Remote sensing

Help from the sky

Unmanned aircraft can be used to increase the productivity of the country's crops

Imagem: USP/EMBRAPAImage shows the regularity of the planting of the eucalyptus plantationImagem: USP/EMBRAPA

In the second half of this year, Brazilian farmers may be able to count on a powerful ally to monitor and to raise the productivity of their crops. This is an unmanned, autonomous aircraft, planned to fly over the plantations and capture images that will afterwards be analyzed and processed by specialized software. The little plane, which is at the final stage of development, has a four-hour flight autonomy and capacity for recording 6 thousand photos a day, flying 100 meters from the ground.

“The images, captured by a digital camera fitted to the aircraft, can be used by the rural producers to detect various problems in the development of the crops, such as failures in the planting, infestation by pests, shortage of nutrients, and the presence of diseases and invasive plants, amongst others”, points out researcher Onofre Trindade Júnior, the coordinator of the product that involves researchers and technicians from the University of São Paulo (USP), Embrapa Agricultural Instrumentation, a unit of the Brazilian Agricultural Research Corporation and AGX Tecnologia, a company from São Carlos focused on the development of technological solutions for the agricultural sector. This company, which is part of the Fazenda Campo Bom group, of Mato Grosso do Sul, invested the major part of the R$ 1.5 million spent to date with the project.

Unmanned aerial vehicles, known by the acronym UAV, can be used in many civil and military applications, explains Onofre Júnior, who is a licentiate professor at the Mathematical Sciences and Computing Institute of USP in São Carlos. The aircraft developed by his group are part of the Arara (Radio-Assisted Autonomous Reconnaissance Aircraft in the Portuguese acronym) Project, which started in 1998, and are aimed at agricultural and ecological monitoring applications. This is done by means of collecting video images and photographs, which are subsequently processed to extract information of interest.

The majority of the images are collected in the spectrum of visible light, but some are taken in the region of far infrared (thermal images), ideal for detecting outbreaks of fires and the presence of animals. At the moment, the researchers are finalizing the modification of one digital camera to get images in the near infrared range to make it possible to identify better several factors that affect the crops, such as lack of water and some kinds of diseases, as well as to get development indices, such as, for example, the quantity of biomass in the crop.

Imagem: USP/EMBRAPAImage from a height of 600 meters makes it possible to count the citrus treesImagem: USP/EMBRAPA

In the case of agricultural monitoring, the images collected are processed afterwards, spawning a series of useful information, like the number of plants per hectare, the size of the crowns of the trees, the distribution of straw on the ground, an important factor in the direct planting technique, and the regularity in the plantation, which measures the variation in the distance between plants. In ecological monitoring, the main information gathered by the aircraft is the mapping and the accompaniment of erosion, detecting and forecasting forest fires, mapping water resources and counting wild animals.

Good resolution
With its 2.3 meters in length and 3.2 meters in span (from one wingtip to the other), the plane is made of fiberglass, with some components of wood and aeronautical aluminum, and is powered by a gasoline engine of the ZDZ make, normally used in model aircraft, with 40 cubic centimeters and 4.8 horsepower (hp). Controlled by mean of radio signals like a model plane, the aircraft is capable of flying up to a thousand meters in distance, and the flight is carried out at a height of between 100 and 300 meters.

According to the coordinator of the project, it offers a series of advantages in relation to other devices also used in agricultural monitoring, like satellites. “The spatial resolution of the image provided by our aircraft is much higher, making it possible to identify characteristics that are not measurable by a satellite, like counting small plants”, Onofre Júnior explains. “Besides this, it also makes it possible to collect images at any moment, while satellites have a limited time slot (when, circling the Earth, they pass the place on which the information is wanted) for getting images.”

There is another advantage, in the capture of images. While the obstruction by clouds can prevent the acquisition of good photos from the satellites at certain moments, the aircraft of the Arara Project usually flies below the level of the clouds and does not suffer from this problem. Finally, the researcher points out, using a satellite, there is a minimum area of cover for the purchase of the image, which can be much larger than the area of interest. Accordingly, with the aircraft developed in São Carlos, it is possible to get images at a lower cost. Conventional planes can also be used for agricultural monitoring, and, in this case, the main advantage of the small plane is its lower costs – both on acquisition and for maintenance and operation.

