A statistical view of the design process

摘要

The application of task-based models of the design process to improving industrial process planning practice is becoming more common (1, 2). Many of these models track progress by measuring aspects of design performance as they approach and meet the design requirements. This means that the value of tasks such as testing, which do not directly alter the underlying design is uncertain in these models, despite their obvious importance. In this paper, a simple statistical model of a generic design process is presented, which embodies these test or simulation tasks as well as those involving the direct modification of the design. This model is then used to show how indirect design activities such as simulation and testing can be given a quantitative basis for inclusion in a task-based model.