Electronic Tongue: Study on the Freshness of Milk

Electronic tongue is an artificial sensory system that uses sensor arrays to collect a number of data for classification purposes. The main objective of our study is to define freshness of milk and identify factors that affect them. Recently, there are some issues concerning the reliability of expiry date labeled by the food industries. There are complaints about lots of product that are already spoilt even though the expiry date is still far from due. We are interested to identify factors such as temperature variation and amount of additive used that makes them to spoil slower. As from our stance, an experiment was performed to describe the effect of temperature variations on the freshness of milk. In the experiment, we used two types of milk; pasteurized and unpasteurized, which were placed at two different temperature; cold (in refrigerator) and room. From the experiment, we observed the changes in pH value of the milk due to the temperature as well as taking into accounts the physical changes of the milk; taste, color, smell and texture. After the experiment has been done, the data obtained was used in Matlab for classification purpose. By using the Neural Network algorithm, we are actually able to find the threshold line which actually separates the fresh and spoilt milk into two different regions. Meanwhile, based on the experiment that we have conducted, we could say that the use of the pH sensor itself which represents the e-tongue is not enough to determine whether the milk has become spoilt or not. As discussed with our supervisor, we have come to a conclusion that, the changes and drop in pH value cannot determine if milk is already spoilt since the one that remains fresh also have noticeable drop in its pH value. Even though we managed to estimate a threshold line that separates good and bad milk, still, the used of extra sensors (e-nose) will be a major boost to obtain a better classification.