STATISTICAL PROCESS CONTROL APPLIED TO THE FOOD INDUSTRY
Statistical Process Control (SPC), Multivariate Control Chart and Food Industry.
The biscuit industry generated 136 billion dollars in sales and increased 9.68% in 2018. As companies in this segment, they need to adopt methods that enable them to optimize or improve their processes, in order to remain competitive in the market. One of the methods used to meet this objective is the Statistical Process Control (SPC) used to decrease the variability of a process based on statistical analysis. The main objective of this research was to implement statistical process control for monitoring of the biscuit manufacturing process, elaborating univariate and multivariate control charts. The biscuit production line was evaluatedat three collection points using two type of molds (A and B to assess the correlation, those impacted on the colleting points for five process variables: raw weight, roasted weight, thickness, length and width. Statistically significant differences were found between these sites when mold A is used, and it is recommended to collect it from the MEIO collection point. This same difference was not evidenced using the B mold. The Anderson-Darling test was performed to assess adherence of the variables to normality. The gravimetric variables (raw weight and roasted weight) are not normal variable (p-value < 0,05), while the dimensional variables (thickness, length and width) showed adherence to normality. The improvement made by process control reduces the quality losses. Before starting this work, the quality losses were 10.37%, when finished this study it was 7.52%, considering a reduction of 27.48%. Improvements were recommended for the company to provide future work.