Technology can be used to monitor the growth of any agricultural crop
Monitor sugarcane growth in order to estimate the best time for harvesting. This practice has already been carried out in some fields in the interior of São Paulo by means of a miniaturized radar system carried on drones.
Developed by the startup Radaz, with support from the Innovative Research in Small Businesses Program (PIPE), in collaboration with researchers from the State University of Campinas (Unicamp), the technology was described in articles published in the Remote Sensing magazine and presented in a lecture during Brazilian Bioenergy Science and Technology Conference (BBEST) 2020-21.
The online event, which took place between Monday (24/05) and Wednesday (26/05), is part of the activities of the FAPESP Bioenergy Research Program (BIOEN) and took place in parallel with the second edition of the Biofuture Summit, promoted by the Platform for Biofuture – a consortium formed by 20 countries, including Brazil, with the aim of promoting low-carbon transport solutions and the bioeconomy.
“The technology can be used to monitor the growth of any agricultural crop. We chose sugarcane because it is one of the most cultivated in the State of São Paulo”, said Hugo Enrique Hernández Figueroa, professor at the Faculty of Electrical and Computer Engineering at Unicamp and project coordinator.
The miniaturized radar system onboard drones operates in three microwave bands simultaneously – P, L and C –, which are emitted by the device over an area, for example, a sugarcane field.
Each of these three microwave bands operates at a different frequency range and has different wavelengths. The longer it is, the greater its ability to pass through an object, such as the top of a tree, and reach and even penetrate the ground.
The P-band, for example, operates at a low frequency of 400 megahertz (MHz) and has a longer wavelength of 75 centimeters. In this way, it is able to reach the surface of a sugarcane plantation and penetrate up to 50 centimeters into the soil of a land with shallow or non-existent vegetation. The L-band operates at a higher frequency, 1,2 gigahertz (GHz), and has a low wavelength of 25 centimeters. And the C-band has a much shorter wavelength, 5,5 centimeters, and operates at a much higher frequency, 5,5 GHz.
The combination of data obtained by the reflection of these three microwave bands on land and captured by radar allows the creation of a digital model of the area with crucial information for farmers.
“In a sugarcane field, for example, it is possible to estimate the height and volume of plants, in addition to humidity, by combining the data obtained. We were also able to detect problems such as erosion and lowering,” explained Figueroa.
field tests
The remote sensing radar motherboard weighs 600 grams. Once it is loaded onto drones, the device can perform complex trajectories, such as circular flights, with variable radii and helical flights.
“In this way, it is possible to improve the spatial resolution of an area flown by 27 times compared to monitoring done through linear flights, as is usual by plane, for example”, compared Figueroa.
Through a project supported by FAPESP, within the scope of the Research Support Program in Partnership for Technological Innovation (PITE) and a cooperation agreement with IBM Brazil, the researchers improved the algorithms that interpret the images obtained by the system in a way to produce information according to the needs of users and depending on the type of plantation analyzed.
To validate the algorithms, studies were carried out in an experimental cultivation area at the Faculty of Agricultural Engineering (Feagri) at Unicamp, where sugarcane was planted and plant growth was recorded using conventional biometric measurements and the radar system. transported by drone.
The study indicated that the results obtained through the technology were very similar to those provided by the conventional system.
To validate the system's ability to estimate the amount of above-ground biomass produced in a sugarcane crop three months after planting, the researchers carried out an experiment on a farm owned by Usina São Martinho, in the interior of São Paulo.
“The results of the study, which was accepted for presentation at an international conference, also indicated that forecasts made using the system were in good agreement with those made by technicians on the farm,” said Figueroa.
original article published on the Agência Fapesp website.
