Salustri_Filippo_A.pdf (626.33 kB)
Download file

Designing as Balance-Seeking Instead of Problem-Solving

Download (626.33 kB)
journal contribution
posted on 21.05.2021, 17:37 by Filippo A. Salustri, Damian A. S. Rogers, Nathan L. Eng
Designing is often related strongly to problem-solving (PS). We will argue that the practise of problem-solving as it is typically understood, has certain characteristics at odds with how designing works. We will suggest an alternative view of designing that could improve its effectiveness. A review of the PS literature reveals various interpretations. We will assume a typical lay-person's conception. In design, PS generally motivates the act of designing: the design will solve a problem. We will show how characteristics of PS in this sense conflict with our understanding of designing. Some of these characteristics are: problems are fixed, and their solutions are permanent; solution methods are insensitive to time delay between problem specification and solution implementation; and the separation of solution from solver. To address the identified issues, we propose that designing be consider an act of balancing a situation; this relates both to situated cognition and Alexander's notion of harmony. We have found consistencies between designing as balancing (DaB) and control theory (which describes and predicts how natural and artificial systems respond to environmental changes), the reactive steady states of ecological systems, and Alexander's pattern languages. We believe the sympathy between all these areas indicate an important and beneficial underlying unity. Finally we consider how DaB will be different from PS-based designing, including: * context and requirements that include measures of imbalance; * design parameters based on intervals rather than target values; * prototyping, modelling, and testing move upstream; * increased analysis of possible post-implementation futures during upstream activities; and * changes in concept evaluation methods. Since a balance-based design process does not exist as far as we know, it is not possible to assess its benefits. However, we will show that benefits are possible.



Ryerson University