https://www.mdu.se/

Component-based development has proven effective in many engineering domains, and several general component technologies are available. Most of these are focused on providing an efficient software-engineering process. However for the majority of embedded systems, run-time efficiency and prediction of system behaviour are as important as process efficiency. This calls for specialized technologies. There is even a need for further specialized technologies adapted to different types of embedded systems, due to the heterogeneity of the domain and the close relation between the software and the often very application specific system. This paper presents the SaveCCM component model, intended for embedded control applications in vehicular systems. SaveCCM is a simple model in which flexibility is limited to facilitate analysis of real-time and dependability. We present and motivate the model, and provide examples of its use.

The component-based strategy aims at managing complexity, shortening time-to-market, and reducing maintenance requirements by building systems with existing components. The full potential of this strategy has not yet been demonstrated for embedded software, mainly because of specific requirements in the domain, e.g., those related to timing, dependability, and resource consumption. We present SaveCCT - a component technology intended for vehicular systems, show the applicability of SaveCCT in the engineering process, and demonstrate its suitability for vehicular systems in an industrial case-study. Our experiments indicate that SaveCCT provides appropriate expressiveness, resource efficiency, analysis and verification support for component-based development of vehicular software. 

In software engineering for embedded systems generic reusable software components must often be discarded in favor of using resource optimized solutions. In this paper we outline a modelthat enables the utilization of component-based principles even for embedded systems with high optimization demands. The model supports the creation of component variants optimized for different scenarios, through the introduction of an entrance preparation step and an ending verification step into the component design process. These activities are proposed to be supported by tools working on metadata associated with components, where the metadata is possible to automatically retrieve from many development tools. This paper outlines the theoretical model that is the basis for our current realization work.

Attractive benefits with successful implementation of component-based principles include managing complexity, reduction of time-to-market, increased quality, and reusability. Deployment of component-based development is however not simple - it depends on many strategic, technical, and business decisions. In this paper we report experiences from our attempts with finding a correct implementation of component-based principles for the business situation of sub-contractors of embedded systems. Findings related to suitable component models, component technologies, and component management are presented. Overall the results confirm the suitability of component-based principles for the domain, but also show the need (and potential) in further development of CBSE theory and technology for embedded systems.