It was always a concern for the food production industries to detect fraud, spoilage, and food safety during mass production and packaging. This can now be discerned with the help of technology. Machine learning is used to train artificial neural networks to distinguish food safety concerns. Graphene-based Ion-sensitive field effect transistors (ISFET) are used in this innovative technology. The integration of graphene-based  ISFETs with machine-learning models allows for the detection of chemical and environmental changes across a wide array of applications. One of the applications is the detection of freshness and quality of the drinks.

The use of ISFETs in modern technology is not new, as it dates back 50 years. However, the results were not satisfactory due to the limitations caused by data variations with changing temperatures.

Chemical neurons collect information related to the chemical composition of liquid and translate it into an electric signal. Among the chemical neurons, ion-sensitive field effect transistors emerged as a promising technology, because of their ability to determine collectively the concentration of ions such as calcium, sodium, and potassium.

Machine learning-assisted graphene-based ISFET could provide promising results in the food industry. The ion-sensitive field effect transistor is a field effect transistor used for measuring ion concentration in a liquid. When the ion concentration of the liquid changes, the current through the transistors will change accordingly. Here the solution is used as the gate electrode.

Initially, data sets are created to authenticate the food products and quantify food adulterations, and food safety.

The chemical sensor, ISFET, detects the freshness of the liquid, by detecting the change in the flow of electric current when the ions from the drink pass through the conductor.

When a liquid or solution comes in contact with the sensor, the chemical composition in the liquid interacts with the sensor. The electrical composition of the sensor is changed as per the concentration of ions in the liquid. This is transmitted as electric signals by the sensor. The variations in the electric signals are then used to determine the freshness of the liquid or fraudulence in food.

This innovation can be significantly used against different kinds of food adulteration happening in the food industry. The everyday food adulterations that are happening globally can be detected with machine learning technology. For example, the adulteration of milk with water, and other contaminants is an everyday scenario interfering with the nutritional benefits of people. There are conventional methods to detect adulteration, but they are inadequate as the contaminants possess physical properties similar to the unadulterated drink. The ISFET technology enables the differentiation of varying percentages of adulteration in the milk.

Another important challenge, in the food industry, is tackling food fraudulence. The authentication of food products is mostly necessary for foods or drinks that are similar in appearance and taste. For example, soft drinks like Diet Coke, Pepsi, coca cola; different types of coffee blends. Artificial neural networks, when trained with data on graphene ISFET, are highly capable of authenticating food items.

The main challenges in the food industry- food adulteration, and food fraudulence could probably be taken care of, with the robust machine learning-assisted graphene ISFET technology.

With further research, this technology can also be applied for use in:

• Healthcare industry - such as blood glucose monitoring

• Environmental changes

• Industrial process control

However, there are certain limitations to this model.

The reliability of ISFET is hindered by differences arising from environmental changes, material properties, manufacturing processes, and design considerations. These are the challenges that need to be overcome for the widespread commercial application of ISFET technology.

But, for now, we can look forward to having it differentiate between the real orange juice and the one adulterated with sweeteners and added preservatives.

References:

1. Pannone A, Raj A, Ravichandran H, Das S, Chen Z, Price CA, et al. Robust chemical analysis with graphene chemosensors and machine learning. Nature [Internet] 2024;634(8034):572–8. Available from: https://pubmed.ncbi.nlm.nih.gov/39385036/

(Input from various sources)

(Rehash/Yagna Prasanna Kondadi/MSM)