The integration work package will provide the backbone of the VIRTUE project through the provision of data and task level management and co-ordination to enable the tools developed within other work packages to be brought together within a holistic environment.
The integration platform will provide state of the art techniques to enable users to:
- Operate existing and future CFD software through a common user interface
- Share data between disparate tools.
- Ensure the maintenance of consistency of data across the VIRTUE platform.
- Manage multiple projects simultaneously.
- Communicate using a variety of differing formats to facilitate collaboration.
- Optimise their designs either locally or globally across the entire VIRTUE platform.
- Visualise the data produced from simulations using a common interface that may itself be used within the collaborative process.
- Mine parametric data generated through the use of the tools either manually or experimentally and display the resulting knowledge within an interface to enable the user to explore what-if scenarios.
- Validate the data generated through the simulations with that generated experimentally as a feedback mechanism to further improve the reliability of the simulations.
The objective of Work Package 5 is to provide a platform that will enable various distributed CFD and design applications developed within VIRTUE to be integrated so that they may operate in a unified and holistic manner. The distributed nature however imposes a number of integration challenges to ensure that the data is consistent between the applications involved, that changes and interactions between the applications are correctly propagated and that the simulations are undertaken in an organised manner. Appropriate technologies will also be delivered to support multi-media collaborative design across a number of designers working both on the same project and different projects, as well as an optimisation toolkit, and knowledge discovery and data mining technology to improve the performance and use of VIRTUE. A validation mechanism using experimental data will also be incorporated to enable comparison with calculated performance and evolution of simulation codes to produce and continually evolve more accurate and reliable predictions. Specific objectives include to:
- Detail the requirements specification for the platform in order to ensure all aspects, features and modes of operation are fully defined.
- Develop and define data management techniques to facilitate data transfer between currently disparate sets of tools.
- Develop and implement the integrated platform to manage data transfer and the platforms performance, process management and user interaction.
- Define, develop and implement an appropriate user interface that shall support multiple user requirements through distributed collaborative working.
- Provide an optimisation tool set to support specific hydrodynamic local optimisations to be integrated within a global optimised design.
- Support the evolution of the platform and integrated tools, and provide a parameterised design model, through the development, implementation and utilisation of data mining and machine learning techniques.
- Develop and implement a holistic visualisation interface through integrated post processing of simulated data.
Description of work:
The relationships between each of the tasks within the integration work package have been clearly identified. Based on this, a dependency matrix has been derived. The following tools will be developed/implemented within the integration work package:
- Common model. The common model will be represented within an accepted database and will be responsible for managing the storage of data that is shared amongst the simulation tools to be integrated. The database will be chosen to accommodate the selected data format and language, as well as to provide distributed access, multiple project management, and data locking.
- Generic wrappers. The generic wrappers form the basis for integrating each of the design and simulation tools within the VIRTUE platform. The wrapping will be loosely coupled to remove any limitations regarding the range of tools that may be integrated, and will provide the management functionality to process: data from the common model; native data; performance data for visualisation purposes; consistency control, and parametric data for optimisation and data mining purposes.
- Integration engine. The integration engine will be responsible for co-ordinating data access and usage, resource allocation, task management, conflict management, and will represent the centre of the control for the VIRTUE platform, ensuring that design and simulation activity is undertaken at the right time for the right reasons.
- User interface. The user interface will represent the window to the VIRTUE platform, which will enable users to operate local and remote design and simulation tools, as well as providing collaboration functionality across groups of designers. The user interface will be developed using a modular plug-in approach, to enable the interfaces of the optimisation, data-mining, visualisation, and verification and validation modules to be plugged in to provide a customisable feel.
- Integrated optimisation toolkit. An existing optimisation toolkit will be used (modeFrontier) and will be integrated via a generic wrapper and interface into the VIRTUE platform to take advantage of the control mechanisms within the integration engine. The wrapper and interface will be developed to ensure that additional optimisation toolkits may be integrated when they are made available.
- Data mining toolkit. The data mining toolkit will take any data that has been generated by the simulation tools in a parametric format, store it in a database, and use data mining techniques to extract information and knowledge which will then be presented within a knowledge interface. The resulting parametric model can then be used to support exploratory conceptual design.
- Visualisation interface. The CFD tools will generate a large amount of data that may be used for visualisation within a plug-in interface. The interface will enable interaction with the data in order to further improve the users understanding of the complexity of the problem.