If you are an agricultural technology company struggling to turn raw satellite and drone imagery into actionable crop health reports — this project developed open-source software for extracting plant traits and photosynthetic efficiency from Sentinel 2 and Landsat 8 data. The tools process hyperspectral time series at daily and seasonal scales, letting you detect stress before it becomes visible to the eye. With 8 demonstrated software deliverables and UAV flight planning procedures already built, you could integrate these into your existing platform.
Open-Source Satellite and Drone Tools That Turn Raw Imagery Into Vegetation Health Maps
Imagine you own thousands of hectares of farmland or forest and you need to know which areas are stressed, drying out, or losing productivity — but you can't walk every field. This project built open-source software that takes raw satellite and drone images and translates them into detailed maps of plant health, water use, and photosynthetic efficiency. Think of it like giving ecosystems a medical check-up from space, combining data from multiple satellites and drones into one clear picture. The tools work with freely available data from ESA Sentinel and NASA satellites, so the raw input costs nothing.
What needed solving
Companies managing large areas of vegetation — farms, forests, nature reserves, utility corridors — need to detect plant stress, drought damage, and productivity changes before they become costly problems. Traditional ground surveys are slow, expensive, and cannot cover enough area. While satellite and drone imagery is increasingly available for free, turning raw spectral data into reliable vegetation health indicators requires specialized processing tools that most companies don't have in-house.
What was built
The project built open-source software for processing satellite imagery (Sentinel 2, Landsat 8) and drone-based hyperspectral data into maps of plant health, photosynthetic efficiency, and biodiversity indicators. Concrete outputs include time series decomposition tools, UAV flight planning procedures with operational protocols, fluorescence retrieval algorithms, statistical upscaling methods, and data assimilation models — all released under GNU licence across 31 deliverables.
Who needs this
Who can put this to work
If you are an environmental consultancy that needs to document ecosystem health changes over time for regulatory reports — this project created statistical methods for upscaling biodiversity variables from remote sensing data across landscapes. The open-source tools process FLUXNET eddy covariance data to derive water use efficiency and light use efficiency metrics. With 13 consortium partners across 7 countries having validated these methods, you get peer-reviewed tools ready for professional use.
If you are a satellite data company looking to expand your product line with vegetation analytics — this project built open-source models combining Sentinel 2 and Landsat 8 data assimilation for retrieving plant traits and ecosystem functional properties, published under GNU licence. The UAV-based hyperspectral processing tools and time series decomposition software let you offer multi-scale vegetation monitoring from ground level to satellite. With 4 industry partners already involved in development, the tools were designed with commercial workflows in mind.
Quick answers
What would it cost to use these tools?
All software deliverables are released as open-source under GNU licence, so there is no licensing fee. Your costs would be integration, customization, and the computing infrastructure to run the processing pipelines. The satellite data from Sentinel 2 and Landsat 8 is freely available from ESA and NASA.
Can these tools work at industrial scale across large territories?
The project specifically developed statistical methods for upscaling from plot-level to landscape-level using multi-satellite data assimilation. The tools are designed to process time series at daily, seasonal, and interannual scales. However, scaling to continent-wide operations would require significant computing resources and validation for your specific geography.
Who owns the intellectual property and can I build commercial products on top?
The software is published under GNU licence, which means you can use, modify, and distribute it. However, GNU GPL requires that derivative works also be open-source, so if you want to build proprietary products, you would need to verify the specific GNU variant used and potentially negotiate with the coordinating university (Università degli Studi di Milano-Bicocca).
How accurate are the vegetation health measurements?
Based on available project data, the project systematically compared different measurement modes — ground-based spectrometers, UAV-mounted instruments, and airborne hyperspectral imagers — to quantify accuracy across scales. The fluorescence datasets were validated across plots with different water and nutrient conditions. Specific accuracy percentages are not stated in the available deliverable descriptions.
What satellite systems does this work with?
The tools are built for ESA Sentinel 2, Landsat 8, and are designed to exploit upcoming missions including ESA-FLEX, additional ESA-Sentinels, and NASA-GEDI. The multi-source data integration approach means adding new satellite feeds is part of the architecture.
Is there ongoing support or is this a finished research project?
The project closed in September 2020, so there is no active funded support. However, the open-source repositories remain available, and the 13-partner consortium across 7 countries trained a generation of researchers now working in this field. The coordinator at Università degli Studi di Milano-Bicocca may be able to direct you to follow-up projects.
How long would integration into our existing platform take?
Based on available project data, the tools were developed as standalone open-source packages, not plug-and-play commercial products. Integration would depend on your existing tech stack. The UAV flight planning procedures and standardized data acquisition protocols are operational and documented, which should reduce setup time for the drone-based components.
Who built it
The TRuStEE consortium brought together 13 partners from 7 countries (Belgium, Germany, Spain, France, Italy, Netherlands, UK), with a healthy mix of 5 universities, 4 research organizations, and 4 industry partners including 3 SMEs. The 31% industry ratio is notable for a training network, suggesting the tools were developed with real-world applications in mind rather than purely academic exercises. The coordinator, Università degli Studi di Milano-Bicocca in Italy, is a strong research university. The geographic spread across Western Europe means the tools have been tested across diverse climate zones and vegetation types, which adds credibility for companies operating in multiple European markets.
- UNIVERSITA' DEGLI STUDI DI MILANO-BICOCCACoordinator · IT
- VLAAMSE INSTELLING VOOR TECHNOLOGISCH ONDERZOEK N.V.participant · BE
- THE UNIVERSITY OF EXETERparticipant · UK
- FORSCHUNGSZENTRUM JULICH GMBHparticipant · DE
- UNIVERSITEIT TWENTEparticipant · NL
- TRILOGIS SRLpartner · IT
- AGENCIA ESTATAL CONSEJO SUPERIOR DE INVESTIGACIONES CIENTIFICASparticipant · ES
- FONDAZIONE EDMUND MACHparticipant · IT
- AEROVISION BVparticipant · NL
- MAX-PLANCK-GESELLSCHAFT ZUR FORDERUNG DER WISSENSCHAFTEN EVparticipant · DE
- THE UNIVERSITY OF EDINBURGHpartner · UK
Coordinator is at Università degli Studi di Milano-Bicocca (Italy). SciTransfer can help identify the right contact person and facilitate an introduction.
Talk to the team behind this work.
Want to know if these open-source vegetation monitoring tools fit your use case? SciTransfer can arrange a technical briefing with the research team and help you evaluate integration options.