A basis for the development and validation of design ideation models for conceptual engineering design.

摘要

A key element in new product development is design ideation, especially during conceptual design where it is estimated that 70 to 80% of the cost of the product is committed. Unfortunately, there is lack of formal methods for design synthesis for generating alternatives. Further, existing ideation methods are not theoretically based and there is little empirical evidence of their effectiveness. The objective of this dissertation is two-fold: to develop and validate an experimentation methodology and to develop a theoretical foundation for the creation of models of design ideation. The experimentation framework developed consists of identification of ideation components, effectiveness metrics, design of factorial experiments (i.e. DOE) and analysis methods. Simultaneous experiments were conducted at the Design (Engineering) and Lab (Cognitive Psychology) levels. An approach was developed to align such multi-level experiments and verify if alignment was achieved.; Experimental results were conducted on six ideation components (Provocative Stimuli, Suspend Judgment, Flexible Representation, Frame of Reference Shifting, Incubation, and Provocative Stimuli). Provocative stimuli (P) and suspend judgment (J) present similar effects. Suspend judgment clearly improved quality and decreased quantity. Frame of reference shifting (F) and Example Exposure (E) have similar effects. Incubation shows a constant positive effect on all metrics. Frame of reference shifting has a strong effect even when interacting. In Set 1 (PJR) Suspended judgment (J) had a perfect alignment (100%). For Set 2, the overall alignment was 80%, with Incubation (I) having the best alignment. The overall results for both sets indicate a good level of alignment (59%), but what may be most useful is derived from the other 41%. Evaluating the reasons of no-alignment provided insight into various relevant issues such as: Ideation principles as more fundamental than ideation components, better understanding of ideation components (some are stronger, some are easier to manipulate, interactions are complex, etc.), distinction between process and outcome metrics (outcome being harder to improve than process), quantity to quality match (more ideas means higher quality), addition of new ideation metrics (Best Novelty and Best Quality). The experience gained will be useful when improving the experimental procedure for future experiments. A series of cognitive models for elements of design ideation were introduced; providing an insight into the mental information and processes occurring during the ideation process. The modeling basis provides a common ground for the combination of theories from various research areas: Engineering Design, Cognitive Psychology, Human Problem Solving, and Human Perception, among others. Such modeling capability is fundamental for the development of CACD for intuitive ideation.