The high level objective of the ANDRES project is to improve the competitiveness of innovative European industries, such as the telecommunication and automotive industry, by reducing the design time and cost of
highly integrated embedded systems. These systems are heterogeneous in nature. They include up to four different domains: software, analogue hardware, static hardware, and dynamically reconfigurable hardware, the latter gaining importance because of its new attractive combination of flexibility and efficiency. Currently no methodology exists allowing to seamlessly specify, simulate, synthesize and verify such heterogeneous systems, because each domain comes with its own computational models, languages and design tools. This prevents early holistic system validations and postpones the system verification to the system integration phase causing long, costly and most importantly time consuming design reiterations.
ANDRES will develop solutions to overcome these incompatibilities by developing an integrated modelling approach for heterogeneous embedded systems. This approach builds on the open-source modelling language SystemC already adopted by many European companies. Next ANDRES will close the gaps in the tool flow for dynamically reconfigurable hardware by developing a tool to translate adaptive models onto RTL descriptions. The prime result of ANDRES is a seamless design flow, which provides the possibility of designing embedded hardware/software systems on a higher level of abstraction emphasising in particular the application of run-time reconfigurable architectures.
Achieving these objectives is a major challenge and requires resources and expertise on the European level. Leading European companies providing application know-how and research institutes with outstanding experience in modelling and synthesis of embedded systems have joint hands in the ANDRES consortium to develop scientifically sound yet industrially applicable solutions for this challenge.