Production industries for the engineering of cars, ships, and aircraft are geared to minimize tooling costs by creating a range of standard models from mass-produced custom components. On the other hand, construction industries for the architecture of buildings aim to create one-off custom design, but with an economy based on the use of standardized components. Of course, this is a simplistic historical view. However, it aims to highlight the different approaches of the two industry sectors. Both achieve a variety of products while exploiting standardization in different ways to achieve efficiency. The advent of digital fabrication techniques has made possible the concept of ”mass customization”, which is blurring this distinction and thereby allowing industries to learn from each other and also to borrow technologies. But the core technology for the shift resides in software engineering.

The success of a piece of software is about the match or mismatch of assumptions between the software designer and the users. We can say that we all learnt from the assumptions made by the software developers of these other parametric systems for other industries. We learnt about what was transferable to architecture and we learnt what additional functionality would be required if the transition of parametric design to architecture was to be successful.

There are two important characteristics of parametric design applied to aircraft or ship design that are not present in terrestrial architecture. The first is that concepts and configurations change relatively slowly. Secondly, a single design, with some minor variations, will be used for a production run of ten, hundreds, or possibly thousands of instances. Therefore, there is the time and sources to invest in the proper “genotype” and insure that this can support the anticipated variations in the phenotypes. Contrast this with buildings where, in the main, each one is unique. There is only one instance so there is no need for a genotype that can support variations in the phenotype. But there are three exceptions to this statement. First, with buildings which are distinctly “rule-based”, it may be advantageous to develop a strong genotype the characteristics of which can be refined and shared with successive variants. Second, buildings which are contain “variation” within a single configuration. Third, all design can benefit from refinement.