CALLISTO: Copernicus Artificial Intelligence Services and Data Fusion


Arti­fi­cial Intel­li­gence (AI) is already part of our lives and is exten­sive­ly enter­ing the space sec­tor to offer val­ue-added Earth Obser­va­tion (EO) prod­ucts and ser­vices. Coper­ni­cus data and oth­er geo­ref­er­enced data sources are often high­ly het­ero­ge­neous, dis­trib­uted and seman­ti­cal­ly frag­ment­ed. Large vol­umes of satel­lite data (images and asso­ci­at­ed meta­da­ta) are fre­quent­ly com­ing to the Earth from Sen­tinel con­stel­la­tion, offer­ing a basis for cre­at­ing val­ue-added prod­ucts that go beyond the space sec­tor. The analy­sis and data fusion of all streams of data need to take advan­tage of the exist­ing DIAS and HPC infra­struc­tures, as well as the Galileo-enabled mobile devices when required by the end users involved to deliv­er ful­ly auto­mat­ed process­es in deci­sion sup­port systems.

CALLISTO project inte­grates Coper­ni­cus data, already indexed in DIAS plat­forms such as ONDA-DIAS, uti­liz­ing High Per­for­mance Com­put­ing infra­struc­tures for enhanced scal­a­bil­i­ty when need­ed. Com­ple­men­tary dis­trib­uted data sources involve Galileo posi­tion­ing data, visu­al con­tent from UAVs, web and social media data link­ing them with open geospa­tial data, in-situ sen­sor data. On top of these data sources, AI meth­ods are applied to extract mean­ing­ful knowl­edge such as con­cepts, changes, activ­i­ties, events, 3D mod­els, videos and ani­ma­tions of the user com­mu­ni­ty. AI meth­ods are also exe­cut­ed at the edge, offer­ing enhanced scal­a­bil­i­ty and time­ly ser­vices. The analy­sis of the extract­ed knowl­edge is per­formed in a seman­tic way and the asso­ci­at­ed ana­lyt­ics are deliv­ered to the end users in non-tra­di­tion­al inter­faces, includ­ing Aug­ment­ed Real­i­ty, Vir­tu­al Real­i­ty and Mix­er Real­i­ty in gen­er­al. Data fusion among sev­er­al types of data sources is pro­vid­ed on-demand, based on the end user require­ments. The AI ​​meth­ods are trained to offer new vir­tu­al and aug­ment­ed real­i­ty appli­ca­tions to water util­i­ty oper­a­tors, jour­nal­ists for the media sec­tor, EU agri­cul­ture and CAP pol­i­cy­mak­ers, and secu­ri­ty agencies.

The CALLISTO team in InfAI is respon­si­ble for seman­ti­cal­ly indexed and enriched data gen­er­at­ed with­in oth­er mod­ules. To this end, we aim at using advanced seman­tic pro­cess­ing frame­works such as the Seman­tic Ana­lyt­ics Stack (SANSA). SANSA con­sists of var­i­ous ded­i­cat­ed lay­ers to per­form SPARQL query­ing, infer­ence and machine learn­ing ana­lyt­ics. Exist­ing com­po­nents will be extend­ed to pro­vide opti­mized big data pro­cess­ing capa­bil­i­ties for data con­tain­ing geospa­tial information.

Dura­tion of the project: 01/2021 – 12/2023
Grant Agree­ment No.: 101004152


This mea­sure is co-financed by the Else Krön­er-Fre­se­nius Cen­tre for Dig­i­tal Health (EKFZ).

This project has received fund­ing from the Euro­pean Union’s Hori­zon 2020 research and inno­va­tion pro­gramme under grant agree­ment No. 101004152.


You can see all project part­ners here.