On the role of computational support for designers in action
The aim of this paper is to provide a clearer understanding of how computational approaches support designers’ actions. To do so, the authors categorize the use of computational tools into four focus areas:
Solution generation
These technologies attempt to assist and automate designer’s actions. Within this category the authors identify:
Parametric modeling systems which represent geometric relationships organized in a hierarchical binary tree structure that is automatically updated and visualized on the screen when changes in the parameters values occur. Although it facilitates design reusability, parametric model has limitations in terms of the variation beyond its preconceived scope, prematurely reducing the range of options to explore.
Expert systems, which apply existing knowledge of design domain in similar situations emulating human decision making based on rules. The main limitation is that these systems are limited to already known solutions.
Case-based reasoning systems, which reuse and adapt previous solutions for similar new problems. The problem is defined as a set of constraints that are satisfied in order of importance. The main disadvantage is that these systems do not learn and have issues synthesizing a large number of constraints for complex problems.
Generative design systems in which the resulting model evolves from an initial state through iteration by operations that imply geometrical and topological transformations. Generative design can lead to apparently creative outcomes since every new combination of parameters brings the opportunity to look for the emergence of new properties or affordances from the resulting composition. Although it is an approximation to the co-evolving dialogue between problem and solution, these systems lack reformulation mechanisms.
Agent-based design systems which rely on the interaction of active autonomous entities to create shapes that satisfy material, fabrication and geometric constraints. These systems have the limitation that they can both produce unexpected remarkable results or fall into constraint circularity and non-sense loops.
Solution evaluation
These systems analyze solution candidates and provide feedback to support decision-making. Within this category, authors identify:
Performance evaluation systems which estimate how a product executes a given function under stress. This design framework focuses on what the product should achieve instead of how it should be built. The disadvantage of these systems is that it is difficult to execute them in early design stages because of ambiguity and incompleteness in the design definition.
Design rule checking systems, which evaluate the fulfillment of rules and provide feedback. However, the designer must interpret the reports and provide solutions. Determining how to make or suggest corrections in real time during the design process remains a challenge.
Constraint-based systems, which capture design knowledge in the form of constraints and requirements that must be satisfied by the design. If constraints are reached or broken immediate feedback is provided, supporting the co-evolution of the problem and the solution. However, constraint-based methods only detect when constraints are violated and do not provide clues about how to address the problem.
Solution selection
These systems reflect trade-offs, values and preferences through quantitative indicators complemented with relative weights to reflect issues of certainty of information and preferences of decision makers. Within this category, authors identify:
Utility methods, which attempt to formally describe the preferences of the decision maker. These methods are hard to implement during the early stages however, because they rely on explicit information that is not always available at these stages.
Multi-criteria decision methods, which establish metrics to evaluate how a design fulfills given objectives. These approaches aggregate data into results to measure performance, which may also be normalized to establish a level of comparison of different design attributes. However, because the comparison is based on quantitative values, it is difficult to apply such methods in early design stages.
Optimization methods, which enable designers to define and search through large spaces of designs.Standard optimization approaches however, require a high level design specifications before design decisions can be made, thus it is not suitable for early design phases.
Integration of the generation, evaluation and selection processes
Within this category, authors identify the following approaches:
Custom, which links generation, evaluation and selection via one-to-one selective data exchanges. E.g. software that links analysis packages with conceptual tridimensional models of buildings to eliminate file exchange and improve the interactivity between design, evaluation and selection. The linkages are labor intensive in terms of implementation and with limited reusability because of the specificity of the integration.
Interoperability, which involves exchanges among systems that are based on industry standard neutral formats. Examples can be found in Building Information Modeling (BIM) paradigm in the Architecture, Engineering and Construction (AEC) industry. These exchanges poorly preserve parametric features and have not yet attempted to represent the simultaneous interaction of different sources of design knowledge while designing.
Model-based System Engineering (MBSE), which implements computational modeling techniques to support the formalization of requirements, design, evaluation, verification and validation of a product.
