Sentinels-based Agriculture Information Service Component

Información del proyecto

SAGRIS

Identificador del acuerdo de subvención: 782116

DOI

10.3030/782116

Proyecto cerrado

Fecha de la firma de la CE 14 Julio 2017

Fecha de inicio 1 Agosto 2017

Fecha de finalización 31 Enero 2018

Financiado con arreglo a

• INDUSTRIAL LEADERSHIP - Leadership in enabling and industrial technologies – Space

Coste total

€ 71 429,00

Aportación de la UE

€ 50 000,00

Coordinado por

GEOMATRIX UAB
Lithuania

Periodic Reporting for period 1 - SAGRIS (Sentinels-based Agriculture Information Service Component)

Período documentado: 2017-08-01 hasta 2018-01-31

Resumen del contexto y de los objetivos generales del proyecto

Due to lower technological advance, developing agricultural markets offer a great opportunity for simple and affordable, yet personalised and efficient B2B and B2C smart farming information services and solutions, focused on small and medium-size traditional or specialised ecological farms, as well as advancing agricultural enterprises looking for smooth transition from traditional farming towards efficient PA practices with balanced investment and return rates. The EC/ESA Copernicus program with its Sentinel-1 active radar sensors provides enabling technology in the form of continuous time-series of SAR imagery independent of cloud coverage in central and northern Europe. SAGRIS will pursue an objective to develop an affordable, problem-centric, personalised and operational smart farming decision support service focused on early warning and advisory support capabilities based on artificial intelligence and data mining. It will offer savings and higher productivity without investment into PA hardware solutions and will be provided on a large scale as a stand-alone solution, but also integrated into the existing smart farming services.

Trabajo realizado desde el comienzo del proyecto hasta el final del período abarcado por el informe y los principales resultados hasta la fecha

It is feasible to develop automated cloud-based processors to transform EO data into object-based signal statistics, further analysed with complex AI algorithms and combined into a multi-purpose back-end service. A complete time-series of Sentinel-1 SAR data calibrated with cloud-free Sentinel-2 MSI will provide frequent and regular assessments of soil conditions and status of crops along with added-value trend analysis, operational mapping, farming recommendations and assessment of damage or loss of crops. SAGRIS business development plan is targeting a large market of ~1 mln. holdings and 44.2 mln. ha of arable land in 15 EU countries, which could generate up to € 100 mln. annually. Service development and initial expansion plan shows steady growth of profits during the initial 5-year period in local markets of 15 targeted EU countries reaching annual net profit of € 2.2 mln., thus proving economical feasibility of the project.

Avances que van más allá del estado de la técnica e impacto potencial esperado (incluida la repercusión socioeconómica y las implicaciones sociales más amplias del proyecto hasta la fecha)

SAGRIS modules are designed as independent blockchain type processors deployed on virtual machines for large-scale automated processing of BIG EO Data time series on DIAS cloud-based infrastructure. Flexibility of the block-chain production work-flow will optimise the processing costs and reduce processing time, enable service scalability through regional service centres and secure sensitive personal data. The ultimate game-changing result of this service will be transition from petabytes of satellite imagery towards time-series of object-based statistical samples and further consolidation into business intelligence, pattern/process analysis and early warning based on problem-centric deep learning and artificial intelligence algorithms.

A combined time series of Landsat-8 and Sentinel-1 SAR images confirms grass mowing date

SAGRIS promotional publication Pan European Networks, Government 23, page1

Damaged winter crops (outlined in red) seen on Sentinel-1 dynamic composite map

Sentinel-1 dynamic crop map (left) and vegetation change map (right)

Sentinel-1/2 image pre-processing work-flows

Sentinel-1 used for assessment of crops damage done by wild animals

Total annual coverage of Sentinel-1 (left) and Sentinel-2 (right) images in Lithuania, 2017

SAGRIS promotional publication Pan European Networks, Government 23, page2

Flooded crops in summer 2017 in Lithuania, as seen on Sentinel-1 time series