The Peruvian startup TrazAr combines artificial intelligence, IoT sensors, and satellite images to alert coffee growers about pests and comply with the EU's anti-deforestation regulations, all without the need for an internet connection.
In a country where only 20.5% of rural households have internet access, a Peruvian startup has achieved what seemed impossible: bringing artificial intelligence to the field without the need for a connection. TrazAr, an AgriTech and Climate Tech company, has developed a system that combines weather stations, IoT sensors, and satellite images to protect coffee and cocoa crops, while also helping farmers comply with the new European regulations against deforestation.
Early warnings for pests without connection
Coffee is Peru's main agricultural export product, but climate change has increased the incidence of pests such as yellow rust and the coffee borer. In 2013, rust affected half of the cultivated area and destroyed 80,000 hectares. In response to this threat, TrazAr has deployed its own network of low-cost weather stations that measure temperature, humidity, precipitation, and solar radiation in real time.
The data is cross-referenced with artificial intelligence algorithms that detect when climatic conditions reach the exact thresholds that favour the emergence of pests. The system then sends alerts to farmers so they can take preventive measures before losing their harvest. All of this works without the internet: technicians download maps in urban areas, travel to remote farms, collect data, and the application synchronises automatically when connectivity is restored.
A digital passport for exporting to Europe
In addition to climate issues, Peruvian coffee growers face a significant trade barrier. The European Union, the main destination for their exports, requires that from December 30, 2025, all coffee and cocoa entering its market must prove that it does not come from deforested areas. This is the EUDR Regulation, and those who do not comply will be excluded.
TrazAr acts as a digital passport. It georeferences the cultivation polygons and, through historical satellite images, verifies whether there has been deforestation in recent years. "I have to prove that the products I am selling to the European Union do not come from deforested areas," explains Wilmer Manay, co-founder of TrazAr, in statements to the Andina agency. With this tool, the farmer can certify their compliance without the need for costly audits.
The system also allows cooperatives to customise technical assistance through a "technology maturity radar" that classifies each producer according to their level of progress. "Each producer is a world; each producer has their level of advancement," says Manay, which allows for tailored training efforts and planning for future integrations via WhatsApp.
From alerts to precision agriculture
Beyond early warnings and regulatory compliance, TrazAr delves into precision agriculture. It uses drone flights to analyse photosynthetic activity, chlorophyll levels, and plant biomass. This information, shared with the banking system, reduces perceived risk and operational evaluation costs, which could facilitate access to credit for small producers.
The startup, co-financed by ProInnóvate, was born to respond to the reality of the Peruvian countryside: an ageing, disconnected, and vulnerable family farming sector. "Climate change is affecting crop behaviour. There is low genetic quality and limited access to technologies in the rural sector," summarises Manay. With its offline solution, TrazAr demonstrates that innovation does not need the internet to reach the most remote places.
For Peruvian coffee growers, the tool represents a double benefit: protecting their crops from climate impacts and ensuring their access to the European market, all from a mobile phone that works even in the mountains. The next step, according to its creators, is to integrate the system with messaging channels like WhatsApp to further facilitate adoption among less technologically advanced producers.
You may also like
