#	Pagina
attuale pagina	/open-h2020/projects/212163/results.html

Report

Teaser, summary, work performed and final results

Periodic Reporting for period 2 - WATER4AGRI (Securing water for food and safety with the world\'s most advanced soil moisture information derived from satellites)

Teaser

Summary

European agriculture is facing important challenges in the coming decades linked to our changing climate. Periods of droughts and floods will further increase which threaten production and require a smart usage of our vulnerable water resources. Climate change also stimulates crop disease. Satellite Remote Sensing has a crucial role to play in tackling these challenges. At VanderSat we have developed a method to provide accurate high-resolution images of soil moisture, at any place on Earth. Soil moisture is a key indicator for the physical status of a plant, and has a strong predictive value with respect to crop yield, floods and droughts. This data is crucial for improving the accuracy of Agricultural models that are the backbone of modern agribusiness.

Three important goals are the heart of our WATER4AGRI project. To develop a scalable software solution that will facilitate short lead times for future implementation of VanderSat in Ag models. Next, we demonstrate the effectiveness of VanderSat soil moisture in two of the most widespread models in agricultural production. Hereby empirically demonstrating the predictive power of VanderSat data for customers. The last goal is to take the final steps in commercialisation.

The VanderSat data service is now built as a solution than can be deployed on different cloud based services for optimal scaling while remaining cost effective serving a large amount of clients. Both the viewer and the API are fully operational. Depending on the type of client data is streamed directly from API to the clientâ€™s systems or the data is being queried and displayed by the VanderSat viewer.

VanderSat data was integrated within two of the most widespread modelling frameworks: one model on plant disease and crop protection and one on drought modelling related to crop yield. Two Beta Programs were successfully executed with key players in the agri and insurance markets; Bayer and Swiss Re. The goal of both Beta Programs was two-fold. First, to establish the requirements for integration of our data in the models and to incorporate model specific adaptations in our APIs. A second goal was to show the value of VanderSat data with respect to crop growth monitoring and crop yield prediction. We successfully showed that for both Beta Programs all the criteria were met, although the results were strongly crop and region dependent. Both pilots resulted in follow-up contracts for an operational service.

By using our website in combination with social media we have managed to establish a significant on-line presence. Furthermore, articles on several websites targeted towards the agricultural and esurance business in combination with presentations on key symposia and interviews with local and global media we have further established the VanderSat brand. Our updated business plan is now focussed on a combination of direct delivery of our core data products to large agricultural and (re)insurance companies in combination with partnering with key players in the water sector. In addition, we have developed a couple of simple products on-top of our core data that are more easily used by customers that require an off-the shelf solution such as a drought indicator that can directly be linked to a payout scheme.

Work performed

We have fully implemented and operationalised our API, viewer and calculation system. DifSat, the patented core of our soil moisture algorithm, has been updated, improved and sped-up considerably. The DamSat operational satellite database has been enhanced and made suitable for a number of cloud services. Thus our system is now extremely scalable and operational costs have been reduced considerably. The system is now serving a number of clients operationally and the API has proven to be stable and scalable. The viewer is operational and is used by all clients. After moving our service to the Google Cloud Computing Engine we are now able to supply a full service to our clients at a low yearly fee. Furthermore our DifSat algorithm that delivers the high resolution in the VanderSat data has been patented.

In the first 9 months (phase 1) the main focus was on the execution of several customized Beta Programs; different test site locations in Europe, Canada and Brazil were defined, and a large database with ground data was compiled. The VanderSat satellite database was set up for the test site locations, and VanderSat satellite data products were then extensively tested using the scientific validation standards as demonstrated in several scientific publications. Besides showing the added value of VanderSat data for disease modeling, a goal of these-Beta Programs was to establish the requirements for integration of our data in several disease and crop yield modeling frameworks. The overall requirements did not differ much for our key customers, and based on these requirements the VanderSat Data Service was adjusted to ensure an optimal data integration. In Phase 2, covering the last 9 months of the project, the focus shifted to scaling up the VanderSat technology with a larger user base. The was reflected by several operational data service contracts that were set up with new customers, and the scaling with existing ones.

VanderSat has used the last nine months to test its go-to-market model and has started replicating this model across all verticals with dedicated teams for each of the market segments. We have launched a new website with a clear message at https://www.vandersat.com. Our website gets a steady 60 visitors per day (coming from 15 at the start of the project). When we do active PR our traffic reaches 300-500 per day. Our PR and marketing activities have been very successful which is demonstrated through the fact that so much of our traffic comes from social media and referrals. With 2523 external links (other websites linking to our website), over 17.000 individual people visiting our website and close to 27.000 sessions we are growing our online presence at a rapid pace.

Final results

We have set a new standard by being able to reliably supply near real time satellite derived soil moisture at high resolution for large areas with consistent quality. Using microwave sensors, we see through clouds and fog without the hassle of having to place thousands of ground sensors at high costs. Not only are we beyond the state of the art in the quality of our service, the product itself is superior to all other current offerings from the commercial or non-profit sector.

With our clients and partners we have been able to make new steps in translating the soil moisture data to actionable products that can determine basis risk for insurance as well as directly serve as a reliable payout index for a parametric insurance. The same data also feeds crop yield prediction models and systems that provided fertilizer advice. By teaming up with world leaders in the reinsurance and agricultural domain such as Swiss Re and BASF we have been able to make a large impact and scale more rapidly. VanderSat measures its impact also in ha of arable land for which its data is being used. At present we serve 121 million ha directly or via our partners (9.5 million at the start of the WATER4AGRI project). The type of impact on the ground differs per application and ranges from reduced insurance premium for a farmer to irrigation advice to small scale farmers.