Overview
DigiFarm is a Norwegian ag-tech company (founded in 2019) developing AI-powered solutions for precision agriculture through advanced satellite data analytics. By leveraging deep neural networks to super-resolve Sentinel-2 imagery (from 10 m to 1 m), the company delivers highly accurate, annually updated field boundaries and seeded area data, forming a critical foundation for decision-making across the agricultural value chain.
Its solutions serve farmers and advisors as well as B2B and B2G clients, including farm management software providers, insurers, financial institutions, commodity traders, public authorities, and IoT and robotics companies. The key market need addressed by DigiFarm is access to precise, up-to-date field data, which is often lacking in traditional and outdated cadastral systems.
To validate the accuracy and real-world applicability of its data, particularly for use in autonomous machinery and field robotics, DigiFarm required independent testing in operational farming environments. The agrifoodTEF Polish node provided the necessary infrastructure and conditions to conduct robust, real-world validation and demonstrate the reliability and machine-readiness of its solutions without requiring upfront investment in its own testing facilities.
Poland
The challenge
DigiFarm needed reliable, independent validation of its AI-driven field boundary data under real-world agricultural conditions. While its models performed well in analytical and laboratory settings, this was insufficient to demonstrate how accurately the data could be translated into executable paths and followed by autonomous machines in operational environments.
This validation was critical for both the company and its customers. Autonomous robots and IoT-based field operations rely on precise, up-to-date boundary data for safe and efficient navigation.
Proving that DigiFarm’s satellite-derived boundaries can be directly used by third-party systems and accurately executed in the field is key to unlocking new market opportunities in robotics and building trust in the solution’s real-world performance.
Without access to agricultural robots, test fields, calibrated sensors, historical data, and independent telemetry analysis, DigiFarm lacked the infrastructure to objectively measure performance and provide credible evidence to customers. Scaling such validation internally would have required significant time, cost, and investment.
The agrifoodTEF service
DigiFarm leveraged agrifoodTEF as a one stop shop providing coordinated access to field testing, data generation, AI validation, and advanced computing capabilities. All three polish partners worked in an integrated manner, ensuring consistency, efficient execution, and seamless collaboration throughout the project.
The support focused on enabling real-world validation of DigiFarm’s technology by translating its satellite-derived field boundaries into executable paths for autonomous systems and verifying their performance in operational conditions. This included end-to-end testing with an agricultural robot, covering data integration, system configuration, and execution of boundary-based navigation in a real field.
In parallel, agrifoodTEF provided high-quality agricultural datasets, historical field data, and expert knowledge to support model development and validation, alongside access to high-performance computing for large-scale processing and simulations. The testing activities combined field experimentation with detailed telemetry analysis, assessing path accuracy, system behavior, and deviations between planned and executed trajectories.
This comprehensive, multidisciplinary support enabled DigiFarm to validate both the analytical accuracy and practical usability of its solution, demonstrating that its data can be directly used by autonomous machines in real farming environments.
Testing and Data Collection of AI-Driven Robotic Solutions: learn more
The impact
The collaboration with agrifoodTEF provided DigiFarm with independent, real-world validation of its technology, confirming that its satellite-derived field boundaries can be seamlessly integrated and accurately executed by autonomous systems in operational conditions. Field tests demonstrated very high navigation precision, with minimal deviation between planned and actual robot paths, delivering quantified proof of performance that can be directly shared with customers.
This validation significantly strengthened DigiFarm’s market readiness, particularly in the rapidly growing segment of agricultural robotics and IoT. It proved that the company’s data is not only analytically accurate but also operationally reliable, enabling real-world applications such as autonomous navigation and geofencing.
At the same time, the testing process identified concrete product improvements, further streamlining future integrations and scalability.
A key added value for the customer was agrifoodTEF’s role as an integrated, one-stop testing ecosystem. The coordinated support, from access to real test fields and datasets, through AI validation, to advanced computing and performance analysis, allowed DigiFarm to move efficiently from concept to verified solution without engaging multiple providers. This significantly reduced time, cost, and complexity.
As a result, DigiFarm has continued its cooperation with agrifoodTEF beyond the initial project, treating it as a strategic partner for ongoing validation, optimisation, and demonstration of new data products and use cases. This provides a scalable and repeatable pathway to support further innovation, expansion into autonomous machinery, and entry into new markets.
