With the increasing complexity of in-vehicle systems and the new generation of communication based (mission critical) applications the size of the state space increases with unmanageable pace. Even when all design-time procedures to guarantee safety are followed adequately, in run-time there irrevocable remain unforeseen and untested system states, state transitions and events. In order to guarantee safety on system level accepting that full testing is in fact impossible, novel methodologies and safety architectures with runtime support must be developed. The safety framework presented in this paper addresses some of the crucial issues and is used in the automotive and transport related projects.
A runtime framework for system safety
2003-01-01
353595 byte
Conference paper
Electronic Resource
English
Runtime Safety Assurance Using Reinforcement Learning
IEEE | 2020
|Runtime adaptation in safety-critical automotive systems
Tema Archive | 2007
|British Library Conference Proceedings | 2022
|