Functional Data Analysis as a Tool to Find Discomfort Evolution Patterns in Passenger Car Seats

More and more, passenger comfort is a topic of interest, mainly in vans which are sold to a public that considers it as a factor when deciding the purchase of a vehicle. However, the classical techniques used to evaluate discomfort show some shortcomings that limit the possibilities of differentiating products that have reach a high degree of performance.

In this work, principal components analysis and functional data analysis are combined to overcome these shortcomings. Principal components analysis is a tool that permits to explain most of the variability of a sample by means of a set of uncorrelated variable. Functional Data Analysis is a technique that permits analyzing data as a continuum, allowing to find patterns and factors in time dependent variables. In the present work, the static comfort evolution of users while seated on second row van seats has been evaluated by means of principal component analysis of functional data and a set of comfort evolution patterns have been found. Although static comfort data have been used, these techniques can be immediately transferred to analyse discomfort data from real vehicle tests or simulation in dynamic platforms.