The work presented in this paper is built upon the work of the authors in recent years, covering the reliability-aware design and implementation of multiprocessor platforms. The authors introduce a framework performing a model-driven design flow for fault-tolerant embedded systems as a new contribution. It allows the designer to provide the abstract description of the application and execution platform that is later taken as input by a design space exploration engine presented in another work, which identifies the most effective solution for the implementation of a reliable embedded system on a multiprocessor platform.

The presentation is well organized and reports the state of the art with respect to the main topics of design frameworks and reliability-aware design space exploration, in order to contextualize the proposal. Once the preliminaries are set, the authors discuss the model-driven framework for the support of the specification of the application, the platform characteristics in terms of available resources and their fault management features, and the requirements and constraints to be guaranteed by the implemented system (such as deadlines).

Part of the paper is devoted to the presentation of the design space exploration engine, possibly for the sake of having a self-contained discussion, although this part of the work has been presented elsewhere, with all of the necessary details. Finally, the last part of the work reports the application of the framework (design and design space exploration) to two case studies. The first one, an adaptive cruise controller, is discussed and presented in a detailed way, to help readers appreciate the effectiveness of the framework. The second one is a production line controller, for which only some details of the final implementation are reported, referring mainly to the design choices selected by the design space exploration engine.

While the work is interesting, the authors’ decision to mix the new contribution with the design space exploration work muddies the real focus and relevance of the design framework itself. The second case study highlights this awkwardness: it provides no evidence of the adoption of the model-driven design framework, while it focuses on the results of the application of the design space exploration work. A reader would expect a comparison against alternative solutions, not only with respect to the design framework, but also to the rest of the flow. Apart from this aspect, the paper is well written and easy to follow.