Using Statistical Process Control to monitor milking process

Twenty milking parlours of Northern Italy were involved in a field study over one year. Milking performance data were extracted from the herd management software of each milking parlour at each milking session. Through the Principal Components Analysis, two new variables were identified (principal components) that can be interpreted as a synthetic index respectively of flow rates and duration and parlour efficiency. The milking process was then analyzed using Shewhart individuals control charts based on the principal components previously identified.