Objectives of the service
SREC addresses the growing need for trustworthy and scalable verification of soil management practices. Many organisations rely on manual audits or farmer self-reporting, which are costly, inconsistent and difficult to validate. This creates significant challenges for regenerative agriculture programmes, insurance risk assessment and environmental compliance systems that all require objective and repeatable evidence.
The SREC service delivers a satellite-based tillage detection layer generated from high-resolution Earth Observation data enhanced through AI. The system identifies soil disturbance patterns and tillage events across large regions and provides results through an API that integrates directly into user platforms. This enables continuous, independent monitoring without the need for field inspections.
The service helps regenerative agriculture platforms verify sustainable practices, supports insurers in evaluating risk, enables carbon and sustainability platforms to strengthen monitoring and reporting, and gives Paying Agencies a reliable source of evidence for compliance monitoring. By combining open satellite data, super-resolution image processing and local training datasets, SREC creates a scalable and transparent tool for environmental claim validation.
Users and their needs
SREC serves a range of users that require reliable and independent information on soil and land management practices. The primary users include regenerative agriculture platforms, insurance and AgTech scoring systems, carbon MRV platforms and Agricultural Paying Agencies. These groups need consistent, high-resolution evidence of tillage events and soil disturbance that can be used in certification, risk modelling, compliance checks and environmental reporting.
Key user needs include:
-
Independent verification not based on self-reported data
-
High-resolution detection suitable for small and fragmented fields
-
Scalable monitoring across thousands or millions of hectares
-
Reliable time series to evaluate seasonal and annual changes
-
Simple integration through API delivery
-
Regional adaptation for different soil types and cropping systems
The project’s initial targeted users are located in Germany and Canada, with applicability across Europe and North America as the service scales.
Service/ system concept
The SREC tillage detection service provides independent, AI-powered verification of soil management practices using satellite imagery. The system delivers binary tillage detection (tilled/not-tilled), event timing, and confidence scores at field level, enabling customers in regenerative agriculture, insurance, carbon MRV, and regulatory compliance to replace costly manual audits with scalable, objective monitoring.
Key Capabilities:
-
Automated tillage event detection with dates and confidence levels (target >80% accuracy)
-
1m Deep Resolution imagery enabling detection on small fields (<2 ha)
-
Multi-index analysis (bare soil indices, soil moisture, vegetation change)
-
API-first delivery via secure HTTPS endpoints with token authentication (GeoJSON, GeoTIFF formats)
-
Regional adaptability through local fine-tuning
The system ingests Sentinel-2 optical and Sentinel-1 radar satellite data covering agricultural fields for a selected time span. The time-series engine calculates soil condition indices (MBSI, BSI, NDTI) as well as vegetation indices and detects temporal anomalies indicating tillage events. A classification module validates candidate events using pattern recognition and segmentation, applied on Deep Resolution imagery (from 10m to 1m resolution). Results are delivered via REST API as field-level GeoJSON outputs with event dates and confidence scores, ready for platform integration.
Space Added Value
SREC relies on Earth Observation satellites, primarily Sentinel-2 optical imagery, enhanced through DigiFarm’s Deep Resolution processing to achieve effective 1-meter detail. This improvement allows the system to detect soil disturbance and tillage patterns that are not visible at standard 10-metre resolution. The service also benefits from high revisit frequency, providing consistent temporal coverage needed to monitor seasonal and regional variations in soil management.
The added value of using satellite assets is the ability to independently observe large agricultural areas without relying on manual field inspections or self-reported data, both of which are costly, inconsistent and difficult to validate. Satellite-based monitoring ensures uniform methodology, repeatability across regions and the capacity to cover millions of hectares efficiently.
By combining satellite imagery with AI and local training data, SREC creates a scalable verification tool that is significantly more transparent and objective than existing approaches. Competitors that rely solely on raw imagery or modelling cannot match the combination of high-resolution detection, interpretability and independence achieved through this integrated space-enabled approach. This makes satellite assets essential for delivering reliable, cost-effective and verifiable environmental claims at scale.
Current Status
Completed Achievements:
-
Processed 60 Lithuanian fields (2023-2025) with 3553 satellite observations as proof-of-concept dataset
-
Successfully adapted time-series methodology from crop emergence analysis to tillage detection, achieving 87.3% mean confidence across all test fields
-
Built cloud-native processing pipeline handling multi-temporal Sentinel-2/1 data
-
Implemented multi-index soil analysis (MBSI, BSI, NDTI) with automated event detection using threshold, derivative, and inflection methods
-
Secured €80,000 contract for 2026 deployment covering 2 million hectares
Currently In Progress:
-
Regional fine-tuning for 50 Canadian fields and German pilot fields
-
API endpoint development and integration specifications
-
Binary classification workflow (tillage/no tillage) further testing and event confirmation based on AI segmentation on Deep Resolution imagery
Prime Contractor(s)
