A multidimensional scaling approach for product design and preference modeling

Taking the needs of a consumer into account is imperative in product design. In many cases, aesthetic properties and, more generally, subjective assessments are as important as technical functions. When one considers the subjective part of the requirements, the feelings, impressions, sensations or preferences of a customer have to be quantified and modeled in advance. This is a main challenge in industrial design. One needs to develop methods and tools adapted to capture the subjectivity of the human assessment and integrate it into the design process. In this context, we propose an approach using multidimensional scaling (MDS) for the building of a perceptual space, in which the perceptual attributes of a family of products can be described. Next, we develop a new technique in order to take the customer's preferences into account for products in this perceptual space. Our results are intended to provide to the designer a description of the perceptual attributes of the product and information to position the next generation of products in a high preference area.