Safety and Risk Assessment in Complex Systems

The field of safety and risk assessment is moving towards more sophisticated and nuanced approaches to evaluating and mitigating potential hazards. Researchers are developing new methods to address the complexities of modern systems, such as autonomous vehicles and embodied AI systems, where traditional approaches to safety assurance are no longer sufficient. A key trend is the development of probabilistic safety frameworks that can provide more realistic and scalable assessments of risk. Another area of focus is the improvement of safety integrity levels and the allocation of safety functions in high-risk industries. Notable papers in this area include:

  • Towards provable probabilistic safety for scalable embodied AI systems, which introduces a new paradigm for ensuring the safety of complex AI systems.
  • Probability of Collision with Tethered Spacecraft, which presents a method for estimating the probability of collision for tethered spacecraft that takes into account the tether's spatial extent and configuration uncertainty.

Sources

SIL Allocation for Mitigation Safety Functions

Stochastic Modeling of Road Hazards on Intersections and their Effect on Safety of Autonomous Vehicles

Probability of Collision with Tethered Spacecraft

Towards provable probabilistic safety for scalable embodied AI systems

Backward Responsibility in Transition Systems Beyond Safety

Built with on top of