This paper introduces the Switching Collective Risk Barrier (SCRB) framework, integrating control barrier functions with hybrid systems theory to ensure safety across discrete operational modes while accounting for collective risk aggregation among agents.
Key findings
Addresses safety during inter-mode transitions, collective risk management, and probabilistic safety guarantees under uncertainty.
Derives necessary and sufficient conditions for safety maintenance under switching policies.
Proposes a compositional verification approach scalable to large multi-agent systems.
Supports multiple risk metrics including VaR, CVaR, and coherent risk measures.
Limitations & open questions
The framework's practical deployment and real-world validation are yet to be demonstrated.
The probabilistic guarantees under uncertainty require further theoretical development.