How a Soft Robot Arm Moves Using Air, Not Motors

Table of Links

Abstract and 1 Introduction

2 Related Works

2.1 Pneumatic Actuation

2.2 Pneumatic Arms

2.3 Control of Soft Robots

3 PAUL: Design and Manufacturing

3.1 Robot Design

3.2 Material Selection

3.3 Manufacturing

3.4 Actuation Bank

4 Data Acquisition and Open-Loop Control

4.1 Hardware Setup

4.2 Vision Capture System

4.3 Dataset Generation: Table-Based Models

4.4 Open-Loop Control

5 Results

5.1 Final PAUL version

5.2 Workspace Analysis

5.3 Performance of the Table-Based Models

5.4 Bending Experiments

5.5 Weight Carrying Experiments

6 Conclusions

Funding Information

A. Conducted Experiments and References

5 Results

5.1 Final PAUL version

Although the layout of the pneumatic bench allows working with up to 4 segments, it was thought that using 3 would allow the different problems linked to redundancy to be tackled without increasing the weight of the robot too much or requiring the tubes –which pass through the interior of the segments – to have an excessive amount of space.

It is true that the tubes of the other three could pass through the first module, nevertheless, it was thought that the stiffness they would introduce by being so compressed could make it difficult to bend the initial segment. Since it is also the segment that has to exert the most force, as it is the one that supports the weight of the other segments, the risk of punctures could be increased.

Therefore, a robot consisting of three identical modules was assembled, standing at a total height of 390 mm (with each segment measuring 100 mm, intersegment connections 20 mm each, and the vision trihedron rod 30 mm). Under these configurations, the estimated weight of PAUL’s arm is around 600 g. The structure protecting the manipulator is a cube with a side of 500 mm. Pressure of the pneumatic line was established in 1.2 bar.

Examples of PAUL reaching different positions are depicted in Figure 13.

5.2 Workspace Analysis

The analysis of the workspace has been carried out experimentally, based on the data taken to generate the dataset. Figure 14 shows the workspace of a segment.

As can be seen, this is a surface, as the segment has two degrees of freedom if the condition that at least one valve should remain deflated is imposed. The surface can be considered as the union of three surfaces intersecting at the central point, which corresponds to the configuration of all deflated bladders. The three surfaces are roughly spherical in shape. If the PCC model were completely valid for the robot, these would be perfect spheres, as the ends of a set of equal-length arcs of circumference with a common origin engrench a circle. Since this is not exactly the case, the generated surfaces only resemble the sphericity predicted by the constant curvature model.

The addition of a second segment already generates a 4-D workspace that is difficult to represent. The generation of this is a consequence of the fact that, from each point on the surface of the workspace of a segment, another similar surface is generated. The

union of all of these surfaces, which arise from the points on the surface of the first segment, results in the two-segment workspace. This is a volume in which, in addition, each point can be reached from two different orientations, thus leaving latent the four degrees of freedom that PAUL would have with only two modules.