Project Snowflake is an algorithmically generated jewelry collection featuring precious pendants that are as unique as nature’s own crystalline yet fleeting masterpieces. The project was conceived as an experimental pitch to McKenzie Liautaud, an award-winning Haitian-American accessories designer. The NYC based jeweler has adorned the likes of Taylor Swift, Cynthia Erivo, Lil Nas X and more with his elevated diamond, crystal and gemstone pieces in unique settings.
Ideation
Initially intended as a seasonal line, the first iteration of Project Snowflake took inspiration from the concept of freezing a magical moment in time. RDL specializes in translating data into highly personalized art and design. We proposed allowing customers to select a special date as the only input into this bespoke algorithm to translate that date into the atmospheric variables responsible for the creation of real snowflakes.
The results were impossibly complex geometrical shapes. The algorithm succeeded in producing combinatorial diversity equivalent to over 4 billion-to-1 odds against a repetition. This meant that McKenzie would be able to offer his customers the truly innovative experience of designing and wearing jewelry as unique as a snowflake.
Experimentation
During the experimentation phase we tested multiple line drawing software libraries before settling on SVGWrite, a powerful and flexible python module.
It took several iterations of code over multiple weeks to discover generated geometry that was clean and precise enough to transform into three dimensional shapes. After that we developed a pipeline to transform the generated SVG files into extruded 3D objects using Blender. Object files were then exported from Blender and finalized in 3D printer software for fabrication.
Prototyping
The prototyping phase was highly collaborative as our client McKenzie Liautaud had existing relationships with jewelry fabricators in NYC. We tested multiple processes on some of the most complex geometry we could produce to benchmark the printers. The idea was to understand the minimum feature size of each printer. Through this phase we realized the limitations of some printers on intricate designes and adjusted the snowflake generation algorithm accordingly.
Through several rounds of prototyping we were able to calibrate on the geometries that when printed at various sizes, ultimately met McKenzie’s high standard of quality for the precious metals he intended to use.