In this three-part exercise, students developed expertise in digital and physical modeling skills enabled by a given set of fabrication machines made available by Autodesk's Pier 9, becoming experts in the relationship between digital representations of form, the capabilities of fabrication equipment, and the constraints of material systems. As a way of benchmarking the aquisition of these skills, and to emphasize the value of craft as a design generator, these experiments will take place in the context of a single canonical test model for fabrication: the Berkeley Pipe.

In computer graphics, a canonical test model (sometimes termed a 'digital primitive') is a three-dimensional model, typically a mesh, which is used as a standard for benchmarking rendering algorithms, demonstrating visual effects, and in comparing results among implementations of a variety of other computational processes. Some of the most widely used canonical objects in computer graphics were produced by scanning physical objects, while others were modeled directly in the computer at times employing procedural rather than explicit modeling methods. Each is concerned with exhibiting a set of specific properties relevant to a computation process for which it serves as a benchmark, and may be associated with a particular set of concerns. The Stanford Bunny, for example, consisting of nearly 70,000 triangles (a simple model by today's standards), was designed to benchmark algorithms for polygonal simplification, compression, and surface smoothing. Other canonical test models have been intentionally designed with flaws, as to provide realistic input for any algorithm that is benchmarked.

This project seeks to transfer the idea of the canonical test model from the domain of computer graphics to that of architectural digital fabrication. The form of the "Berkeley Pipe" is proposed as an initial digital primitive that exhibits useful properties for the benchmarking of digital fabrication techniques.