Artificial intelligence can detect foods unfit for consumption

In an article published by the Scientific American portal, Elaine Nsoesie of the University of Boston presented research that could change and facilitate recall processes. The proposal is based on artificial intelligence algorithms programmed to read the analysis of websites that sell, among other things, food.

This artificial intelligence comes in response to major concern from the FDA, the Food and Drug Administration, which is the federal agency of the US Department of Health and Human Services. Between 2012 and 2014 the first tests have begun to read reviews, identifying about 20, 000 potentially hazardous products (containing color, smell or other altered aspects) - which could assist in the detection and removal of batches of these goods.

According to a survey by the Paulista Association of Supermarkets (APAS), conducted by IBOPE, 20% is the percentage of the purchase intention of cold foods and dairy products in Brazil. Besides the lack of time, the greater variety of products is what motivates this habit in the country.

The World Health Organization also warns that more than 600 million people worldwide get sick every year from eating inappropriate foods. Nsoesie believes the initiative could help with the process of removing suspicious products from the world market faster.