An Industrial Method for the Model-Based Development of Automotive Systems

Software and systems engineering offers a multitude of innovative methods for the design, realization, and management of systems that support software and systems engineers from the initial idea through design and specification to the operation and disposal of the system. This thesis therefore investigates how these methods can contribute to the efficient and effective specification and design of automotive systems and explores ways to utilize these methods for the development of such large and complicated systems. Although systems engineering achieved remarkable results in developing cyber-physical systems in selected examples, many previous works in the field of automotive systems engineering have yet to consider several real-world applications in a wider context of their utilization. Therefore, many aspects that appear reasonable in a single application of the methods lead to significant problems in their industrialization as soon as several development projects and different possibilities to decompose the system into system levels are considered. For example, while the execution of the same task from different perspectives may seem careful and suited to reveal further insights in a single application or a single system model with a few elements, the scaled application for systems with hundreds or thousands of model elements reveals that performing the same task multiple times using an inefficient modeling method leads to excessive efforts in system modeling that frustrates the system engineers. Consequently, this thesis considers several industry projects and provides tailored methods for defining and applying system models.