Researchers develop New Perception System for Soft Robotics

A team of researchers from Bioinspired Robotics and Design Lab at UC San Diego have created a new perception system exclusively for soft robots. Theycame up with an ideafrom the way human beings process information about their own bodies in space and in relation to other objects and people.

The main aim of the researchers is to develop a system that can predict a robot’s internal state and movements without the need of any sensors. The researchers have stated in their Science Robotics paper that they have successfully achieved the goal for a soft robotic finger.

The applications of the system can be used in wearable robotics and carry out human-robotic interactions. Apart from all that, the system can be used as soft devices to fix disorders affecting bones and muscles.

The system mimics different environmental navigation components for human beings: for vision, there is a motion capture system; for brain functions, touch sensation and finger movements there are neural networks. The neural network is trained by the motion capture system and after training, the motion capture system is discarded.

The researchers inserted soft strain sensors randomly within the soft robotic finger. After that it allowed them to predict to predict forces applied to, and movements of, the finger.

The approach will pave the way for the researchers to create models that can predict deformations and forces experienced by soft robotic systems while moving.