STIFTUNG FRANKFURT INSTITUTE FOR ADVANCED STUDIES

Their core work

FIAS is a theoretical research institute in Frankfurt focused on computational science, brain modelling, and intelligent autonomous systems. Their work bridges fundamental neuroscience (Bayesian models of cortical perception) with applied robotics and high-performance computing for solving large-scale physics simulations. They contribute deep mathematical and algorithmic expertise to consortia tackling problems from robotic inspection systems to exascale computing engines. Their strength lies in turning theoretical insights about learning and computation into models that power real-world autonomous systems.

What they specialise in

How they've shifted over time

FIAS's H2020 activity is concentrated in a narrow window (2015–2016 project starts), making a clear temporal shift difficult to detect. Their earliest projects (BayesianHumanCortex, ExaHyPE) focused on fundamental theory — brain computation models and exascale numerical methods. The slightly later GOAL-Robots project (2016–2021) suggests a move toward applied autonomous learning, connecting their theoretical strengths to embodied robotic systems.

FIAS appears to be translating its deep theoretical expertise in computation and neuroscience toward autonomous robotic systems that learn and adapt — a direction relevant for industry partners needing intelligent automation.

How they like to work

FIAS predominantly joins consortia as a participant (3 of 4 projects), contributing specialist theoretical and computational expertise rather than leading large projects. They coordinated one project in their core strength area (computational neuroscience). With 25 unique partners across 10 countries from just 4 projects, they engage in medium-to-large international consortia and appear open to diverse partnerships rather than repeating with the same groups.

FIAS has built a broad network of 25 partners across 10 countries from only 4 projects, indicating participation in sizable international consortia with diverse European partners.

What sets them apart

FIAS occupies a rare niche as a theoretical institute that connects fundamental brain science with applied robotics and high-performance computing. Unlike application-focused robotics labs, they bring deep mathematical modelling — Bayesian inference, PDEs, autonomous learning theory — to practical engineering challenges. For consortium builders, they offer the kind of rigorous theoretical backbone that turns an engineering project into a scientifically grounded one.

Highlights from their portfolio

• GOAL-Robots
  Largest EC contribution (EUR 709,875) and longest project (2016–2021), representing FIAS's investment in autonomous robot learning at the intersection of AI and embodied systems.
• BayesianHumanCortex
  FIAS's only coordinated project, directly in their core strength of computational neuroscience and Bayesian brain modelling.
• AEROBI
  Demonstrates FIAS's ability to contribute to applied industrial problems — aerial robotic bridge inspection — beyond pure theory.