Objectives of the service
The PerPlant Sensor System addresses the operational and financial challenges faced by small to medium-sized arable farms, particularly inefficient chemical usage and high technology adoption barriers. Traditional precision farming systems often carry premium price tags that remain out of reach for operations managing between 100 and 700 hectares.
The primary objective of the project was to validate an affordable, plug-and-play retrofit system capable of providing plant-level care. By fusing real-time camera vision with satellite historical trends and high-accuracy positioning, the service optimizes variable-rate and spot-spraying applications. The objective was fully met during the 2026 operational pilot phase, delivering validated input reductions of 70% on herbicides, 45% on fertilizers, and 35% on fungicides, while increasing overall crop yields by up to 10%.
Users and their needs
The primary user community consists of small to medium-sized arable farmers managing 100 to 700 hectares of land within Europe, alongside professional plant advisors and precision agriculture equipment resellers.
Through intensive user field research, the project identified five core operational needs:
- Cost-Effective Retrofitting: Solutions must easily integrate with existing tractors and sprayers without requiring massive capital machinery investments.
- Network Independence: Robust offline functionality is essential to preserve system precision in remote rural fields lacking reliable mobile internet coverage.
- High Operational Accuracy: Centimeter-level application is required to ensure spot-spraying efficacy under dense canopies or complex layouts.
- 24/7 Functionality: The capability to perform high-accuracy weed detection during nighttime spraying or low-light, cloudy conditions.
- Simplicity and Clarity: A clear, intuitive interface that removes technical complexity and builds operator trust in automated AI adjustments.
Service/ system concept
The PerPlant Sensor System utilizes a sophisticated multi-component architecture that decouples localized edge execution from cloud processing to ensure total reliability in the field.
The system operates across three core layers:
- On-Edge Sensing & Inference: Tractor-mounted RGB and multispectral cameras capture live canopy imagery. An on-board edge-AI computing node processes this visual data instantly, executing weed detection and plant health classification locally on the machine without requiring a cloud connection.
- Space Asset Fusion: The edge data is continuously combined with Copernicus Sentinel satellite data to evaluate historical vegetation indices (NDVI) and zone productivity. Positioning is secured via multi-constellation GNSS integrated with RTK correction networks to maintain a strict ±10–15 cm accuracy baseline.
- Terminal Execution: The system calculates real-time variable-rate instructions and streams them directly to the implement via standard ISOBUS protocols, maintaining application rates within ±5% of the target prescription. Farmers review comprehensive operational field maps through the cloud-based PerPlant Console SaaS dashboard post-operation.
The system architecture (on a high level)
Space Added Value
The "Space Added Value" is the core differentiator of the Agronomy on the Edge platform. Rather than relying solely on localized camera views, the system uses a dual-layer data fusion approach.
Positioning relies on multi-constellation GNSS (including Galileo and GPS) paired with RTK base stations. This architecture enables an offline fail-safe mechanism: if a mobile network drop occurs in a remote area, the system maintains centimeter-level spraying precision uninterrupted.
Simultaneously, the continuous ingestion of Earth Observation data from Sentinel-1 and Sentinel-2 satellites via the Copernicus DataSpace Ecosystem provides historical crop health mapping at a 10-meter resolution. This allows the system to ground-truth live camera data against multi-year productivity baselines, optimizing chemical applications even in changing weather conditions or adverse low-light environments.
The Maps above shows a prescription map based on historical satellite data and next to it sensor map and all layers of the PerPlant system and will have a prescription map combining both the layer and make precise agronomy insights through Agronomy On the Edge.
Current Status
The project is successfully completed, reaching full Technology Readiness Level 9 (TRL-9). Following successful Factory Acceptance Testing (FAT) and Site Acceptance Testing (SAT), all system verification requirements have been met. Extensive operational pilots were executed across sites in Denmark, the United Kingdom, and the Czech Republic during the 2026 spraying season.
The commercial feasibility of the business model has been validated through immediate market traction, resulting in direct sales and leasing conversions of pilot units during the project timeframe. Commercial distribution is now established via authorized dealer networks, including Ag Precision in Denmark and Leading Farmers CZ in Central Europe.
Image 1 - Latest satellite map
Image 2 - Historical satellite map - productivity map
Image 3 - Combined sensor + satellite map - Prescription map
Prime Contractor(s)
