Create a program that generates 3D objects. More specifically, your program should generate a family of 3D objects that are all parameterized in the same way, but that differ when their parameters are set to different values. The output from your program should be suitable, at least in principle, for prototyping in the real world.
Give consideration to the way in which your parametric object operates as a cultural artifact. How might your software attain special relevance by generating things that address a real human need or interest? Can it make a dataset tangible or wearable? Is it possible for a generated object to be critical or tactical? Is it a tool? Garment? Decoration? Can it be funny, surprising, or unexpected? If you are hunting for a concept, it may be helpful to think of things in the world around us that are mass-produced but that could be (or ought to be) personalized. Document your object so that it can be shared with relevant audiences.
Generative design is the activity of authoring a system of rules for automating design decisions. In the case of the parametric object, a form is produced with properties articulated by certain variables. Changing the values of these variables changes the form in response, and incorporating elements of contingency or randomness can produce unique objects in every iteration. Among other approaches, parametric forms can be assembled from different arrangements of modular components, lofted from mathematical curves and surfaces, or produced through the actions of simulated physics.
3D parametric objects can be rendered in physical materials with the help of a wide range of fabrication technologies. These include additive technologies like 3D printers, which accrete or accumulate material, and subtractive technologies such as mills, which remove material from a piece of stock. The earliest of these, the CNC (computer-numeric controlled) milling machine, was developed in the 1950s through a military-funded, Cold War initiative to create mathematically precise propellers for aircraft. The early 1960s saw the parallel development of CAD (computer-aided design) tools to simplify the process of specifying geometries for CNC manufacture, and it wasn't long before artists finagled access to these technologies. In 1965, around the same time as the first exhibition of computer-generated plotter art, pioneer Charles “Chuck” Csuri became the first artist to employ CNC tooling for expressive ends, in his abstract Numeric Milling sculptures.
The creation of generative forms does not necessarily require computer programming. Traditions of process-based, open-ended and rule-based conceptual art create meaning in the tension between algorithmic logics and the material or social contingencies of the real world. Highly systematic thinking underpins works like SCUMAK by Roxy Paine (a metasculptural machine that extrudes randomly-shaped plastic blobs); Nathalie Miebach's handmade sculptures, structured according to weather data; and Laura Devendorf's Being the Machine project, an “alternative 3D printer” in which instructions typically provided to fabrication machines are instead given to human makers.
Manufactured objects gain a role in our lives through their utility as furniture, housewares, tools, toys, and ornaments. At times, however, the “one size fits all” approach in industrial design means “no size fits any.” Parametric design counters this by enabling “mass customization”—offering personalization akin to handcrafted and homemade forms, but at a heretofore impossible scale. Eyeglasses, prosthetics, and other wearables, for example, can be customized using measurements or scans taken from an individual's body. Maps, histograms, and time series can be generated from data and rendered into tactile media; this “physical visualization” approach can be useful in widening the accessibility of data for the visually impaired, or in producing mementos or souvenirs that encode information from highly personal experiences. In situations where uniqueness is prized, algorithmic design techniques can guarantee that no two items are alike. Parametric design also holds the alluring promise of creating “optimal” forms for a given situation (e.g., solving for the greatest strength per unit of material). While this is important for reducing the material impact of design processes, we must also recognize that over-optimization risks producing unduly specific outcomes in a dynamic world.