A Model-Based Framework for Developing Real-Time Safety Ada Systems [chapter]

Emilio Salazar, Alejandro Alonso, Miguel A. de Miguel, Juan A. de la Puente
2013 Lecture Notes in Computer Science  
This paper describes an MDE framework for real-time systems with safety requirements. The framework is based on industry standards, such as UML 2.2, MARTE, and the Ada Ravenscar profile. It integrates pre-existing technology with newly developed tools. Special care has been taken to ensure consistency between models and final code. Temporal analysis is integrated in the framework in order to ensure that the real-time behaviour of the models and the final code is consistent and according to the
more » ... pecification. Automatic code generation from high-level models is performed based on the Ravenscar computational model. The tools generate Ravenscarcompliant Ada code using a reduced set of code stereotypes. A case study is described for a subsystem of the on-board software of UPMSat2, a university micro-satellite project. different types of concepts. Model-driven approaches introduce solutions for the specialization of the models for specific concerns, as well as the interconnection of concerns based on models transformations. It improves communication between stakeholders using the models to support the interchange of information. But the separation of concerns often requires specialized modelling languages for the description of specific concerns. This paper describes an MDE framework for the development of real-time high-integrity systems. The functional part of the system is modelled using the Unified Modeling Language (UML2) [12] ). Real-time and platform properties are added to functional models by means of annotations, using the UML profile for Modelling and Analysis of Real-Time and Embedded Systems (MARTE) [13] ). An analysis model for verifying the temporal behaviour of the system using MAST 1 [6] is automatically generated from the MARTE model. Finally, Ada code skeletons are generated, based on the system model and the results of response time analysis. Code generation is based on the Ravenscar computational model [3] , and generates D.13.1]. Related work includes the Ada code generator in IBM Rhapsody 2 [5], which generates complex Ada code but does not support MARTE or the Ravenscar profile. Papyrus 3 [9], on the other hand, supports functional Ada code generation from UML models, but cannot generate Ravenscar code and does not fully integrate temporal analysis with system models. The tools developed in ASSERT 4 follow a closer approach. Two sets of tools were developed in this project, one based on 14, 2] , and the other one on AADL 5 [7, 8] , which later evolved to the current TASTE 6 toolset [15] . Both can generate Ravenscar Ada code and include timing analysis with MAST. The main differences between these toolsets and the framework presented here are: i) This framework uses up-to date industrial standards such as UML2 and MARTE, instead of ad-hoc adaptions of UML; ii) the transformation tools in this framework have been built with standard languages; iii) the extensive use of standards in this framework makes it possible to use it with different design environments, without being tied to a specific development platform.
doi:10.1007/978-3-642-38601-5_9 fatcat:noar3cbcfveddhh52jh3x5atd4