top of page


Rhinoceros and Grasshopper, Digital Culture and Fabrication


Designed in Fall 2018

​Focusing on creative workflow, implementation and fabrication, this project aims to create bench using voice algorithm-generated geometry.

From Poetry-generator to Waveform

When Bohemians became a revealing metaphor context, Baudelaire mixed women and death in a third imagery to the narrative the city of Paris. Poetry deconstructed language, which magnifies the metaphor and atmosphere infinitely.


In The Poetry Generator, I'm aiming to generate propinquity Baudelaire's poetry through a semi-quantitative prediction of the options of the random subjects and inheritable elements. I base on the features showed in The Flowers of Evil developed a quantitative framework to let the program transfer the influence of emotional factors in random poem generating. Through the loop inheritance of functions, the program constructs similar random poetry with unified emotions. However, the correspondence of imagery and metaphor is random, which means that the meaning is infinitely possible. Completing the illusion of our poets' dream by borrowing this process of exploring unpredictable vague fables.

As an extension of the previous Workflow Experiment (Through the Looking Water), I added a voice module to the previous poetry generator. Capture a series of waveforms from the poems, generate a 3D model from the waveforms, and cut the model into slices for digital fabrication. Laser-cut the slices to build a small prototype of the physical poem.

The following are the basic steps to approach:


First, classify the vocabularies into the category.

- Re-generate the category giving the generator multiple choice.

- Set up the random vocabularies function.


Second, set up the function of the form.

- Generate candidate function.

- Call viable function.

- Use random function select viable function.


Finally, generate Semi-quantitative Baudelaire's propinquity poetry.

Random Sonnet Diagram (Figure 1.):


Figure 1. The poem uses for audio recording.

By using grasshopper (Figure 2.), I was able to transfer the audio waves into physical waveforms. Furthermore, I have chosen the most satisfied pieces to create a 3D model (Figure 3.). 

Capture_1 copy.PNG

Figure 2. Grasshopper process.

formOwl_1 copy.PNG

Figure 3. The 3D model from the waveforms.


Figure 4. Transfer the 3D model into a 3D printable bench.

bottom of page