|
Abstract:
|
Ingredient levels of food/beverage products are typically optimized using response surface methodology (RSM) with a single optimization goal: maximize consumer liking.
However, when a product is in its early stages of development, consumer testing costs are difficult to justify. Subjecting RSM treatments to a sensory test is an alternative, with the sensory profile of an existing, highly-liked product as optimization goal. In this test, intensities of pre-defined product attributes on treatments are determined. By definition, a treatment’s sensory profile consists of its intensity scores on all product attributes. Response surface regression of the intensities vs. ingredient levels determines the ingredient levels that will deliver a profile that closely matches the target product’s profile. This approach has a key technical challenge: if, for example, a product profile consists of 15 attributes, then the RSM optimization needs to calculate ingredient levels that satisfy 15 optimization goals simultaneously.
This paper will discuss case studies that used target sensory profiles as goals to optimize ingredient levels. Results of validation experiments will also be discussed.
|
