The ITEA 2 projects AMALTHEA and AMALTHEA4public are part of a 'string of pearls' in the automotive domain; successes that have pushed this domain into the next phase of its development. AUTOSAR, a result from the former ITEA project EAST-EEA, defined a methodology for component-based development of automotive software and a standardised software architecture for automotive electronic control units. However, AUTOSAR offered only limited support for detailed behaviour descriptions, which are indispensable for developing much more complex multi-core systems of high quality. Those require an increased exchange between tools. Multi-core optimisation especially relies on additional information like detailed timing behaviour. AMALTHEA set about adapting existing development methods and tools and creating a common model that offers the required description capabilities on different abstraction levels. The follow-up project AMALTHEA4public was set up to foster the transfer into application and to create a sustainable open (“public”) platform and a vibrant community of users and contributors.
One of the major achievements of the AMALTHEA consortium was a common meta-model for multi-core software and hardware modelling that enables integration of heterogeneous tools in a custom tool chain to gain easy and efficient access to the overall characteristic of a multi-core system. The AMALTHEA platform is distributed under the Eclipse open source license (EPL) and allows efficient data exchange between different cooperating companies but also between different (new and/or existing) tools used by a single organisation. The AMALTHEA model was taken to a next level by the AMALTHEA4public consortium by adding additional features like verification and test generation as well as traceability of requirements.
In 2015 the AMALTHEA4public framework was moved to the newly created open source project Eclipse APP4MC, an open tool platform for modelling, analysis and optimisation of embedded multi- and many-core software (www.eclipse.org/app4mc/). APP4MC was created by AMALTHEA4public project participants Robert Bosch GmbH, itemis AG, Timing Architects Embedded Systems GmbH, Dortmund University of Applied Sciences and Arts and Eclipse Foundation Europe GmbH. In 2016 a second Eclipse project Capra (www.eclipse.org/capra/) was created. It contains the traceability management tools mainly contributed by the Swedish AMALTHEA4public partners.
AMALTHEA4pubic partners AVL, BHTC and Bosch demonstrated the applicability of the project’s results in industry in several projects incorporating different companies. The AMALTHEA data model or platform is currently in use with some major automotive vendors from several countries around the world.
Robert Bosch uses the consolidated model as input for tools (commercial and in-house) and to exchange timing information with customers. Bosch intends to publish some tools as open source to further support the APP4MC platform. The source code analysis tool “SCA2AMALTHEA” was the first contribution of that kind in August 2017.
For Timing-Architects Embedded Systems GmbH (now part of Vector Informatik), AMALTHEA is the ideal format solution for many of its customers working with software modelling for real-time analysis. The Eclipse Framework APP4MC offers AMALTHEA users an easy start and supports the first steps on the path of introducing model-based timing analysis with well-developed and maintained basic functions, such as Import/Export APIs.
With the help of the AMALTHEA format, real-world benchmarks are available and the exchange between research and industry is alive and ongoing. A fundamental cooperation between academic and industry has been built to master the upcoming performance challenges within the new EE architectures. Thereby, the AMALTHEA and AMALTHEA4public projects contribute to the training of future specialists and managers.
AMALTHEA and its successor projects have shown how an idea for a narrow application (timing analysis of multi-core automotive and telecom systems at some companies) has become an extensive application in many core systems for the most diverse applications in a big number of companies.