Overview

Leveraging the power of algorithms to address societal challenges is becoming increasingly common in today’s world. This project focuses on utilizing such models to create more efficient and automated Automatic Milking Systems (AMS), based on data from a farm in Northern Italy. During each session, the milking robot collects extensive data on individual cow productivity and milking behavior. Important metrics such as fat, protein, and lactose levels are measured to evaluate milk quality. However, these measurements are currently obtained using colorimetric methods that require calibration every two weeks. The goal of this project is to infer lactose content using data from the milking robot. If this approach can be scaled to include fat and protein, it could eliminate the need for traditional measurement methods, thereby simplifying the evaluation of milk quality.

Full Report

The full Report, with detailed explanations and results, is present in the 'Final_Report.pdf' file.