The Pro­ject PAIRS


P
rivacy-Aware, intel­li­gent and
Resi­li­ent Crisis Management

Pro­ject details

PAIRS

Iden­tify cri­sis sce­n­a­rios fas­ter with AI

The AI light­house pro­ject, initia­ted and coor­di­na­ted by Adva­neo and fun­ded by the BMWK with around €10 mil­lion, focu­ses on the deve­lo­p­ment of an AI-based data space for cri­sis manage­ment.
The PAIRS pro­ject (short for Pri­vacy-Aware, Intel­li­gent and Resi­li­ent Cri­siS Manage­ment) is deve­lo­ping a ser­vice-ori­en­ted, open data infra­st­ruc­ture that can be used to fore­cast the impact of cri­sis situa­tions. The AI hybrid tech­no­logy of PAIRS is inten­ded to anti­ci­pate both the initial cri­sis event and the reac­tions of various actors in a cross-domain data space in order to gene­rate tar­ge­ted recom­men­da­ti­ons for action on this basis.

Mar­ket per­spec­tive and pro­duct promise

Excep­tio­nal situa­tions such as the glo­bal coro­na­vi­rus pan­de­mic have shown how important it is for ope­ra­tors of cri­ti­cal infra­st­ruc­tures (e.g., health, energy, etc.) as well as poli­ti­cal actors (e.g., government insti­tu­ti­ons, NGOs, etc.) to quickly derive and effec­tively imple­ment tar­ge­ted mea­su­res from what is hap­pe­ning in cri­sis situa­tions. Accord­in­gly, the PAIRS rese­arch pro­ject pur­sues the deve­lo­p­ment of a cross-domain lear­ning plat­form for cri­sis manage­ment that com­bi­nes AI and human intel­li­gence. This AI hybrid approach will use machine lear­ning methods to enable the plat­form to iden­tify cri­sis sce­n­a­rios, dyna­mi­cally pre­dict their con­se­quen­ces, and recom­mend appro­priate actions to users. This will secure the avai­la­bi­lity of essen­tial resour­ces and ser­vices of eco­no­mic and orga­niz­a­tio­nal eco­sys­tems, sus­tainably streng­t­hen their mar­keta­bi­lity, and sup­port over­all socie­tal resilience.

Chal­len­ges and innovation

In order to effec­tively sup­port cri­sis manage­ment via AI, it is important that the app­li­ca­ti­ons to be deve­lo­ped are able to pre­dict the evo­lu­tion of cri­sis situa­tions and also anti­ci­pate the simul­ta­ne­ous reac­tions of the various actors (government, com­pa­nies, etc.). Only by taking into account the reac­tions to an initial cri­sis event in spe­ci­fic sce­n­a­rios will dyna­mic cri­sis manage­ment be pos­si­ble. The amount of data requi­red by the AI app­li­ca­tion to fore­cast such highly com­plex sce­n­a­rios is cor­re­spon­din­gly high. The lack of data that is com­mon in a Big Data sce­n­a­rio must be over­come and exis­ting data must be con­ti­nuously kept up to date while ensu­ring data pri­vacy and sov­er­eig­nty at all times. PAIRS sol­ves these chal­len­ges through a plat­form archi­tec­ture with fede­r­a­ted ser­vices that can access a wealth of rele­vant data and col­la­bo­ra­tively enable eco­no­mic and poli­ti­cal actors to anti­ci­pate mutual influ­en­ces of indi­vi­dual actions and incor­po­rate them into their own decisi­ons. An important role here is played above all by the abi­lity to share data in a trus­ting man­ner, if necessary even pre­ser­ving data privacy.

Solu­tion

PAIRS enab­les dyna­mic fore­cas­ting of and cor­re­spon­ding response to cri­sis situa­tions in three work steps. First, data sources from eco­sys­tems such as the Euro­pean cloud infra­st­ruc­ture GAIA‑X and other domain-rele­vant data infra­st­ruc­tures (e.g., data spaces from the energy and health sec­tors) are built up and inte­gra­ted into the PAIRS plat­form archi­tec­ture via open inter­faces. Secure data exchange is ensu­red here by using the stan­dards of the Inter­na­tio­nal Data Spaces (IDS) refe­rence archi­tec­ture. Based on the avail­able data, the AI hybrid tech­no­logy iden­ti­fies cri­sis situa­tions and their effects and deve­lops appro­priate response mea­su­res. In a third step, the response actions are recom­men­ded to the users via a cus­to­miz­able inter­face and the actors are sup­por­ted in the best pos­si­ble selec­tion of their response stra­tegy. A fully com­pre­hen­sive AI & data mar­ket­place ser­vice is avail­able for this inter­ac­tion. The selec­ted response actions of each actor are fed to the PAIRS plat­form anony­mously. This enab­les dyna­mic and detailed fore­casts and recom­men­ded actions that incor­po­rate the macroeco­no­mic and policy respon­ses of all plat­form par­ti­ci­pants as well as anti­ci­pa­ted interactions.

Use Cases

Sup­ply chains & logistics

Using the data of highly net­wor­ked pro­duc­tion sys­tems and uti­liz­a­tion chains, bot­t­len­ecks for spe­ci­fic pro­ducts are to be iden­ti­fied at an early stage in cri­sis situa­tions (e.g. incre­ase in demand for hygiene pro­ducts) and appro­priate mea­su­res taken and moni­to­red (e.g. build up stocks, set up tem­porary sto­rage loca­ti­ons, etc.).

Health Care

Based on data from hos­pi­tal infor­ma­tion sys­tems, ser­vices are to be crea­ted in PAIRS that pro­vide early and spa­tial warning of epi­de­mics and pro­vide par­ti­ci­pa­ting sta­ke­hol­ders with con­crete input for for­ward-loo­king demand plan­ning for cri­sis management.

Energy

Ana­ly­sis of energy sec­tor data should make it pos­si­ble to fore­cast the impact of cri­ses on energy demand.

Com­pa­ri­son

Prior to PAIRS With PAIRS
Cri­sis situa­tions are often dif­fi­cult for ope­ra­tors of cri­ti­cal infra­st­ruc­tures to manage, so that no tar­ge­ted mea­su­res can be taken quickly. PAIRS allows cri­ti­cal infra­st­ruc­ture ope­ra­tors to gain a much bet­ter over­view of cri­sis situa­tions and pro­vi­des tar­ge­ted recom­men­da­ti­ons for action based on a wealth of data.
In cri­sis situa­tions, actors are only able to react to the initial event, if at all, but not to the over­all sce­n­a­rio, which is also gui­ded by the reac­tions of other actors. PAIRS also takes into account the reac­tions of a large num­ber of actors to an initial cri­sis event in its ana­ly­sis, thus enab­ling dyna­mic cri­sis management
In order to be able to take tar­ge­ted mea­su­res in cri­sis situa­tions, a large amount of data is requi­red, which has not been avail­able to date. As part of the PAIRS infra­st­ruc­ture, nume­rous data sources are lin­ked so that signi­fi­cantly more effec­tive mea­su­res can be iden­ti­fied based on the inte­gra­ted AI methods.

Down­load detailed pro­ject infor­ma­tion for free!

The joint pro­ject PAIRS is fun­ded as an AI light­house pro­ject wit­hin the frame­work of the “Inno­va­tion Com­pe­ti­tion Arti­fi­cial Intel­li­gence” by the Federal Minis­try of Eco­no­mics and Cli­mate Pro­tec­tion (BMWK).

Copy­right © 2022 PAIRS Projekt