The PAIRS project

rivacy-Aware, intelligent and
Resilient Crisis Management

Project description


Faster crisis scenario identification with AI

The AI lighthouse project, initiated and coordinated by Advaneo and funded by the BMWi 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.

Challenge 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.


PAIRS enables dynamic forecasting of and the 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.).


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.


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


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.