COEGSS · Project

Supercomputer-Powered Simulation Platform for Modeling Global Risks and Market Dynamics

digitalPrototypeTRL 5Thin data (2/5)

Imagine you could build a digital twin of entire populations — millions of virtual people with realistic behaviors — and then run "what if" scenarios on a supercomputer to see how markets crash, diseases spread, or energy demand shifts. That's what COEGSS built: an HPC platform that generates these synthetic populations and lets decision-makers stress-test policies before implementing them in the real world. Think of it like a flight simulator, but instead of training pilots, you're training governments and businesses to handle global-scale risks.

By the numbers

consortium partners across the project

countries represented in the consortium

portal releases delivered to the GSS/HPC community

total deliverables produced

industry partners in the consortium

The business problem

What needed solving

Businesses and governments need to understand how complex global systems — financial markets, energy grids, supply chains — behave under stress, but traditional analytics tools cannot model millions of interacting agents in real time. Without large-scale simulation capability, decision-makers are flying blind when anticipating cascading failures, market shifts, or policy impacts across interconnected systems.

The solution

What was built

The project built an HPC-based web portal (delivered in 4 successive releases with evolving architecture) for generating customized synthetic populations and running Global Systems Science simulations. Across 29 total deliverables, the consortium created tools for high-performance data analysis, population modeling, and real-time risk assessment accessible through a community portal.

Audience

Who needs this

Reinsurance companies modeling systemic and catastrophic risksEnergy utilities forecasting demand across volatile global marketsGovernment policy simulation units testing intervention scenariosFinancial regulators stress-testing market contagion scenariosConsulting firms offering scenario planning to enterprise clients

Business applications

Who can put this to work

Insurance & Risk Management

enterprise

Target: Reinsurance firms and catastrophe modeling companies

If you are a reinsurance firm dealing with the growing difficulty of pricing systemic risks like pandemics or financial contagion — this project developed an HPC-based platform for generating synthetic populations and running large-scale simulations. It could help you model cascading failures across interconnected global systems, moving beyond traditional actuarial tables to agent-based risk scenarios powered by supercomputing.

Energy & Utilities

enterprise

Target: Grid operators and energy trading companies

If you are an energy company struggling to forecast demand across volatile global markets — this project built simulation tools that blend big data streams with high-performance computing to model global energy use patterns. With 13 partners across 6 countries contributing expertise, the platform can generate realistic population-level consumption models to stress-test your grid planning assumptions.

Government & Public Policy

enterprise

Target: National statistics offices and policy simulation units

If you are a government agency needing to model the impact of policy changes on millions of citizens before rolling them out — this project created customized synthetic population generators running on HPC infrastructure. The portal went through 4 documented release cycles and was designed to provide real-time assessments of global risks and opportunities for decision-makers.

Frequently asked

Quick answers

What would it cost to use this simulation platform?

The EU contribution for this project is not available in the dataset, so specific development costs cannot be quoted. The platform relies on HPC infrastructure (supercomputers), which typically involves significant compute-time costs. Access models would likely depend on the hosting institution (Universitaet Potsdam) and available HPC allocations.

Can this scale to model real national or global populations?

The project was specifically designed for global-scale modeling — the core innovation was generating 'customized synthetic populations' for Global Systems Science applications using High Performance Data Analysis. The platform was built to handle very large datasets including social media data streams, suggesting it was architected for population-scale simulations.

What about IP and licensing — can we use this commercially?

This was an RIA (Research and Innovation Action) project with 13 partners across 6 countries. IP arrangements would be governed by the consortium agreement. The portal was released to the 'GSS/HPC community,' suggesting an open or semi-open access model. Commercial licensing would need to be negotiated with the coordinator at Universitaet Potsdam.

Is the platform still maintained after the project ended in 2018?

The project closed in September 2018. With 4 portal releases documented and 29 total deliverables produced, substantial infrastructure was built. However, post-project maintenance status is unclear. Based on available project data, you would need to check with the coordinator whether the platform remains actively supported.

How does this integrate with existing enterprise IT systems?

The platform was designed as a web portal (4 documented releases with updated architecture), which suggests browser-based access rather than deep on-premise integration. It connects to HPC back-end infrastructure for computation. Based on available project data, integration with enterprise data systems would likely require custom development.

What data sources does the platform actually use?

According to the project objectives, the platform ingests 'very big datasets including data streams from social media' and blends them with other data to generate synthetic populations. The exact data sources per use case would vary, but the architecture was built to handle High Performance Data Analysis at scale.

Consortium

Who built it

The COEGSS consortium brings together 13 partners from 6 countries (Germany, Spain, France, Italy, Poland, Sweden), with a mix of 4 universities, 4 research organizations, 3 industry players, and 2 other entities. The 23% industry ratio and only 1 SME suggest this was primarily a research-driven effort rather than a market-oriented one. The coordinator, Universitaet Potsdam in Germany, is an academic institution, which aligns with the project's nature as a Center of Excellence. For a business looking to adopt this technology, the limited industry presence means commercial readiness may require additional development beyond what the consortium delivered.

UNIVERSITAET POTSDAMCoordinator · DE
CSP INNOVAZIONE NELLE ICT SCARLthirdparty · IT
COSMO TECHparticipant · FR
CONSORZIO TOP-IX - TORINO E PIEMONTE EXCHANGE POINTparticipant · IT
INSTYTUT CHEMII BIOORGANICZNEJ POLSKIEJ AKADEMII NAUKparticipant · PL
DIALOGIK GEMEINNUTZIGE GESELLSCHAFT FUR KOMMUNIKATIONS UND KOOPERATIONSFORSCHUNG MBHparticipant · DE
ATOS SPAIN SAparticipant · ES
GCF - GLOBAL CLIMATE FORUM EVparticipant · DE
SCUOLA IMT (ISTITUZIONI, MERCATI, TECNOLOGIE) ALTI STUDI DI LUCCAparticipant · IT
ISTITUTO PER L'INTERSCAMBIO SCIENTIFICOparticipant · IT
UNIVERSITY OF STUTTGARTparticipant · DE
ATOS IT SOLUTIONS AND SERVICES IBERIA SLthirdparty · ES
CHALMERS TEKNISKA HOGSKOLA ABparticipant · SE

How to reach the team

Universitaet Potsdam, Germany — reach out to the HPC or computational science department

Order a report on this consortium See UNIVERSITAET POTSDAM profile →

Next steps

Talk to the team behind this work.

Want to explore how supercomputer-powered population modeling could strengthen your risk assessment or market forecasting? SciTransfer can connect you with the COEGSS research team and help translate their platform into your business context.

Order a report on this topic More in Digital & ICT