Sensors driven system coupled with artificial intelligence for quality monitoring and HACCP in dairy production

Abstract

The maintenance of good milk quality standards is still a challenge for dairy farmers that requires a rapid control system that is compatible with both the environment and production cost. A patented Hazard Analysis and Critical Control Points-like remote diagnostic (sensor driven) system named BEST was implemented to enable both quality monitoring and traceability in the dairy chain. BEST was daily tested in a dairy farm to identify new reliable indicators of anomalies (safety and quality) in milk production based on a Machine-Learning approach. The database obtained in four months of sensoristic analysis was subjected to a statistical study with AI algorithm to identify outliers. BEST proved ability to spot cows with specific characteristics in the whole herd's database. In particular, AI highlighted the sole cow from a different breed, the only cow that recently gave birth and the only cow in the herd that received treatment with the drug Micospectone® (Lincomycin + Spectinomycin).