Project
After designing and building functional controller prototypes for visuals performance, an idea emerged: the buttons on a multi-button grid controller could be more than just a grid of small buttons. Combining a love of playing with words, exposure to inspiring electronic art pieces, and an interest in design and observation of human interactions with technology led to the concept of creating an ambitious interactive electronic art piece.
Challenges
Large scope, which required learning unfamiliar processes on the job and extended the timeline
A significant amount of manual labor was required to complete the project.
Role
I conceived the project, acted as the designer and fabricator for the entire project, researched construction and build methods, sourced materials from springs to LEDs, and led multiple soldering and LED building sessions with volunteers. I installed the Word Wall at various events, presented the project, spoke with people about how certain aspects were accomplished, learned necessary technical information, and received useful fabrication ideas.
Process
Word List: The main part of the project was the front panel, which required a list of 128 words that would evoke a Pacific Northwest theme. Initially, I used popular words from the internet, but they lacked descriptiveness and did not evoke moods. To make the words more useful, I created sentence structures and edited the list accordingly. I realized the list would be more like haiku or fridge magnet poetry, which was perfect for the project. The words were printed onto paper and cut out, and then matched up into poems and sentences to test if the combination made sense. This process led to further editing of the word list. I used Adobe Illustrator and Inkscape to create a detailed layout, including holes to mount the panel to the case and the buttons to the back of the panel.
Lettering: The characters were cut through the aluminum panel, which required a stencil font to hold pieces like the center of the letter "O". I found a font and edited it to have larger sections of metal within the letters for durability during use and transport. I also smoothed out some characters for consistency with the rest of the font.
Button Mechanism: Initially, the design focused on creating mechanical buttons using combinations of bolts, springs, and plastic sleeves that would contact part of a spring with a washer to make an electrical connection. However, this proved to be fiddly and unreliable over time. A better solution was found: using the same springs and bolts to connect the button assemblies to the back of the front panel, but tightening the lower connection points to create a hinge. This allowed the piece of plexiglass to press forward on the top while staying in place on the bottom when pressed. A standard lever switch was added behind the panel near the top to complete the button mechanism. Initial tuning of all buttons was required to set the height and firmness of the spring action, but it has been rock solid since then. The end result was a spring-mounted plexiglass panel behind each word.
LED Lighting: Early testing of an LED sitting behind the white plexiglass button material looked good. However, due to highly differing word sizes, a single standard LED or assembly couldn't be used. A lighting electronics friend had leftover surface mount LED strips and spools of surface mount LEDs in two different colors: yellow and orange. Several test pieces were soldered together to see how different combinations looked when held behind the plexiglass, and a general rule of two oranges surrounding a yellow was used.
Every word had at least one section of 3 LEDs, and longer words have 15 LEDs. Each 3 LED section, with the addition of 2 resistors, required 12 volts. Additional sections were soldered together to keep the voltage requirement the same for every word. There were not enough strips to complete the wall, so additional strips were hand-fabricated. These strips, as well as the switches that the plexiglass panel presses to engage a button press, were mounted on a wood panel located behind the button panel at the height that gives the best glow and coverage for the words.
Fabrication: The creation of this piece required a lot of manual labor. Fortunately, the first piece was completed by an industrial waterjet company in Seattle that typically cuts insulation for boat hulls.
Wiring: Typically, a piece of electronics with a grid of buttons would use a standardized connection method using bundles of wires aligned vertically and horizontally. However, since the words on the front panel were intentionally designed to not have a uniform horizontal alignment, there was no easy and clean way to connect the electrical switches and LED light boards. This led to a messy process, but it made more sense after enough connections were soldered. Although the end result isn't the cleanest, it does work.
Learnings & Future
Scope: The project was ambitious given the timeline. Originally intended to be completed for an Ignition Northwest event, it was worked on during and after the event. Soldering and assembly sessions were held in Seattle with the help of event attendees. However, the estimated time for manual labor was far off, and in future projects, more time should be budgeted for similar endeavors.
Tool Time and Shop Space: A friend's shop was used for a significant portion of the early fabrication, which was not planned. In the future, shop space and time should be included in funding estimates.
Transport: To facilitate transportation, the wall was designed to be split into two pieces. However, the vehicle Ian owned at the time was not suitable to transport the halves. In future projects, sizing for expected transport or budgeting for rental vehicles should be considered. Breaking the project up into smaller pieces for ease of moving and working on the pieces could also be an option.
Repeatability: The project relied heavily on manual labor, shop space, and tool time, which made estimating the cost and time for creating a custom version difficult. Ian has been designing a new version with fewer mechanical parts and build steps that will allow for quicker creation of custom variations, and may lead to a product based on this work.
Electronics: Initially, boutique electronics were used to decode button presses and control lighting. Unfortunately, a mishap caused damage to the initial electronics, and subsequent attempts required a different connection method, resulting in the need to redo all the wiring. To prevent similar issues in future projects, Ian plans to use standard and easily accessible electronics.
Wiring: The wiring was an afterthought, and while it worked as conceived, Ian did not anticipate the amount of wiring the project required. Using wire that tended to pull away from solder joints made the wiring unreliable. In the future, more time should be allocated on the front end to avoid dealing with unreliable wiring.
Methods: Laser cutting was considered too expensive, so waterjet cutting was used instead. While initially disappointed with the cut edges not being perfect due to melting, it ultimately worked out as the soft edges were safer for user interaction. Testing any unfamiliar processes on a small scale before implementation in larger projects should be considered.
Complexity: This was Ian's most complex project yet, with every step requiring learning new processes. The initial waterjet cut job was expensive enough that more time was spent reviewing the Illustrator file before sending it in, which affected the time budget.
Word Outliers: Some of the words used were longer than average, requiring more mounting points in the panel to prevent the button panel from flexing when pressed. The letter "I" was particularly difficult to mount as it was small enough to almost not fit either its button or LED lighting strip. It was also the only word to use only two mounting points, potentially making it more breakable or problematic over time. Ensuring that all buttons are close enough in size to use the same mounting, triggering, and lighting methods would be ideal.
Ian describing the design of the Word Wall in comparison to a Monome MIDI grid controller at an artist fundraiser brunch that he and the wall were invited to.
Money: Ian initially envisioned this project being brought to music and art events, with the cost being amortized over a few years. However, this has not come to fruition, resulting in an expensive electronic art piece for his apartment. Despite this, Ian is proud of what he has achieved, grateful for the support he received, and glad to have met new people. Starting with a smaller project may not have led to something of similar scope, making the experience worthwhile.
Outcome
The Word Wall was partially funded by Ignition Northwest and installed as a work-in-progress at the organization's signature event, Critical Massive. Additional funding was provided by a group of Seattle artists who put on a brunch series.
This interactive sculpture consists of a 6' x 6' aluminum front panel with 128 words cut out using an industrial waterjet cutter. Behind each word is a spring-mounted plastic panel that triggers a button when pressed. Behind each plastic panel is a strip of LEDs. A computer takes input from the buttons and gives output by lighting up a word. One of the main activities for users is to enter a sequence of words that the Word Wall will play back at random, resembling haiku or fridge poetry over time.
The piece was built in two halves, allowing the wall to be assembled with the halves side by side, which makes it accessible to children and facilitates transportation. The Word Wall has been popular with both kids and adults, who love playing with it and being captivated by its glowing words.
