AI for cyber-physical systems
AI-guided simulation-based design and verification of cyber-physical systems, particularly in aerospace, critical infrastructures, transportation, logistics.
At RAISE Lab we focus on AI-guided simulation of mission- and safety-critical cyber-physical systems to enable their verification and optimisation via design space exploration.
We develop methods based on both symbolic AI and machine learning as well as software for:
Modelling of system environments and generation of operational scenarios
A cyber-physical system must operate safely, effectively, and efficiently in its operational environment, which typically envisions the possible occurrence of uncontrollable events, up to some extent and subject to some constraints.
We provide means to declaratively model the system operational environment and its constraints, and to randomly sample or enumerate (any-horizon) system scenarios of interest from within such a constrained space.
Simulation-based verification on system digital twins
We provide methods and tools to perform system-level verification by orchestrating the simulation of system digital twins over scenarios extracted from operational environment models.
Statistical model checking
Our methods provide sound guarantees of error probability & accuracy of KPI estimations, as required for system certification and qualification.
AI-guided black-box optimisation on system digital twins
We perform intelligent design space exploration on system digital twins to find configurations optimal wrt. one or multiple objectives.
Massively parallel HPC simulations
Our methods and software heavily exploit massively parallel high-performance computing infrastructures (clusters of computers) to keep the computation time within the use-case requirements.
What-if analyses and decision support
Our methods enable interactive as well as completely automatic what-if analyses and decision support for system (single- and multi-objective) optimisation.
RAISE 














