed data from its sensors to predict a configuration, and compared its predictions to the ground truth data collected simultaneously by the motion-capture system. In doing so, the model “learned” to map signal patterns from its sensors to real-world configurations, matching the robot’s true position.