A computer-aided Markovian model of the design process
De Leon Fajardo, Augusto
Master of Architecture
This thesis is based on the hypothesis that design problems are conflicting situations of social systems which, as open systems, exist at a steady-state. The process by which these systems reach steady-states is described by Markov chains. Two basic characteristics of this process can be identified: the openness of social systems allows for the cancellation of design problems, and also the need for these systems to exist at a steady-state produces ever changing situations, hence ever changing design problems. The structure of design problems at that state consists of an arrangement of its components in such a way that some of them are more or less important than others. This can be analyzed by the proposed model in order to generate a description of the structure and a strategical decision sequence to solve the problem. The first part of this thesis discusses the above concepts as well as formulates the model and develops an algorithm for the associated computer program RESTOP, whereas the second part demonstrates the validity of the model with two case studies. This validation has some consequences in the field of environmental design, such as in teaching models, studies of the processes of upsurgence and cancellation of design problems, studies of particular problems in both their aspects of change and solution, and studies of management of design activity. Finally, some current developments in Diagraph Theory will soon permit further improvement of this model.