According to P&G inventors, perfumed products are typically designed and/or formulated using empirical methods or basic modeling. Such efforts are time-consuming, expensive and, in the case of empirical approaches, generally not optimized because not all components and parameters are considered.
Furthermore, aspects of such methods can be limited to only existing components. There is thus a need for an effective and efficient methodology that removes the shortcomings of such methods.
The modeling systems described in this P&G patent application meet this need; particularly to determine the malodor-reducing capability of raw materials to produce superior perfumed products. The models also are faster and more efficient than traditional approaches.
Mathematically modeling perfumed products
U.S. Patent Application 20160306909
Publication date: Oct. 20, 2016
Assignee: The Procter & Gamble Co.
This invention relates to methods of designing and making not only perfumed products, but also perfume raw materials. In particular, the mathematical model comprises a technique selected from options such as: multiple linear regression, genetic function methods, principal components regression, decision trees, nearest-neighbor classifiers and the like, and/or combinations thereof (see the original patent for the complete list).