Designing the Post-Natural - Measuring Device - Construction Noise Realizer

IMG_0097.JPG
IMG_0109.jpg

At this moment, the existing version of my project “Construction Noise Realizer” unfortunately is not yet working. There is a fundamental incongruity in my code that I realized too late to fix for this week, however, I have a firm idea for pivoting the design to make it fully functional in the near future.

These tweaks won’t affect the core conceit of the piece, which is to make visible (and physical) the nearly omnipresent noise pollution caused by building construction in our contemporary urban environment. The device will be placed by a construction site during the workday and over the course of an hour, it will “record” the noise levels by spraying water over a tray of dry cement in a pattern representing the prevalence and volume of the different audio frequencies present. Then after a few hours, this cement will dry, forming a sculptural record of the noise at that location. Performing this measuring process in different locations will then allow for the “data” to be compared by showing these cement sculptures side by side.

This idea developed over the course of a series of iterations and began with thoughts about the detrimental effects of construction on the few plants that do still live in our city. I then pivoted slightly to look at noise pollution specifically, and that was the point at which I started experimenting with cement sculptures as a form of data. Initially, I imagined that water would simply drip into a bucket of cement in order to create abstract shapes, but after a series of tests, I realized that the outcome tended to just be more like cement blobs than any sort of meaningful shape. This lead instead to the idea of having the cement laid out in a tray and to the project’s current iteration.

In it’s current form, the electronic components of the piece are an electret microphone module, 12V water pump, and motorized slide pot controlled by an Arduino Nano and motor driver. There is also a relay controlling a light bulb which simply turns off and on to indicate when the pump is activated.

Screen Shot 2019-10-02 at 6.08.13 PM.png
Screen Shot 2019-09-26 at 1.41.39 PM.png

From early on in the process, I wanted the form of the device to echo the construction sites it would be measuring, and so I made use of bare construction materials like plywood and metal brackets to create that effect. The layout was also heavily inspired by the artist Tom Sachs and his commitment to precision and organization of design even while using humble, unfinished materials. In order to efficiently interface all the parts I also modeled and 3D printed a small bracket to sit on top of the slide pot.

in progress.

in progress.

The goal of the Arduino code is for the microcontroller to take in audio from the microphone and analyze its various frequencies using this FFT library. Then if any of the frequency levels spike above a set threshold the motorized pot will slide the water nozzle to the corresponding spot over the cement and spray a small amount of water in that spot. Over the course of the hour, the water will accumulate in specific areas creating a pattern unique to the sound profile of that site.

the audio levels coming in across eight values.

the audio levels coming in across eight values.

The position changing, motor activating and light turning on in sync.

The position changing, motor activating and light turning on in sync.

As I mentioned earlier, this project unfortunately still only exists as a whole in a conceptual form. I had the entire piece fabricated and had each separate function of the code working. However, it was when I attempted to combine them all into the final piece that it ceased to work. What I learned much too late is that the FFT library I am using works by manipulating the timing of the microcontroller on a very low level and doing so causes all analogRead() functions to stop working. In order to operate, the motorized slide pot I am using to move the water hose has to read its own position in order to work, and so my program crashes when I attempted to do both things.

That being said, everything is not totally grim. I already have ideas about replacing the slide pot with a servo that would hopefully serve a very similar function and do so without the need for analogRead(). At that point, it should all work, and I am very enthusiastic about seeing what sorts of sculptural data this device will create.