A Croatian digital services firm has used data from an ESA Third Party Mission (TPM) to transform its vision of a next-generation green space management tool into commercial reality.

Developed by 3D EMS, E2O.GREEN employs artificial intelligence and remote sensing to generate actionable insights for the maintenance of grassland areas, such as golf courses, sports pitches and urban parks.

Bridging the gap

Dario Vuljanko, remote sensing specialist and chief executive of 3D EMS, says, “We created the E2O.GREEN platform to bridge the gap between Earth observation and sustainable action for managers of green spaces in or near urban areas.

“Early on in the development process, we identified subscriptions from golf course managers as a potentially important revenue stream – the platform was fine-tuned with this purpose in mind, but it can also be applied to a range of other use cases.”

Backed by the AI4Copernicus initiative, E2O.GREEN uses machine learning algorithms trained on remote sensing data and drone imagery to track the health of green spaces and identify areas that require irrigation or other interventions.

The project drew on Copernicus Sentinel-2 data, available on a free and open basis, and PlanetScope data, delivered via ESA’s TPM programme as part of its commitment to supporting entrepreneurs and start-ups across Europe.

Advanced AI algorithms

Satellite and drone imagery collected over the area of interest across different time points are first fed into the system, before being pre-processed by applying cloud masking and resolution adjustments.

Using advanced algorithms, this imagery is analysed to calculate Normalised Difference Moisture Index (NDMI) values, which represent levels of vegetation moisture. An NDMI change map is generated to illustrate how vegetation moisture evolves over time.

Artificial intelligence models developed by 3D EMS are then used to classify grassland conditions according to the NDMI change map, highlighting drought-stressed, over-irrigated and disease-affected areas.

E2O.GREEN delivers these insights to customers via an augmented reality-enabled mobile application, which overlays information on disease and irrigation status onto high resolution Earth observation imagery. This easy-to-understand system enables golf course managers to quickly identify problem areas and intervene to maintain the health of their greens and fairways.

Vuljanko comments, “The support of ESA’s TPM programme was crucial for the success of E2O.GREEN.

“The revisit cycle, sensor type, and spectral bands of PlanetScope made it a perfect connecting layer between lower resolution Sentinel sensors and very high resolution drone camera imagery. Together with the excellent alignment of the bands with the requirements of our computer vision models, this synergy enabled faster and more reliable monitoring of change detection and grassland classification.”

E2O.GREEN has continued to gain traction since its launch in 2022 and has now been implemented by more than 10 golf courses across Europe. As well as generating significant revenue, this uptake has promoted a more sustainable use of water resources by ensuring that irrigation is targeted to where it is needed the most.

3D EMS now plans to further scale up the product and expand its use to other green spaces.

As well as demonstrating the value of TPM data for growing companies, the project serves as an example of how ESA’s TPM programme supports activities within ESA Cooperating Sates (ECS), following Croatia’s designation as an ECS in 2023. 

New ventures

Vuljanko adds, “Building on the development of the E2O.GREEN augmented reality mobile application, we are now investigating a new business idea, catering to the gaming industry, that could also draw on ESA TPM data.   

“Earth observation imagery could provide developers with the possibility to build more realistic gaming environments – but this potential has yet to be fully unlocked. Our proposed new platform aims to address this challenge by empowering game developers to embed high resolution Earth observation data into the worlds of video games, as well as virtual reality experiences.”

3D EMS is currently in the early development stages of the platform and a prototype system is expected to be created in the near future to test out this new concept.