Paper Title
Innovation in the Field of Safe Software in Safety-Oriented Autonomous Systems
Abstract
The objective of the scientific paper is to set up a new safety strategy for an autonomous mechatronic system with
the different operating modes. The new scientifically investigated functions in the field of autonomous driving are described.
An autonomous mechatronic system of the future not only allows a safe feeling among others, but also must meet the
functional requirements for automated driving. Research and development is currently working on new platforms for safe
autonomous systems that meet the requirements for future automated driving/steering. As long as the autonomous system has
capabilities that allow the driver to intervene (for example, a steering wheel), it must be clear beyond doubt for every moment
who is responsible for driving the vehicle. For this purpose, there must be a function that takes responsibility into account. A
further goal must be to predict the driver's intention with the help of various sensors. Now it has to be stated, what are the
possible modes of operation for an autonomous system? In the research work among other things the questions are
scientifically examined.In order to ensure the safe functioning of an autonomous system, it is necessary to define a safety
concept that describes how the system can be controlled by the driver in the event of errors occurring, or how the system can
"safely" switch to a safe state. The challenge of a safe and autonomous system is that errors of various types must be safely
controlled. The stochastic approach therefore follows the assumption that redundant and diverse monitoring of the calculated
function signals is guaranteed. The redundancy can be achieved here with double signal lines. Another possibility is to use
different sensors to generate the same signal. In this scientific paper, the diversity calculation means that the calculation is done
in different ways. Thus, the diversity calculation can be based on two independent ranges of values.
Keywords - Software Reliability Growth Model, dependence failure, mechatronic systems, autonomous vehicle motion
software system.