Mapping from Product Shape Parameters into Kansei Image Using ANNs
Authors: Kesheng Wang, Lijuan Dai
Abstract
In today’s highly competitive market, the function of products has become the basic requirement of consumers. Consumers don’t just buy a product; they buy value in the shape of entertainment, experience and identity. The external shape expression of products is a critical determinant of consumer response and product success. However, because of subjectivity, this particular dimension of the consumers’ need is usually difficult to be expressed, quantified and assessed. In this paper, an Artificial Neural Networks (ANNs) based approach for Product Form Design (PFD) is addressed to deal with this problem. 2D draft modeling, semantic differential method and ANNs are used to set up an experiment for modeling and simulation. The experimental data are analyzed and modeled by the use of a Standard Back-Propagation (SBP) neural network. The ANN model establishes the mapping between product shape parameters and adjective image words. As a result, designers can input the product shape parameters of their new design configurations, evaluate the extent of their pleasant sensation, and then modify the shape parameters of the product until they can fit more closely to the consumers’ desire. To illustrate this approach, a LCD monitor design problem is taken as a case study. Due to the generalization of the approach, it can be extended to the development of other products.
| Attachment | Size |
|---|---|
| Mapping from Product Shape Parameters into Kansei Image Using ANNs.wmv | 11.62 MB |
Dear authors,
In your paper, you commented that designers can input the product shape parameters and evaluate the desired sensation, and then modify the shape parameters of the product until they can fit more closely to the consumers desire.
I have a concern here, do you think the designs will be a optimal one (for e.g. meeting basic functional requirements and desired sensation)? I am concerned that the designer has to use trial and error many times before he/she can obtain a design that meet desired sensation and functional requirements.
Thanks.
ang
Hi. Thank you for an intriguing paper for product form design.
I hope you could elaborate the advantages of your application in comparison to the works of the following researchers who also used ANN for product form design:
Hsiao, S-W. & Tsai, H.-C (2005) Applying a hybrid approach
based on fuzzy neural network and genetic algorithm to product form design. International Journal of Industrial Ergonomics, 35:411-428.
Lin, Y.C., Lai, H.H., Yeh, C.H. & Hung, C.H. (2005) A hybrid approach to determining the best combination on product form design. In Knowledge-Based Intelligent Information and Engineering Systems. Berlin, Springer pp.235-241
Thank you.
K.W. Ng