Sensors evaluate fruit quality non-destructively

The use of technologies such as artificial intelligence, big data, machine learning algorithm , internet of things e sensors, among others, have been showing excellent results in large crops such as soybeans, corn and cotton. But its dissemination in fruit growing is not yet a reality. A scenario that could change with developments such as optoelectronic equipment for the non-destructive measurement of quality attributes of fresh fruits, and an "artificial fruit" for measuring and transmitting temperature and humidity, which show promise for the fruit market. Aagribusiness and Agriculture Ddigital from Ptermination.

"This optoelectronic equipment allows tests to be carried out quickly and simply, without the need for destructive tests in the laboratory, which contributes to reducing costs and speeds up decision-making in the management and management of agricultural crops. The equipment already has patent protection ", says professor Bárbara Teruel, from the Faculty of Agricultural Engineering (Feagri) at Unicamp, who works in the line of non-destructive tests aimed at precision digital agriculture.

The Unicamp professor explains that fresh fruit evaluations are normally carried out in a destructive way, with waste in the analysis process to obtain product quality characteristics, in addition to the need for laboratory infrastructure. "Losses and financial losses occur. Fresh fruits on the shelf (for the final consumer) are visually evaluated regarding color, size and shape. In laboratory evaluation, there are different types of analyses: total soluble solids (TSS), total titratable acidity ( ATT), SST/ATT ratio, color by method L (luminosity), C (chromaticity) H (hue angle), firmness of skin and pulp, weight, fruit geometry. There are ranges that determine the ripening stage of the fruits."

According to Bárbara Teruel, the decision-making process for harvesting grapes for wine production, for example, is done through sampling, removing some berries from bunches in a given area. "These samples go to the laboratory, where TSS (°Brix), total sugars, pH and total acidity are measured. One of the problems is that maturation does not occur homogeneously in the area. And these analyzes are used to make harvest decisions , after a certain period, they need to be done two to three times a week. These are manual and time-consuming actions. SST are reliable tests, but not the only ones for determining the ripeness of a fruit. Samplings are generally done in 5 % of the total of a batch of fruits, depending on the level of inspection."

Bárbara Teruel supervised two doctoral studies in this line, and one of them resulted in OptoFruit, the name under which it was patented. The equipment was developed by Daniel Dos Santos Costa, teacher from the Federal University of Vale do São Francisco (Univasf), which reports on various trends in winemaking, with selective harvesting. "It consists of harvesting the ripest fruits to produce excellent wines, which increases the added value of the final product. But this is only possible with the development of equipment or devices that measure quality attributes in a non-destructive, non-invasive way and in real time, like OptoFruit."

Daniel Costa notes that once the quality attributes are determined, it is possible to identify the stage of grape maturation and make decisions regarding harvesting, without the need to take samples from the field and take them to the laboratory to carry out conventional measurements. "With this, we optimize the production process, reducing the time taken to make harvest decisions, improving management. We also reduce the need for an analysis laboratory, which requires reagents, inputs and different equipment."

Also according to the Univasf researcher, the device works by emitting beams of light at different wavelengths; light interacts with the fruit and provides a response that is directly related to the quality attributes - which, in this case, were calibrated for total soluble solids (°Brix) and phenolic compounds (anthocyanins and flavonoids). "It is worth noting that OptoFruit uses artificial intelligence models and that the information is transmitted wirelessly from the sensor to a computer, tablet or cell phone. These devices are non-existent in the country and, when imported, they are expensive."

Artificial fruit

A second doctoral thesis resulted in the “artificial fruit”, a technological product with a deposit of the French Patent & Trademark Office (INPI) (National Institute of Industrial Property) and by professor Osvaldo Campelo de Mello Vasconcelos, from the Federal Institute of Science and Technology of Amapá (IFAP). “It is an instrument for measuring and transmitting temperature records (of the product and the environment) and humidity (the middle one), with friendly human-machine interface. The artificial fruit works using two equipment: the first It has a casing thermally similar to a mango peel, with pulp simulated through a gel with thermal sensors (internal and external), strategically positioned to capture data on internal and external temperature and external humidity", explains the author of the research. 

Compliant vasconcelos, the artificial fruit hasi in its central equatorial region, an integrated circuit that captures information and sends it for the second equipment,  a server (Raspeberry PI), which processes and makes available in an operator access platform (Web, Android, IOs). "It is an innovative tool in digital agriculture, helping rural producers to monitor everything from harvest to the final consumer of the product sold. The equipment is capable of simulating the currentdios of fruit maturation and also to collect information on external temperature and humidity through wireless sensors."

In this collected information, says the IFAP researcher, it is possible to estimate the ripening stage of the fruits through parameterized equations, through information on the quality attributes that determine this ripening. "Such parameters can also help companies to build refrigerated chambers, providing information on density, specific heat and thermal diffusivity. The artificial fruit has an autonomy of 25 hours."

Precision Digital Agriculture

In the opinion of researchers involved in this line of research, agriculture is one of the activities that most incorporates enabling technologies, serving mainly more demanding international markets, and the expansion of agribusiness has been occurring due to the innovation of technologies aimed at precision agriculture. Digital agriculture (4.0) is aimed at increasing the relationship between productivity-area and maintaining product quality until the final consumer. As an example, we have low-cost weather stations with communication via Web and Wi-Fi, inspection of the harvest point with equipment that evaluates in a non-destructive way, traceability technologies and maintenance of the quality of the agricultural product, improving thermal conditions in the chain. of the Cold.