Aesthetic research on embryonic morphologies, part of MHOX design agenda. A generative application is programmed to produce virtual embryos composed by two main elements, an internal nucleus and an external membrane. The two elements estabilish a mutual dependence of generation and amplification based on seamless gradients in densities and opacities.
Each time the generative algorithm is run, its parameters change, and a new embryo is generated, with different distributions of densities and opacities.
The application is written in Processing, programming opensource framework based on Java.
The approach driving this graphic experimentation focuses on a few generative design key concepts.
GENERATIVE APPROACH. The designer is able to control a morphogenetic simulation by tweeking its parameters, which do not refer to visible features of the output but to more abstract constructs in the generative algorithm. This leads to a bottom-up generative process, in which the aim is orienting and exploring a simulation that allows to achieve unexpected results, instead of imposing top-down control on an object’s morphology.
GENOTYPE / PHENOTYPE. Variations in code parameters generate new, different graphic outputs, allowing the exploration of a space of embryonic configurations; it is possible to think of the generative algorithm as the genotype of a population of phenotypes, the individual embryos.
POPULATION BREEDING, FITNESS, SELECTION CRITERIA. Instead of optimizing a single solution, the aim is generating populations of possibilities; each embryo is intended as an individual, whose fitness has to be defined by the confrontation with selection’s criteria defined by the designer. Since each embryo relates differently with the selection’s criteria, it is possible to evaluate different balances and models of fitness and performance, not definable a priori.
The generative approach presented in this graphic experimentation shows opportunities for different areas of contemporary design; among others, the possibility of using a generative simulation to create differentiated graphic designs or objects on the basis of differentiated preferences or target performances models, keeping transversal identity elements, without having to reconfigure or reactivate the creative process each time.