Recent advances in computer technology, such as the development of global navigation and geoprocessing systems, are expanding the possibilities for use of unmanned aerial vehicles, or drones, in agriculture. Relatively cheap and easy to use, these aircraft, which come equipped with more and more efficient and accurate sensors and image resources, can help farmers increase productivity and reduce crop damage through survey data able to detect pests and estimate the growth rate of plants, to name a few. Given the possibilities of using these aircraft, computer scientists Bruno Squizato Faiçal, Heitor Freitas and Professor Jó Ueyama, of the University of São Paulo’s Institute of Mathematical Sciences and Computation (ICMC-USP), located in São Carlos, have developed an autonomous intelligent system for agrochemical spraying using drones.
The use of agrochemicals is essential in large-scale agriculture. These defensive chemicals, in general, are sprayed on crops by hand or with the help of tractors. Even with the use of some sort of protection, such as masks, farm workers are exposed to the product, which can cause serious health problems such as cancer and adversely effect the central and peripheral nervous system. Brazil is the largest consumer of pesticides. Sales in Brazil have grown substantially in recent years, jumping from $2 billion in 2001 to more than $8.5 billion in 2011, according to a report by the National Cancer Institute (INCA) on the risks to human health from the use pesticides. Controlling the amount of agrochemicals applied to crops, in turn, s very difficult. Spraying is often subject to meteorological factors such as wind speed and direction, which may compromise its application in the growing area and spread it to neighboring areas.
The system developed by the researchers at ICMC-USP orchestrates the use of a drone with rotating wings, in the form of propellers, and a wireless sensor network installed around the growing area. It is based on an artificial intelligence system capable of adjusting the route of the aircraft according to specific weather conditions. According to the researchers, this is accomplished through an exchange of data between the drone and real time sensors installed along the periphery of the area to be sprayed. “First, the drone does some test runs at different altitudes and under different weather conditions to ascertain the system’s pattern of spray dispersal and the influence of weather conditions, says Faiçal. “This information is stored and used later to build a knowledge model, which enables the drone to make decisions when spraying in weather conditions similar or not to those previously encountered.”
As it approaches the sensors installed around the sprayed area, the drone compares the information it generated with the information obtained in real time by the equipment in the ground. After comparing this information, the system is able to regulate the release of the chemical on the crop. The idea is for the aircraft and sensors to work autonomously, with a control station and a technician to monitor the progress of the process.
The drone’s coordinates recorded in the navigation system, based on comparing the calculations between the aircraft and the sensors, determine the power of a pump that regulates the amount of agrochemical released. The higher the power, the more product that is released. According to the researchers, it makes spraying safer and more accurate, improves not only agrochemical coverage but also the quality of the growing process, ensuring greater use of these substances by the plants with less harm to the environment. The system was evaluated using a rotary-wing drone with eight electric motors powered by batteries and a 2.5 kg load capacity in open fields on university property.
The prototype was effective in releasing controlled amounts of agrochemicals in predetermined areas, while accounting for meteorological conditions and using routes calculated by the GPS system. “Our system can guarantee a specific and intelligent application with less waste and less exposure on the part of the farmer to the pesticide,” says Faiçal. Another advantage, he says, is that the same system can be adapted and installed in other vehicles used on land, such as tractors, for example, by connecting them with sensors scattered in the fields. A long road lies ahead, however, before the technology is available to farmers. One of the challenges is to adapt the entire system for use in larger aircraft, capable of spraying large agricultural areas. The ICMC-USP team has already filed a patent application with the USP Innovation Agency.
In the past 15 years, farmers in many countries have begun to see drones as an opportunity to apply so-called precision farming to their fields, based on using the instruments and information technology resources to implement improvements in agricultural production. The advantage of drones over other monitoring systems is that they can make weekly, low cost flyovers throughout the entire production period. Since 1990, Embrapa Instrumentação, based in São Carlos, São Paulo State, has invested in the development of new systems and aircraft capable of operating and performing well even under harsh field conditions. Under the coordination of electronic engineer Lúcio André de Castro Jorge, its projects seek to expand the possibilities of adapting drones to various agricultural operations; they use high definition conventional cameras, sensors and thermal and multispectral cameras, which generally monitor crops to estimate production volume and gauge the presence of disease and pests.
