STiR coffee and tea magazine

38 STiR coffee and tea / Issue 5, 2018 (October / November) By Peter Keen Brazil's favorable climate ripens coffee in a concentrated period of weeks, making mechanical harvesting an eco- nomically reasonable choice but that much mechaniza- tion comes with its challenges. Up to 20% of coffee cherry falls to the ground during the harvest and from other causes, including rain, wind, disease, and pests. Agronomist Tiago de Oliveira Tava- res at Säo Paulo State University in Brazil observes that "ground coffee" is retrieved through mechanical sweep- ing but the machinery is heavy and compresses the soil, interfering with the trees' root growth and their ultimate levels of production. In response, growers use a process called subsoiling to break up this hard soil. A long blade is pulled behind a tractor to break up the subsoil very effectively but leaves an uneven surface. Uneven ground means more ripe coffee berries are left behind by the machines, and more dirt and stones are picked up. Growers must use additional soil manage- ment techniques to smooth the soil. For example, a har- Photo by Tiago do Oliveira Tavares and Rouverson Pereira da Silva. A mechanical coffee harvester straddles a row of coffee trees. Drones apply pesticides and conduct surveillance for disease detection Precision Agriculture recision agriculture is the emerging mainstream of farming. It uses technol- ogy to focus, zero in on, and narrow down, for example, land mapping, soil sampling, fertilization, pest, and disease control, with highly localized weather alerts. "Precision" means right down to the tightest degree of control and minimization of resource usage. Just one striking instance is the ability to apply fertilizer, pesticide, and water to a single plant, delivered from a drone or an internet connected tractor. Forty percent of commercial drones in Brazil are used in agribusiness. Edgard Bressani c.e.o. of Brazil's O'Coffee in Pedregulho, São Paulo, has been an early adopter. He reports that aerial high-resolution satellite mapping and interpretive geospatial software have cut overall production costs by up to 30%. The drone is a flying farmhand that, according to many reports, so accurately focuses spraying that it cuts water use by 30 to 50% and chemicals by 90%. Planting costs are decreased by as much as 90% and weeding time sped up by a factor of 5 with labor savings cut by a factor of 10. Brazil coffee growing is well advanced in many areas of precision agriculture (PA) but pace and penetration are fragmented. It's a general management discipline not a specific technology. There are wide variations in global applications and adoption. In Europe there is an emphasis on livestock farming, for instance, but not in Brazil, a diverse crop economy. Surveys show Brazil overall as very receptive to PA, somewhat lagging in areas of application such as artificial intelligence (AI), and marked by many startups that are expanding into foreign markets. Here are a few examples of precision in Brazil that either are specific to coffee farming or that apply generic tools to it. • XMobots is Brazil's leading de- veloper and operator of a wide range of UAVs (unmanned aerial vehicles). It was founded in 2007 and grew out of r&d and incubation projects at the University of Säo Paolo. Brazil has many strong univer- sity and trade associations to draw from in both technology, engineering and appli- cation expertise. XMobot's strength is in UAVs for aerial mapping and inspection. Its cameras provide an image resolution of up to 2.5 centimeters per pixel. They are increasingly being applied to hyperspec- tral – invisible light – imaging to detect incipient plant disease and pest infection. O'Coffee has used XMobot UAVs in its continuous mapping of 3,000 hectares on six of its farms, for four harvests. • Stara: A fast-growing Brazilian producer of seed spreading and planter equipment, Stara was early in applying PA on the ground to help small farmers improve their core operations. It added real-time telemetry and GPS-guided au- tosteering sensors to its machines. These sensors turn a tractor or spreader into an intelligent device via IoT: The internet of things. IoT makes anything an internet address to link to images, applications and databases. Star's high-end seed spreaders have 20 IOT sensors, amounting to one- third of the cost. The sensors adjust the rate and amount of the seed dispensed to take into account the exact soil condi- tions, location, seed quality (measured in size and shape) and crop status. • Strider is focused on coffee dis- ease monitoring, via software on mobile tablets that access satellite images and decision support systems, including heat P

• Cover