8. Soft robotics¶

Research¶

In our lecture this week we learned about the myriad of ways soft robots are being designed to perform tasks that are not suited for "hard robots". This was a concept I hadn't ever considered. What's interesting about "soft" actuators is how they can be used in ways that reduce the potential for physical harm or damage when working with humans, animals or in sensitive environments.

There are three principals of design that we focused on...

Soft Actuators¶

In very simple terms, this video by RealPars talks about the purpose of an actuator in general, which is to trigger movement. The movement can be linear or rotary. The video also explains that there are typically three different sources of energy for this movement: pneumatic, electric, or hydraulic.

Video: Link to YouTube Video ¶

Sensors and Grippers¶

This article, titled New 3D-Printed Soft Robotic Gripper Functions Without Electronics for Tech Explorist, by Ashwini Sakharkar, talks about robots that can grip and hold objects using only pneumatic channels and no electronics. I also love how this kind of gripper might be used to perform tasks in environments like water or without the need for complex connections and programming.

New 3D-Printed Soft Robotic Gripper Functions Without Electronics¶

Source: Tech Explorist

Artificial Muscles¶

I found it interesting that "muscles," as we know them in animals and humans, are being incorporated into the design of soft robots too.

There are 3 types of muscle contraction:

Isometric: The muscle fibres stay the same length. They do not elongate or shorten.
Concentric: Muscle fibres contract and shorten to produce movement.
Eccentric: Muscle fibres stretch and lengthen.

Pairing these principles with soft robots, researchers are now making soft robots that can hold up to 1000 times their own weight, as seen in the video on this site by the Wyss Institute for Biologically Inspired Engineering.

FOAMs: Soft Robotic Artificial Muscles¶

Source: Wyss Institute for Biologically Inspired Engineering

Inspiration¶

1. Studio Drift¶

Studio Drift has some incredible soft robotic, biomimetic, high immersive installations. In some cases, elements like the viewer's own heartbeat are captured to influence the movement of the art. I'm particularly fond of thier project "Shy Societey" in which soft robotics and lighting design are blended seamlessly into mesmerizing, constantly moving organic shapes.

2. AeroMorph¶

“Could this futuristic inflatable fabric change industrial design?” – Curbed.
For my actual project this week I was inspired by this article by Asad Syrkett. In it Syrkett goes into detail about AeroMorph, a material being developed by MIT Media Lab. The article contains a video showing the work of researchers who have been experimenting with origami folding methods for inflatables. This technique results in materials that have the potential to fold by themselves. They can be used as things like grippers and self folding packages.

After reading this I was inspired to explore the kinetics of different folding techniques. So first I would need to do some fast prototyping.

Process and workflow¶

I started my work with vinyl, using the Roland CAMM-1 GS-24 vinyl cutter.

To get me started I watched this tutorial by Stahls' TV on how the vinyl cutter works.

How to Cut Vinyl Start to Finish with the Roland® GS-24 Vinyl Cutter¶

Watch the video: How to Cut Vinyl Start to Finish with the Roland® GS-24 Vinyl Cutter

I used the baking sheet/ vinyl method explained by Adriana Cabrera in our tutorial. In the tutorial the instructions are to cut two pieces of your vinyl or what ever material you choose, that can be fused by heat. To make the air channels, baking paper is inserted in between your two pieces to stop the material from fusing together.

Below you can see my steps:

1.I used AutoCad to come up with a number of designs. I saved my files as a .dfx and popped them on a USB so I could take them to the lab computer.

2.First I had to secure the baking paper to the cricut sheet. Baking paper is slippery so this prevents it from getting dragged by the printer or shifting. After a number of failed attempts with spray glue I finally switched to stick glue and this worked way better.

3.I loaded my sheet and selected how the print should read it. In this case I wanted it to find the edge of the material.

4.When I was happy I set the origin of the cutter, and then I moved to the computer where I opened my file in Inkscape.

5.From Inkscape I sent it to the Roland Studio cutting software. I made the force of my cut 50 and the speed to 5 (both half of the default setting).

6.Once I was happy with the position of my file on the template bed, and all the lines looked correct, I sent it to the cutter.

After a number of failed attempts with the paper shifting because of the wrong glue, this was finally the result:

7.Next I repeated the same step, only this time using a vinyl sheet instead of the cricut paper holder. The process was the same except I told the printer to read the "piece" of vinyl instead of the "edge".

Note: I originally peeled off my vinyl shapes in my first test, but then Annie showed me that I should leave the vinyl where it is and peel around. We need the clear plastic layer to remain on the shape when heat pressing. But that part can come off after pressing.

I also found the cutter didn't always read my lines. I don't know why this was happening, so I upped the power setting to 10 on the vinyl cutter. This didn't work, so the last couple of lines that it missed I manually cut myself with a ruler and an X-acto blade.

8.With my pieces cut, I headed to the heat press. We heated it up to around 155°C. When it was hot enough, I pressed all my pieces together for about 30 seconds.

9.Once my piece was cut and pressed, I peeled off the clear plastic. Then I very carefully picked the spot to create a small hole where the air valve would go. This needs to be in a region where the wax paper sits underneath, and only on one side. Otherwise, you are just poking a hole in the balloon and it won't inflate.

In some instances I included this tiny hole in my CAD File. In others, I did it myself. I think I'll stick with doing it in CAD, even though it takes a bit of extra time. Doing it yourself can get finicky and it's easy to slip and make extra holes.

10.I placed my valves in the balloon and inflated it with a small motor.

Results¶

The first prototype was the closest thing I came to a soft "gripper" robot. It had the ability to grasp things very lightly. Here is a shchematic of the air flow:

I also found that with most of the balloons, I could manually manipulate them to decide which way I wanted them to fold.

After studying the results, I played with the crease lines in my next prototypes to see how that affected the fold. When air bubbles bump up against air bubbles too soon it makes it harder for the shape to fully fold so I played around with more and less distance between the folds.

Below you will find all the tests I did using the same step process as above: