Comparing the Rules in Digital Algorithmic Generation Methods for Architectural Designs

Abstract

The paper deals with digital generative algorithmic systems. Its aim is to investigate the extent to which the difference between rule-based algorithmic systems affects the definition of rules themselves. To achieve this aim, the paper put forward a theoretical framework of rule-based systems including the characteristics of rules, and the way they work. The framework was adopted as a basis for the comparison of four common rule-based algorithmic design approaches in architecture. They are Shape Grammars, L-systems, Cellular Automata, and Swarm Intelligence Algorithm. The results showed that the differences between the rule-based algorithmic systems are slightly reflected in their rules. The similarities are identified in many rule's properties and their way of work such as: using symbols in the representation of rules, using substitution rules in the generation process, using the same procedures included in rules, the possibility of rule's repetition and rule's termination. However, shape grammars showed some individuality, especially in the reference of rules to architectural precedents.