Effective Application of Model-Based Software Engineering Technology to Automotive Embedded Software 2024-01-2049

The major driving factors for adoption of model-based software engineering (MBSE) practices are increasing complexity, time-to-market pressures, and certification requirements. Adoption of MBSE technology can add tremendous value to an automotive software development organization, however to realize maximum efficiency it is critically important to scale and customize these practices appropriately. This paper offers a practical, tool vendor-independent framework for application of MBSE technology to automotive embedded software development, based on tendencies and outcomes observed while working with automotive OEMs and Tier 1 suppliers of various sizes. The core of the presented approach is to first define the MBSE technology strategy and then to operationalize the strategy. There are three main elements that constitute the MBSE technology strategy of a software development organization. The first element is technology scope – it defines the balance between MBSE and traditional coding methods. This is rarely a binary decision, and it takes into account not only the application domain, but also the culture of communication between engineering functions internal and external to the organization. The second element is integration paradigm, i.e., whether to integrate software in model or code domain. This is typically driven by balance between effort, performance, and program objectives. The third element is software stack design. Commonly discussed in the web and mobile space, it is sensible to address software stack design specifically as it applies to automotive embedded software. Stack design choices are directly tied to business objectives of an organization, as they frequently (and sometimes implicitly) define critical interfaces, draw ownership boundaries, and inform make vs. buy decisions. Finally, once the MBSE technology strategy is defined, it is operationalized by leveraging traditional software DevOps practices, modified and extended to support MBSE methods in the automotive embedded space.