RESUMO
The current approach to the ergonomic design of furniture often relies on a gender-based segmentation, assuming that it yields homogeneous groups. However, this assumption has proved to be inconsistent in various contexts, particularly in regions characterized by ethnic diversity and economic inequalities, leading to diverse anthropometric characteristics. Considering gender alone fails to encompass this heterogeneity. The proposed approach introduces an 'anthropometric segmentation' methodology in which the population is clustered based on a wide range of anthropometric dimensions. Additionally, kernel density estimation is used when normality is not fulfilled. This step is critical, as it informs furniture dimensions by using selected percentiles from these density functions. The proposed approach is applied to a dataset of university students in Peru, a country marked by significant heterogeneity within its male and female populations. Ultimately, our methodology promises to improve the ergonomic design of furniture by accounting for a broader spectrum of anthropometric diversity.