The importance of diagnosis in on-board automotive systems is increasingly blending with the complexity of the systems. This important role has no equivalent in the design chain since diagnosis is currently only the last step in the design process. This situation leads to diagnostic functions that are often a compromise that could sacrifice part of environmental, customer and sometimes also safety requirements. The IDD project aims to solve this critical situation through the formalisation and standardisation of the diagnostic design process, and the realization of new tools for the designers that can help them in evaluating and understanding the effects of each choice on the system being designed. A success in reaching these objectives will lead to new systems with a higher degree of reliability and higher diagnostic performances with advantages for environment, safety and customer.
The project aims to make a contribution to re-organising the design process to include aspects of diagnosis in early steps. Main goal of the project is to develop a methodology for integrating the analysis of diagnosability and avoidance of fault effects in the design chain and a set of tools that support the designer in this analysis. This will be done by creating interfaces between currently used tools, such as CAD and numerical modelling, and simulation and advanced Model Based systems. The ultimate goal of this step is to improve performance of cars with respect to reliability, safety and environmental impact.
The re-organisation of the design chain with the introduction of diagnosis in an early stage of the system development can be achieved giving to designers appropriate tools and methodology. Applying techniques of Model Based Diagnosis, a new approach for the design process of OBD (on-board diagnosis) functions in vehicles can be defined, specified, implemented, integrated in a more usual model-based co-design process for functions.
During the project some potential problems in the current design process have been identified. Possible solutions have been proposed and tested introducing a new design process. In the new process different activities can be performed in an interleaved way, such as the selection of components and their layout, definition and simulation of a system model, diagnosability analysis, and generation of on board diagnostics: this can lead to a significant transformation of the design cycle, causing a reduction in the design time.
In order to support the new process, a toolbox has been defined that integrates software platforms for design and simulation (namely Matlab(R)/Simulink(R)) and model-based reasoning tasks for performing the activities related to diagnosis. The main results have been tested on real applications and are ready for the next phase, i.e. industrialization. The results achieved can demonstrate the feasibility and the potentiality of the approach developed.
Even if the project showed that the model-based reasoning can play a fundamental role in the new design process, it also pointed to new open problems to be faced by researches in the next years. The most important and interesting one is deriving qualitative models from quantitative ones. This is not an easy task because of the nature of quantitative models which are developed for control purposes. Still a lot of research will be need to be carried out to produce a solution that works in general for any type of quantitative models.
The automatic derivation is a fundamental opportunity for model based reasoning and diagnosis and it will provide a wider diffusion of these technologies. A further aspect that has to be taken into consideration in the application of the project results is the modification of the design process. This can have a strong impact on company organisation and should be analysed and evaluated on a company-by-company basis in order to achieve the expected results.
No results directly relevant to this theme. However, please note that some findings relevant to the project's key theme (Vehicle Technology) are generically applicable.
<?xml:namespace prefix = o ns = 'urn:schemas-microsoft-com:office:office' />