MELLITUS RESEARCH
Mellitus, which I initiated in 2007, laid the groundwork for my later Islet series by transforming the invisible rhythms of my body into tangible forms. At its core was a simple yet radical idea: intercept the radio transmissions of my early continuous glucose monitor (CGM) and translate each blood-sugar reading into the language of geometry. What began as a one-off proof of concept evolved, over months, into a fully realized parametric system capable of producing an entire collection of “islets”—organic sculptures shaped directly by three months’ worth of glucose data.
Notably, the data window coincided with my planning of a major conference, introducing distinct physiological contexts to the project. During the first phase—High Stress/Poor Diet—intensive logistical coordination and irregular meals drove marked glucose spikes. In the second phase, my readings settled toward a middle range as the conference logistics stabilized—Moderate Stress/Moderate Diet. Finally, in the third phase—Low Stress/Moderate Diet—post-conference decompression and more consistent nutrition yielded smoother, lower fluctuations. By aligning these three conditions with the standard hemoglobin A1C timeframe, each “islet” became not merely a record of metabolic history, but a sculptural chronicle of stress, diet, and embodied experience.
The technical pipeline began with hardware interception. I turned a 433 MHz RF receiver—built around an RFM01 module and a small whip antenna—to the same frequency used by my CGM transmitter. Each broadcast, sent every five minutes, was an unencrypted data packet. Feeding these raw bursts into a software-defined radio (SDR) front end, I employed a custom Python script to filter out noise, isolate valid frames via the sensor’s unique preamble, and parse the payload for precise glucose readings.
Rather than simply logging these readings in a linear text file, I devised a “circular file” on disk: a fixed-size binary buffer that continuously overwrote its oldest entries once filled. I maintained a rolling data window that mirrored the clinical timeframe by calibrating the buffer size to match the number of transmissions within a standard three-month hemoglobin A1C cycle. This ensured that every sculpture in the series embodied a coherent span of my metabolic history, whether under the strain of conference deadlines or during the restorative lull that followed.
With the live data buffer in place, I turned to generative design. A Processing sketch reads the circular file in real time, normalizing each glucose value against the three-month average. These normalized values then drove key parameters—radius, extrusion depth, vertex count—in two- and three-dimensional parametric forms. Low readings pulled the form inward; spikes in glucose pushed surfaces outward, creating undulating, island-like shapes that visibly “breathe” with my body’s fluctuations.
Initially, this pipeline animated a single object on-screen and in rapid-prototyped wax. Over subsequent months, I expanded the system to spawn a new form for each reading and aggregated these into a series of pendant-sized sculptures. The result is a visceral portrait of my metabolic life: a collection of “islets,” each shaped by the ebb and flow of blood-sugar data, stress levels, and dietary rhythms—together mapping the intimate, temporal contours of my lived experience.