Startup supported by FAPESP has developed equipment for biological control of pests that affect soybean and cotton production (photo: Tarvos)

Smart traps monitor caterpillar infestation of crops

12 de janeiro de 2021

By Eduardo Geraque  |  FAPESP Innovative R&D – The simplicity of the device set up in the middle of a plantation may give a false impression of this powerful and effective tool for combating caterpillar infestation of crops. Developed by Tarvos, a Brazilian startup supported by FAPESP’s Innovative Research in Small Business Program (PIPE), the smart trap will be ready to operate on a commercial scale in the 2021-22 crop year.

The device consists of a 1.80 meter metal pole, on top of which is a trap containing a specific pheromone chosen to attract insects such as the Fall armyworm (Spodoptera frugiperda), the main cotton and corn pest in South America. When the adult moth enters the trap, it trips a smart camera that photographs the insect and transmits the image by satellite to a central processing platform.

“An algorithm runs the image against a database and the system produces a report on the number of insects recorded on a given day. It also displays the species identified,” says Andrei Grespan, co-founder and CEO of Tarvos.

The goal is to give the user a comprehensive pest and disease control overview and heat map. One of the device’s most distinctive features, Grespan notes, is that it operates independently of the telecommunications networks that cover rural properties. “All the data is uploaded to the cloud and the platform is available online 24 by 7,” he says. Telephone and internet coverage is patchy in Brazil, especially in rural or remote areas.

Biological pest control

Students at the University of Campinas (UNICAMP) began developing the invention three years ago. The device and accompanying software are designed to help farmers cut the cost of pest control. “Continuous monitoring is a major challenge for farmers because of its cost,” Grespan says. This is particularly the case for soybean growers, whose profit margins are slender, he adds.

Besides soybeans, the system is being tested on cotton plantations, where S. frugiperda can cause significant crop losses, second only to the damage done by the Boll weevil (Anthonomus grandis).

“We’re currently partnering with AgBiTech [an Australian-American supplier of large-scale biological control solutions] on experimental testing of a caterpillar monitoring service using our equipment,” Grespan says.

Testing on soybeans and cotton is being conducted on farms in Querência, Sapezal, Diamantino and Primavera do Leste, Mato Grosso State, and Luiz Eduardo Magalhães and Barreiras, Bahia State.

“The system detects moths that have yet to mate, before the emergence of the caterpillars that are the actual predators on these crops, so there’s time to take protective measures,” Grespan says.

Automatically mapping the types of insects that are on the plantation and estimating their quantity gives growers important decision-making support, he adds.

The conventional monitoring process, often also using traps with pheromones but without electronic equipment of any kind, is much more laborious and inefficient, according to the firm’s researchers. Manual inspection, usually carried out once a week, requires the use of personnel in the field, and is a subjective, error-prone process. In addition, crop pest control cannot be tightly managed if data is not collected every day, Grespan insists.

Tarvos will also be responsible for developing the sensors installed in the traps. The system dates from the first projects in the history of the firm. “From the word go we set out to use computing and image processing,” Grespan recalls. “When we understood that pests are a really serious problem for farmers, we decided to take this path as well.”