In Gavião Peixoto, a municipality near São Carlos, researchers are testing new components on drones similar to a mini-helicopter, with a propeller measuring 2.80 meters (m) in diameter. They make periodic flights over orange groves to detect greening, a disease that affects fruit ripening, turning the leaves of the plants yellow, and it is usually identified only at an advanced stage. More flexible and accurate during spraying, without the use of sensors and manual operation, and also used on rice, soybean and wheat crops, the Embrapa drones are part of a broader set of research focused on software development and embedded image capture systems suitable for various agricultural applications, on both small and large properties.
The projects are supported by the Precision Agriculture Network of the Brazilian Innovation Agency (FINEP), the National Council for Scientific and Technological Development (CNPq) and Embrapa itself, and are being developed at the new National Reference Laboratory in Precision Agriculture (Lanapre), opened in 2013. Working in the new laboratory, Jorge’s team are in the process of designing new integrated systems, testing them to see how they perform in soy and corn plantings, as well as in cultivated pastures on Embrapa’s farm next to the laboratory. “The tests are done using different models of aircraft, such as the Isis from Hórus, the Echar from XMobots, and the Otus from Aérials, all of which are Brazilian companies; they are loaded with free software, which helps bring down the cost of the innovation process,” says Jorge.
The systems now being developed by Jorge’s group can potentially identify flaws in planting, estimate the rate of plant growth and detect different levels of nutritional stress. They can also discover anomalies caused by rust, water stress, fungi and pests by using multi- and hyperspectral cameras capable of simultaneously obtaining images with high spatial, spectral—with various electromagnetic wavelength ranges on the formation of the photograph—and infrared resolution. Others, still in development, will help in spreading eucalyptus seeds and releasing the natural enemies of some pests for biological control.
The first drones in Brazil began to be developed in the mid-1980s with the Acauã aircraft, a mini-airplane designed by the Aerospace Technical Center (see Pesquisa FAPESP Issue nº 211) in partnership with the Avibras company, initially for military purposes. Beyond military applications, industry growth is reflected in a variety of applications, from environmental monitoring to aerial inspections to combat drug trafficking operations and humanitarian missions, such as delivering drugs and vaccines to remote areas.
Even with the rapid development in this sector, drone flights still lack specific rules. Brazilian law has advanced somewhat in that direction based on a recent request for public comment carried out by the National Civil Aviation Agency (ANAC), according to Jorge. “There will be no restrictions on the use of drones over fields, provided that established altitudes are respected,” he says. Companies in the industry are selling aircraft, parts and software, but they await more specific regulations. On December 14, 2015, the U.S. Federal Aviation Administration (FAA) announced new rules on using drones in its airspace. Owners of these aircraft in the U.S. will have to register their name, address and e-mail in a national database, which will generate a certificate of registration for the aircraft.
Drones are largely responsible for the expansion and development of the global aerospace industry, with estimates of huge investments in the future, from the current $2.7 billion per year to about $8.3 billion annually in the next decade. The possibilities are many and with industry regulation civil use of the technology should be even more robust. While Brazilian regulatory bodies look for the best way to deal with security and privacy issues, Lux Research, a U.S. consulting firm specializing in market research, estimates that more than one million drones will be sold by 2025.
Embedded technology is likely to account for $670 million of the $1.7 billion spent in this market during that period. In a 620-page report published by the U.S. company RnR Market Research and entitled: Commercial drones: Market shares, strategies, and forecasts, worldwide, 2015 to 2021, industry analysts say that drones are changing the nature of agriculture all over the world and that these aircraft represented an expenditure of $609 million in 2014. They estimate this number will reach $4.8 billion worldwide by 2021.
1. Using computational intelligence and UAVs to reduce drift in the application of pesticides (nº 2013/18859-8); Grant Mechanism Scholarships in Brazil – Regular – Doctoral; Principal Investigator Jó Ueyama (ICMC-USP); Grant Recipient Bruno Squizato Faiçal (ICMC-USP); Investment R$ 72,650.00.
2. Exploring the sensor web approach and participatory sensing approaches for urban river monitoring (nº 2012/22550-0); Grant Mechanism Regular Research Grant; Principal Investigator Jó Ueyama (ICMC-USP); Investment R$60,529.50.