Abstract
KEYWORDS – Software Redundancy, Hardware Redundancy, Triple Modular Redundancy, M-n-systems, n-version programming
ABSTRACT –
Ongoing advances in mechatronic components and power electronics help to improve control systems within automotive applications. New developed or designed components enable more efficient system architectures and control. Based on these development trends, mechatronic systems within automotive engineering are gaining more and more importance. At the same time, quality and safety requirements become challenging for automotive manufacturers as well as their suppliers regarding the decrease of default risk and increase of component reliability in a high degree.
To fulfil nowadays safety standards for complex mechatronic systems, several different analysis methods have to be applied. Only the connection of any safe fault recognition with a safe fault reaction enables a system to avoid harmful consequences. The increasing amount of electronic components in safety related environments creates a need for smart architectures and methods that can compensate single failures of its components. In all cases a reliability analysis method is in charge to identify and assess safety related systems as proposed by the ISO 26262. Systems or components which are evaluated with an ASIL A or higher are legitimate candidates for fail-operational architectures.
Regarding product development, there is an ongoing change from routine tests (durability tests) to testing selected parts of a safety function (fault injection tests). How action is taken is changing, with a trend towards a further development of software tools, supporting functional safe systems holistically, including hazard and risk assessment, integrated system analysis of systematic and random failures, and hardware metrics. Safety-relevant aspects in the development of modern mechatronic systems have to be considered thoroughly. The high number of technical properties and complex connections of mechatronics systems in the development of modern vehicles are very challenging for state-of-the-art analysis methods. For this reason, new and innovational safety concepts are required, to optimize existing safety concepts using conventional components and methods in combination.