Component-Level ASIL Decomposition for Automotive Architectures

The Automotive industry is evolving towards a more electronics-assisted driving and self-driving functionality. The addition of complex subsystems has a great impact on the current vehicle architectures, leading to safety concerns. In this work we present a technique that follows the ISO 26262: Road Vehicles - Functional Safety standard to introduce redundancy in the architecture by using ASIL decomposition, and perform a safety analysis of the modelled system. A three-layer model is used to describe the application, the resources, and the physical space of the vehicle. In this paper we introduce novel model transformations to replicate parts of the application following ASIL decomposition rules. Finally, we perform a cost analysis and a probabilistic fault tree analysis on the architecture, making a comparison between different possible solutions. The advantages of these techniques, such as traceability and scalability, are shown by modelling and analysing the lateral control application of a real truck platooning system.