A systematic study of conventional, sensor-induced, and machine learning-based methods for milk quality analysis and adulterant detection

Abstract

Milk is important constituent of daily food worldwide. Quality of milk is compromised because of multiple issues like sub-standard practices in dairy industry, frequent adulterations, microbial contaminations, storage and climatic conditions. The problem gets multifold grave because of lack of portable solutions of milk quality assessment and testing. Still the milk quality testing remains centralized and lab oriented. So, the need of the hour is to study and figure out solutions which can be portable, accurate and provide results in timely fashion. The current study tries to bridge this gap by analyzing recent development in the field of milk testing especially sensor-based techniques coupled with the power of Machine Learning based classification strategies. The search is for a milk quality testing solution which is reliable, all-in-one testing solution, low cost, portable, anywhere accessible, friendly user interface and operates in real time. It should employ state of the art technology like connectivity and novel Artificial Intelligence and Machine learning methods. The study concludes that sensor infused machine learning solutions provides an upper edge with respect to traditional lab based slow and costly solutions. Internet of Things and Sensor-based technology is opening doors for real time portable milk testing kit which is showing promising result with backend prowess of Machine learning methods. Our study is driven of societal impact especially in the area of food safety and rural empowerment.