THRUST-4RESST
THRUST-4RESST is a multifunctional forest condition assessment tool that is based on AI-aided analytics of time-resolved satellite data

Covering more than 38% of land area within the European Union, forests are a vital component of Europe's natural environment, providing a range of ecological, social, and economic benefits. However, forestry resources are being continuously challenged due to deforestation, spread of diseases, damage by wildfire, floods, storms, and other threats. Though prevalent, manual maintenance & control processes cause serious time lag in decision making, therefore most up-to-date countrywide information is required.
To this end, we have developed THRUST-4RESST – Forest REmote Sensing via Satellite Technology – a multifunctional forest condition assessment tool, which is based on AI-aided analytics of time-resolved satellite data that can be combined with ultra-high-resolution imagery collected by UAVs and manned aviation.
With THRUST-4RESST, we aim to bridge the gap between high-resolution, low-frequency aerial data and low-resolution, high-frequency satellite data to create an efficient decision support system for private and public foresters.
THRUST-4RESST system detects deforestation, fire, flood, or storm damage on a country-wide scale efficiently, which helps to apply control measures without delays.

THRUST-4RESST relies on change detection in Sentinel-2 MSI data and classification of the detected changes based on their spectral footprint. Due to the universality of the change detection approach, THRUST-4RESST can be employed to analyze most of man-made and natural threats to forest resources and after customization would be able to perform:
- Detection of deforestation
- Assessment of damage by storms, identification of windfall
- Detection of waterlogged, flooded areas
- Assessment of damage by wildfire
- Detection of pest infestations
- Detection of illegal construction work
- Detection of landfill
- Classification of tree species or type (e.g., coniferous vs. deciduous)
- Identification of invasive species.

One of the key use cases of THRUST-4RESST is detection of deforestation in large areas. Due to the highly optimized analytics approach, THRUST-4RESST can present a deforestation map for entire country within minutes. The analytics system takes pairs of Sentinel-2 MSI data from different time points as an input, and performs change detection followed by change classification to indicate clear-felling zones in forested areas. The identified clear-felling patches are then vectorized and reported as a GIS polygon vector layer with relevant attributes. Thus, THRUST-4RESST solution is based on edge computing principles, as no surplus data transfer is executed. The validity of the approach has been tested by comparison with high-resolution country-wide LIDAR data.
The output format can be adjusted and tailored to the client needs. Please contact us for the demonstration of THRUST-4RESST functionality for your custom use case.
This sub-project has received funding from the European Union’s Horizon 2020 research and innovation programme within the framework of the AI4Copernicus Project funded under grant agreement No 101016798.
