Integrating Soft Robotics with ROS - A hybrid pick and place arm

TL;DR

This paper presents an open-source ROS-based framework for integrating soft robotic components, demonstrated through a hybrid pick-and-place system to facilitate adoption in research and industry.

Contribution

The authors develop and validate a modular, soft-hard hybrid robotic system framework built on ROS, simplifying integration of soft robotics into existing platforms.

Findings

Successfully demonstrated pick-and-place tasks with the hybrid system

Framework reduces development time for soft robotic integration

Open-source release encourages wider adoption and experimentation

Abstract

Soft robotic systems present a variety of new opportunities for solving complex problems. The use of soft robotic grippers, for example, can simplify the complexity in tasks such as the of grasping irregular and delicate objects. Adoption of soft robotics by academia and industry, however, has been slow and this is, in-part, due to the amount of hardware and software that must be developed from scratch for each use of soft system components. In this paper we detail the design, fabrication and validation of an open-source framework that we designed to lower the barrier to entry for integrating soft robotic subsystems. This framework is built on ROS (the Robot Operating System) and we use it to demonstrate a modular, soft-hard hybrid system which is capable of completing pick and place tasks. By lowering this barrier to entry we hope that system designers and researchers will find it easy to integrate soft components into their existing ROS-enabled robotic systems.