The main motivation for carrying out the Arara Project, the researcher points out, came from the aeromodelling activities carried out, on the insistence of one of his children. “After many hours of involvement of the sport, the idea arose of using aircraft in research and development”, he explains. The kickoff for the project was made with the work for a master’s degree of a researcher from Embrapa, Nilda Pessoa de Souza, in which she tried to evaluate the quality of images obtained with model aircraft and their applicability in agriculture. From then on, several other works for masters’ degrees have been carried out, giving priority to subsystems of the aircraft and techniques for processing the images.

The aircraft uses a distributed architecture of sensors and servomechanisms, relying on a total of 11 onboard microprocessors. Servomechanisms are devices that transform electrical signals into movements of the controls in the wings and body of the aircraft, such as the rudder, ailerons, elevators and flaps. The microprocessors include the central processor, barometric sensors (altitude and aerodynamic speed) engine monitor (rotation, temperature etc.), electricity generator and battery charge control, servomechanism controllers, flight stabilization system and automatic pilot. Besides this, the vehicle also has equipment purchased on the market, such as the video camera, a high precision GPS receiver (satellite-based geographical locator, an 8 megapixel digital camera, and radio and video transmitters and receivers.

Imagem: USP/EMBRAPA2.3 meters in length, the plane is produced with fiberglass, wood and aeronautical aluminumImagem: USP/EMBRAPA

One of the challenges for the Arara Project is to build autonomous aircraft, capable of making decisions by means of an analysis of the data collected in real time, such as, for example, to follow a road, the course of a river, or an electricity transmission line. To finalize the little plane, it is still necessary to develop an onboard mechanism for collecting images automatically and recording their coordinates, to refine the adjustments of the current controllers for greater precision, and to develop a route controller making it possible to perform the necessary maneuvers for getting images on missions of visual scanning of agricultural areas. When it is ready, probably in the second half of the year, it is going to work like this: before the flight, the operators define a mission, containing points for collecting images and points of the route (latitude, longitude and altitude) to be followed. This data is loaded into the aircraft’s central processor while it is still on the ground, and it carries out the mission autonomously, without intervention of the pilot. The parameters of the mission can be modified in flight, by means of radio communication with the control station or any other communication channel, such as mobile telephony or satellite.

Tests and regulations
Up until now, ten prototypes of the vehicle have been produced, which have been submitted to tests in sugarcane, orange, soybean, corn and eucalyptus plantations since the beginning of 2004. “We are happy with the results. We found that the positioning of the images was precise, and there was a correct identification of the characteristics of interest in the digital processing of the images, confirmed by a visual analysis of them.” The researchers also carried out pilot projects with some companies from the agricultural and forestry sectors to sound out the receptivity of the market and to evaluate the usefulness of the system. According to Onofre Júnior, to go into commercial operation, the aircraft is going, first of all, to have to undergo a licensing process that is now being handled by USP, through the USP Innovation Agency, and by Embrapa.

“As regulations do not yet exist in Brazil on the operation of autonomous aircraft, to start with, the operations may be remotely piloted or in the form of projects, intermediated by a research institution with the farmers interested”, Onofre Júnior explains. “Independently of this, there are several companies in Brazil interested in the equipment, and we believe that there is also a market for it abroad.” A plane like this will cost between R$ 60 thousand and R$ 70 thousand.

In the world and in Brazil, there are several types and sizes of unmanned aerial vehicles at the operational stage or under study, the great majority of which acting in the military area. Although unmanned planes are no novelty, Onofre Júnior guarantees that there is no other equipment offering a complete solution for agricultural and forestry monitoring. “The Arara Project covers everything from collecting the images in agricultural and forestry area, with equipment focused on this application – allowing takeoff and landing in rural areas, low minimum flight speed (40 km/h), high resolution cameras for evaluating small details, and low cost -, to the digital processing of the images, extracting them, and showing the data to the producer.”