Candidate
Jordan Ross
Title
Synthesis and Exploration of Multi-Level, Multi-Perspective Architectures of Automotive Embedded Systems
Supervisor
Krzysztof Czarnecki
Abstract
Exploring candidate architectures is a routine step in the design process for any complex system, including automotive electric/electronic (E/E) systems. Today, many solutions exist to find optimal candidates given a small number of design decisions (i.e., degrees of freedom or variability). However, these solutions lack by not considering all abstraction layers (i.e., feature model, functional architecture, and hardware architecture) that make up the architecture as well as limit the possible variable design decisions one can make. These limitations not only lead to solutions that may not be globally optimal but also restrict the design space that engineers can explore and analyze.
In this presentation I will first present a reference model suitable for early design of automotive E/E architectures which can be used to express the possible candidate architectures. I will then show the types of design exploration that we can achieve when using the structural modeling language, Clafer, to encode our models and its supported backend solver to synthesize the candidates. Furthermore, I will show how using our reference model and Clafer it is possible to explore scenarios, automatically, which current state-of-the-art tools are unable by synthesizing the possible candidates with respect to some design decisions, objectives, and constraints.