An Early Detection System for Proactive Management of Raw Milk Quality: An Australian Case Study

Milk is a highly perishable product, whose quality degrades while moving downstream in an imperfect cold dairy supply chain. Existing literature adopts a reactive approach for evaluating and preventing milk with a high microbial index from moving further downstream in a dairy supply chain. In this paper, we argue that such an approach is not the best response if the intention is to maximize milk life in terms of quality. We propose a proactive approach that monitors the metrics of the temperature and the level that are the building blocks of microorganisms in milk. This information is then used to determine the status at which the storage tank should hold the milk in accordance with standards. This status is then compared with the tank's actual status, and if they are different from one another, it will prompt the farmers to take the required preventive actions to manage the quality of milk. The developed proactive management of raw milk quality approach is modeled by using a rule-based system and machine learning techniques with a high level of accuracy. To test the validity of our approach and demonstrate its applicability, we apply it to a milk farm in Queensland, Australia.