ChatGPT for agriculture? Brazilian startup launches AI for agribusiness powered by autonomous drones
- Emilly Ketzer
- Apr 8
- 3 min read
Psyche Aerospace, a Brazilian startup based in São José dos Campos, launched Turing on Tuesday (29), an artificial intelligence (AI) platform called the ‘ChatGPT of agriculture’ by its developers.
While the original tool, developed by OpenAI, works only with public data, Turing aims to feed the database with information collected by drones manufactured by the company.
“Our proposal is to enable an integrated ecosystem, in which drones, field sensors, soil analysis and mapping communicate with each other, generating qualified data that rural producers can interpret and use immediately”, says Gabriel Leal, CEO of Psyche.
At an event held in São Paulo, the company detailed plans for the future and inaugurated the series B round, seeking to raise funds for the construction of a new factory, capable of producing drones on a large scale. The project is expected to cost around R$100 million.
Psyche Aerospace's drones
At just 24 years old, Leal founded Psyche in 2022 and two years later managed to raise around R$15 million in its first round of investment. The funds were used to produce the company's first drones for agribusiness.
One of them is the Harpia P-71, presented as the largest spraying drone in the world for agricultural use. According to the CEO, the drone is capable of operating around 40 hectares per hour, carrying 400 kg of agricultural pesticides.
The equipment is also autonomous, without the need for someone to operate it, and it even refuels itself at a mobile refueling station called “Beluga”.
In addition to the Harpia P-71, the company has developed five other models, three drones and two ground robots. Each one has different specifications to capture more precise data and forward it to the Turing database, Psyche's AI.
The CEO states that the equipment will not be sold, but rather managed by Psyche itself, which intends to build bases in various regions of the country, getting closer to rural producers. Turing, the 'ChatGPT of agriculture', should be made available to producers free of charge.
“For producers, they will basically register their farm on the website and when we open a base in their region, our team will send a notice saying that Turing is available in the region”, says CEO Gabriel Leal.
The mapping and spraying performed by drones will be offered at R$25 per hectare, in a five-year contract, with 25% advance payment in the first year. According to Leal, the price is the lowest on the market.
Turing, the AI of agribusiness
According to Psyche, Turing's proposal is to offer real-time diagnostics and recommendations, capable of optimizing resources and increasing productivity. A version of the tool is already available for download in app stores, but it doesn't have much information yet. The most complete mapping should be available in six months.
“The big difference between Turing and ChatGPT is that GPT was developed for general use, while Turing will have a specialty. It will adapt to you (the producer), to your problem, in the context of Brazilian agribusiness”, says Anderson Rocha, director of the AI laboratory at Unicamp and consultant at Psyche.
In the app, it is possible to view information about a sugarcane plot, which is part of a demonstration field of the Tereos Group, located in Baguaçu, in the city of Olímpia. The tool offers suggestions such as the need for irrigation adjustments or checking soil nutrients in certain areas.
According to Rocha, the idea is that the data captured will be of increasingly higher quality when compared to images captured by satellite. With more precise reading of problems in the field and the use of drones in spraying, Psyche estimates a productivity gain of 100% to 150% in agricultural activities.
“The big difference is in the longitudinal collection, from one year to the next, from one harvest to the next. Last year, you collected the data and know exactly what happened, detecting the problems and working with suggestions for improvement. This year, you can know exactly how much you improved,” says Rocha.
The consultant emphasizes that the productivity gain will come with the continuous use of the data provided by the tool. “So, if, for example, you improved production from one year to the next by 10%, you have a much greater chance of improving from the second year to the third, because you learned everything that worked and saw what didn’t work so well,” he explains.
Commentaires