Interactive Evolutionary Concept Optimization System
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for gathering consumer preferences in product development, such as consumer clinics and Conjoint Analysis, face challenges including limited design exposure, biased group dynamics, and the assumption of pre-existing preferences, leading to inaccurate results and increased respondent burden.
Innovation Solution
A system and method that uses genetic or evolutionary computational techniques to generate and present design alternatives to participants, allowing them to rate and refine preferences through a flexible interface on mobile devices, enabling iterative feedback and reducing interpersonal dynamics.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional consumer clinics and Conjoint_analysis are used to gather consumer preferences, then consumer input can be collected, but the results are biased due to group dynamics and limited design exposure
Solution Approach 1:
The patent uses virtual representations of consumers (digital twins or avatars) that can be replicated without physical constraints. These virtual consumers can be presented with numerous design alternatives simultaneously, eliminating the bias introduced by physical focus groups where only a few designs can be shown at once. The virtual consumers provide consistent, unbiased feedback across multiple design iterations.
Solution Approach 2:
The patent transitions from two-dimensional physical consumer interactions to multi-dimensional virtual environments. Consumers interact with designs through immersive interfaces that can present numerous alternatives simultaneously across different dimensions (visual, functional, emotional), enabling comprehensive preference measurement without the limitations of traditional one-on-one or small group consultations.
2Measurement precision
If consumers are presented with many design alternatives to improve measurement accuracy, then preference identification becomes more accurate, but respondent burden increases
Solution Approach 1:
By using virtual consumer representations, the system can present thousands of design alternatives simultaneously to each virtual consumer without increasing their cognitive load. The virtual consumers process information at computational speed, eliminating the time and effort constraints that limit human respondents when faced with numerous options.
Solution Approach 2:
The system presents design alternatives in periodic, managed batches to virtual consumers, allowing efficient processing and feedback collection. This periodic presentation mechanism enables comprehensive evaluation of many designs while maintaining manageable interaction cycles, unlike continuous human response collection which becomes overwhelming.
3Loss of information
If focus groups are used to gather consumer feedback, then interpersonal dynamics can be observed, but biased group dynamics distort preference data
Solution Approach 1:
The patent creates virtual consumers that replicate individual consumer behaviors and preferences without the social dynamics of real human interactions. These virtual consumers provide honest, unbiased feedback because they lack the social pressures, conformity biases, and groupthink that distort preferences in physical focus groups, while still capturing rich individual consumer insights.
Data Source
AI summary
A method includes determining a plurality of data points, determining a distance between each data point and each of the other plurality of data points, choosing a first one of the plurality of data points, identifying all of the other plurality of data points within a maximum distance of the chosen data point, repeating steps the previous steps to choose a different one of the plurality of data points until all of the data points have been chosen, identifying one or more clusters each having a predefined minimum number, K, of data points within a predefined search radius and analyzing the one or more clusters with respect to K linkages.


