Soft Robotics - Week 6 (02/28/2025)

Programmable Air - Part 1

This week we were introduced the the programatic side of pneumatics and played with Programmable-Air pumps created by a former ITP student ​Amitabh Shrivastava.

​For this introduction week, we were asked to experiment with inflating/deflating a balloon and programming a pumping air sequence.

I will preface - I did pop a balloon...

Troubleshooting

On Tuesday night, March 4th, I attempted to complete the labs for this week, and for two hours struggled to upload the factory firmware program onto the Arduino Nano.

At one point I thought I had fried the microcontroller. It wasn't until 30 minutes later a classmate reminded me that the processor had to be updated to the Old Bootloader version.
​Then I couldn't remember how to factory reset the pumps. And just as I was about to email Kari, I figured I would double check the slides and guess what?

All of this information was actually on the class slides. So I am documenting those details here to not forget again.

Inflating a Balloon

I started off with the timing control program in the example library to understand what the code does.

I also experimented with different needle widths attached to the vent valve.
​I preferred the purple thinner needle as it releases less air, leaving more observational time in the deflating process.

I then tinkered with the code a little by adding a "sucking" state to the sequence and see what would happen - my prediction was that after sucking for 2 seconds, the balloon will hold in the inflated state.
​
This was proven wrong as the pump was not programmed to close any valves, so it continued to suck air out of the balloon.

Added variation in timing as well as the duty cycle percentage.
It is not very noticeable in the video, but the balloon is inflates for 8 seconds, vents for 1 second, deflates at 90% suction for 2 seconds, vents for 1 second, deflates at 75% suction for 2 seconds, vents for 1 second, deflates at 50% suction, and vents for the remainder of the cycle.

​Observations:
A 25% difference in suction power is not very noticeable but the 40% difference is definitely observable. Additionally, the 50% suction is very similar to the venting with a skinny needle (purple). 

In this example, I experiment with alternating between inflation, venting, and deflation. Introducing some inflation midway through the deflation sequence to see if the "strength" of inflation is the same when there is existing air pressure in the balloon.
​
It seems that the inflation behavior is about the same, and that the air pressure in the balloon does not affect the pump adding more air.

It was also around this time when I popped the balloon. I realized that I had been stretching the base section over and over again, and the air was not getting into the rest of the balloon. So in this process of inflating the balloon further, I ended up overfilling the already thin latex at the base, and boom! It popped.

Inflating The Silicone Cast

The silicone cast that I created last week was meant to actuate through suction, however, I made the walls of the structure too thick, leading to little footprint reduction (it's original intent).

​Though the ability for it to deflate AND inflate makes it an interesting tool to observe air flow in physical form.