X-Domain describes the merging of different domains (i.e., braking, steering, propulsion, suspension) into single functionalities. One example in this context is torque-vectoring. Different goals can be pursued by applying X-Domain features. On the one hand, savings in fuel consumption and an improved vehicle driving performance can be potentially accomplished. On the other hand, safety can be improved by taking over a failed or degraded functionality of one domain by other domains. The safety-aspect from the viewpoint of requirements is highlighted within this contribution.
Every automotive system being developed and influencing the vehicle safety must fulfill certain safety objectives. These are top-level safety requirements (ISO 26262-1) specifying functionalities to avoid unreasonable risk. Every safety objective is associated with an Automotive Safety Integrity Level (ASIL) derived from a Hazard Analysis and Risk Assessment (HARA).
Current HARA-approaches are conducted on a domain-specific basis only. Such procedures lead to safety goals only addressing the steering, the braking, or the propulsion system, respectively. This contribution presents a new methodology that meets the new requirements arising from the introduction of X-Domain features. First, the exposure of an X-Domain operation space considering speed, lateral and longitudinal acceleration and the friction coefficient of the street is derived from representational fleet-data. In a second step, a generic driving situation is derived to assess the severity of different malfunctions. This driving situation is applied to the operation space. Furthermore, the controllability is assumed conservatively as 3. Finally, the ASIL for the complete operation space that considers both longitudinal and lateral accelerations is determined by the superposition of exposure, severity, and controllability.