Use of AL & ML In Agriculture
ML & AI Applications: Many people believe that the food products sold in grocery stores don’t taste good. There are various reasons for this. One of them is that the food industry doesn’t encourage Plant breeders and Scientists to focus on the taste when developing new varieties.
When they want to focus on flavour, don’t have any tools to quickly taste fruit from thousands of cultivars. Researchers at the University of Florida describe a new method for “tasting” agricultural products based on their chemical profile.
Relation between the Flavour & Chemical Components:
The researchers behind this study focused on dozens of varieties of tomatoes and blueberries, including commercial cultivars sold in supermarkets, traditional varieties more likely to be found at farmers’ markets.
For each cultivar, they had two sorts of data. First, a chemical profile that specifies which chemical components are present in its fruit and in what quantities. Second, they used consumer panels to gather data from hundreds of real individuals who assessed each tomato or blueberry variety on factors such as sweetness and overall liking.
Their approach indicated how well the various chemical components matched with the ratings given to each variety by human tasters. Surprisingly, sugars and acids in the fruits only accounted for nearly half of the difference in taster preferences between varieties.
Breeders will be able to predict the Taste of their Produce:
Breeders have no direct control over which compounds are present in a cultivar’s fruit. Rather, they have an impact on genes that code for metabolic pathways that produce the chemicals that eventually determine the flavor of a fruit. Even with modern technology, it is a cumbersome procedure that is usually carried out on a big scale.
Measuring flavor has been significantly more challenging because the breeder has just two options: sample the fruit himself or create a panel of tasters. Sampling is highly subjective, and testing people in a systematic manner is costly.
“You can’t serve 1000 types in the same day if you put together a standard consumer sensory panel and gather 100 individuals into a room,” one of scientist involved in the research explains.
According to him, this new study is “a proof of concept” that “shows we can now develop models to achieve the same thing” by monitoring molecules.
Conclusion: This might be a termed as a breakthrough in the world of Agriculture but it has yet to face a lot of criticism as many experts believe this kind of research will never provide a flawless version of a specific fruit. For starters, taste preferences change over time and across cultures. Because machine learning models can only generate predictions based on the data they were trained on, they are limited in their predictions.
Show your support to Agri-Journalism
Dear patron, thank you for being our reader. Readers like you are an inspiration for us to move Agri Journalism forward. We need your support to keep delivering quality Agri Journalism and reach the farmers and people in every corner of rural India.
Every contribution is valuable for our future.