03

The feedback was mostly on how to proceed with this experiment and I was glad that they aligned with what I wanted to proceed with, such as making more iterations and user testing. I assume that it means my supervisor and peers find this interesting as well. Although for now I’ve only shown them the motion of the brachiograph and it didn’t function very well at the school for some reason. Which also reminds me to come early and set up when I need to present or share my tasks with others.

Although I had taken a lot of images and videos of myself making the drawing machine and the final drawings, many of them seemed redundant. For this week, while I do my user testing, I will be able to document the actual interaction and variety of outcomes. Also, the pulse sensor wasn't prperly calibrated, which is something I should start with.

Moving on to the iterations, this week I will focus on refining the code and movement of the Arduino, rather than experimenting with physical elements like using different mediums for drawing. Although I will move on to user testing afterwards, I would like to note that the goal here isn't about establishing functionality. I’m trying to document the interaction between objective data and subjective interpretation, later shaping my prototypes based on both feedback and reflections.

The chart I am using is based on a morphological chart, a tool commonly used in product design for morphological analysis. This method involves breaking down a system into its basic components and analysing the possible variations of each to consider different possibilities. I believe this approach aligns well with my process, as I'm applying an analytical framework while introducing subjective criteria based on how I perceive the results. The options will become more refined later on.

The first category of options is the type of input data. Initially, I tried to think of different options for normalising the data but soon realised I simply assign an angle and range as well, and so normalised all the data into a range between zero and one. For the angles, I divided them into two categories:

one for the starting position and another for the range of movement. By doing this, I no longer need to adjust the normalisation. I also considered calculating the variance (HRV) as an option for the input, but decided not to because of how long it takes to get the reading in real time.

I would also like to try out different motors. However, the motors that I have ordered will be arriving next week so I will make more iterations using them next week. Also, my EEG device is coming in next week so it would be nice to experiment with both, and also use different mediums as well when they arrive. As for the duration, the user just stopped when they felt that the outcome was 'finished'.

One less straightforward factor I included is the stability of the adhesive used to connect the servos to the sticks. This idea emerged by chance when I tried using tape instead of a glue gun. While the tape was sticky enough to hold, it introduced a lot of variance in the movements, meaning it was less stable and more flimsy and bouncy compared to the glue gun.

User: SH (BPM: 62~82)

" This was very cool. I like that this outcome is something only I could create, like a finger print. I would also be curious to see how it looks when it becomes more refined. "

This leads to an interesting paradox I noticed during user testing. When users can set their own parameters, their sense of agency increases, as confirmed by their feedback. However, they also expect to see a unique, individualised outcome in comparison to others based solely on their input (when the parameters for the settings are the same). I’m excited to explore more of this balance between agency and individuality / self. My plan for next week is using different options for the input and output, slowly evolving this overall idea of biofeedback to physical. At the moment, I’m experimenting with digital outcomes like p5js or processing, so I will be sharing that as well.

I am also categorising my concept into three main themes, mainly to organise my reading list for my dissertation.​​ I find myself circling around three central themes that seem to consistently guide my thinking. All three inform one another and create a framework that holds together my project. Defining them helps me clarify my research questions and stay focused as I move forward.

The first pillar—Serendipity in Generative Art—is one of my most consistent interests, spanning from last year. I keep coming across them from my experiments. I’ve been attracted by the idea that not all creative outcomes are intentional, and often, the most compelling works arise from unexpected accidents or errors. In generative art, machines produce these unplanned variations, and I also find myself wondering how much of the final result is the machine’s doing and how much is mine. I hope to examine the tension between control and randomness.

The second pillar is Electroencephalogram (EEG) as Involuntary Feedback. I’ve chosen to use EEG because of its capacity to capture involuntary activity, as well as the narrative of consciousness, tying back to machine ego and agency, which were my entry points to this entire topic. Also, there’s something very intriguing about the idea of using one’s own brain waves, an element of ourselves that reflects our physiological state. This lets me discuss the definition of self representation as well.

And the distinction between voluntary and involuntary feedback, which pushes the boundaries of traditional interactive systems that rely on deliberate inputs, often keeping clarity as the main consideration. EEG’s role as a form of data input also challenges ideas of authorship, making it the primary aspect of my research into how artists interact with generative systems. Which brings us to our third pillar.

The third pillar, Sense of Agency through Interactivity, is both an introduction to the narrative and objective of the research as well as establishing a connection between the preceding two pillars. Do we still feel a sense of authorship and self representation if the input is something I can’t control? I want to understand how this sense of agency, or lack thereof, affects both the creative process and the interpretation of the work.

This pillar is necessary because, ultimately, my research isn’t just about how machines and biofeedback generate art but about how humans relate to these systems and perceive their own role within them. While they may change as I move on with my project and research, I know all three will be a theme worth exploring in order to build up to the outcome.