Works in progress

Remote Sensing in Agriculture

Research line: Digital Agriculture

We work in remote sensing and artificial neural networks applied to agriculture, in crops such as peanuts and soybeans. These projects aim to evaluate the biophysical characteristics of cultures using remote sensing (soil, aerial and satellite) and the data are processed in statistical software such as Python.

(Dr. Franciele Morlin Carneiro)

Tools for automated management of crops

Research line: Digital Agriculture

This research project is focused on developing tools for automated management growth, yield and quality of crops where the economic interest raw material is located below or above ground with implementation of remote sensing techniques, computational vision, machine learning, and deep learning. 

(M. Sc. Danilo Tedesco-Oliveira)

Selective mechanized harvesting of coffee

Research line: Coffee harvesting

The selective mechanized harvesting of coffee is carried out with the objective of increasing the quality of production and, consequently, its value. However, the success of this operation is related to the strength required to remove the fruits from the plant. This way, and based on hypothesis that the force required for the release of coffee fruit varies according to the time of day and the face of exposure to sunlight, directly influencing the efficiency of mechanized selective harvesting, we are evaluating the behavior of the detachment force of coffee fruits during the day, as well as in relation to the maturation stages and the face of exposure to sunlight.

(M. Sc. João de Deus Godinho)

Quality of the coffee mechanized harvest

Research line: Coffee harvesting

One of the limitations of the coffee harvest is the unevenness of the maturation of the fruits, a factor of extreme influence in the choice of the ideal moment for the beginning of the harvest, obtaining less losses and a final product of higher quality.

It is assumed that coffee crops in an organic production system present soils with higher organic matter contents when compared to coffee plantations in a conventional system, consequently this soil presents physical condition, providing greater water retention, and better development of the root system, contributing in the formation, filling and maturity of the fruits, thus providing the need to evaluate the quality of the mechanized harvest of the coffee crop according to the physical attributes of the soil under an organic and conventional production system. 

(Agron. João Paulo Leme Donadelli)

Neural networks and remote sensing to support decisions

Research line: Digital Agriculture

This research project is about developing methods to apply neural networks associated with remote sensing to develop support decision system for precision agriculture in corn, coffee and peanut.

(M. Sc. Mailson Freire de Oliveira)

Quali-quantitative losses in peanut harvesting

Research line: Peanut harvesting

Analyzing the impurities from mechanized peanut harvesting, we seek the best way to quantify losses in the crop and, thus, analyze the economic impact of these losses. 

(Agr. Eng. Armando Lopes Brito Filho)

Peanut maturity estimation

Research line: Peanut harvesting / Digital Agriculture

We developed research in the field of agricultural mechanization, precision agriculture and remote sensing, working on the development of a new methodology to predict the maturity of peanuts using remote sensing.

(Agr. Eng. Jarlyson Bruno de Souza)

Mechanized sugarcane harvesting: performance and impacts

Research line: Sugarcane harvesting

The mechanized sugarcane harvesting can cause irreparable damage to the crop, such as the shattering of stalks, trampling and even the pulling of the brass knuckles. The emergence of technologies contributes to the evolution of the harvesting  method, however it requires knowledge and study in order to achieve the expected objectives. Based on the hypothesis that mechanized sugarcane harvesting can present better effective quality, we seek to determine ideal standards for adjusting the harvester mechanisms, in order to obtain better quality of the raw material, longevity of the cane field and less wear on the active organs of the harvester.

(Agronom. Bruno Rocca de Oliveira)

Tests on agricultural tractors for spraying citrus crops

Research line: Balance and operational performance

We carry out tests on agricultural tractors promoting changes in the ballast index, tire calibration, dynamic advance of traction and verifying the reflection of these changes in the operational performance of the spray in the citrus crop. We also carry out ergonomic, visibility and safety analyzes of agricultural tractors.

We are currently looking for a new methodology for testing visibility using Drone images.

(Agron. Alex Rangel Gonzaga)

Digital Ag and Remote Sensing in cotton crop

Research line: Cotton harvesting / Digital Agriculture

We have conducting research with intelligent systems on farms to optimize cotton yield prediction and the use of images of large land with cotton for decision making and traceability of the big data obtained from the harvester and its relationship to fiber quality.

(M. Sc. Franciellle Morelli Ferreira)

Precision Mechanized Sowing

Digital Agriculture

The main goals with the project are to test if different vertical loads of planter's active hydraulic downforce systems in distinct soil textures have influence on seeding depth, plant development (seedlings emergence, plant height, dry mass) and yield of cotton, corn, soybeans and peanuts.  

(M. Sc. Luan Pereira de Oliveira)

Are Machine Learning techniques able to measure the slippage index of a agricultural tractor?

Balance and operational performance / Digital Agriculture

The main goals with the project are to test if different vertical loads of planter's active hydraulic downforce systems in distinct soil textures have influence on seeding depth, plant development (seedlings emergence, pOne of the most relevant factors for a fuel consumption optimization is the operational balance. The slippage index is a key indicator of how well the set tractor plus implement is adjusted. When the set of tractor & implement is out of the slippage range, many side effects arise such as excessive fuel consumption, increase of maintenance costs, premature tire wear and poor operational performance. Based on this background, the aim of this Project is to develop a mobile application based on Machine Learning techniques to recognize visual patterns and then measure the slippage of a agricultural tractor on the Field. Convolutional neural networks algorithms results will be compared to the commercial device called "slippage meter". See illustrative picture for details and numerical example lant height, dry mass) and yield of cotton, corn, soybeans and peanuts.

(Electr. Eng. Rafael Feih)

Data-mining in telemetry data

Digital Agriculture

Digital agriculture is inserted in practically all agricultural processes at a time when information and computer technologies are growing exponentially. Decision-making must be carried out with greater speed and precision from data from systems connected to the internet. From this scenario, there is a demand for tools to transform Big Data into Right Data to support decision-making that will optimize the use of inputs and rationalize natural resources.

(Agron. Caio Donadon)

Methodological innovations for evaluation of sugarcane harvesters

Research line: Sugarcane harvesting

The field evaluation of quali-quantitative losses in mechanized sugarcane harvesting is a difficult activity to be carried out.. So, we are developing a inovative method to evaluated sugarcane harvesters performance. 

(Agron. Antônio Maurício Loureiro Jr)