#	Pagina
attuale pagina	/open-h2020/projects/206578/results.html

Report

Teaser, summary, work performed and final results

Periodic Reporting for period 2 - COMPOSELECTOR (Multi-scale Composite Material Selection Platform with a Seamless Integration of Material Models and Multidisciplinary Design Framework.)

Teaser

Summary

The mission of H2020 COMPOSELECTOR project is to develop a Business Decision Support System (BDSS) to support the complex decision process involved in the selection and design of polymer-matrix composites (PMCs). This will be achieved by means of an open integration platform which enables interoperability and information management of materials models and data and connects a rich â€˜materials modelling layerâ€™ with industry standard business process models. The delivered product of the project is a software system for the reliable selection of composite materials considering multi-scale and multi-field coupled interactions. The pre-existing commercial software tools will represent the starting point for the implementation. The pre-existing software codes will be further developed and integrated during the project with the objective to obtain a mature software system that can be used in an operational context. The project is industry driven thanks to the significant involvement of the industrial partners. As a consequence, the project will be led by specific industrial needs. This guarantees a stable foundation and reduces the risk of the development activities. It also ensures cost effectiveness because the project will maximally use pre-existing technologies. The existing software tools will be further developed and integrated in this project. Although these tools are mature, they are extremely complex products. However, the objective at the end of the project is to obtain mature software platform commercially exploitable.

Work performed

A business Decision Platform Prototype: COMPOSELECTOR BDSS is moving from the drawing board into real life. Prototypes of some BDSS components are already publicly available for partnersâ€™ access at the web address https://business.composelector.eu. The BDSS provides a top-level entry point with single sign-on authentication and authorization implemented with state-of-the-art security mechanisms. The business layer of the BDSS supports the definition of business processes and their associated decisions rules by using the BPMN and DMN standards respectively. Business processes can be executed with the BDSS process engine. The engine supports user tasks to enable business users to interact with the system to take business decisions. Of course, automatic decisions which have been defined using the DMN standard are also supported.
Multiple business actors performing different roles are supported. Business data provenance information and the whole information about the business decisions taken during process execution are kept in persistent storage by the business layer. In this way, evidence about the procedure by which business data were generated and details about the actors that have interacted during business process execution, including the decisions taken by them, can be accessed at any time from the web interface. From the business layer, the business users get the possibility to start simulation processes which will be run on the MuPIF integration platform (see below), accessing data stored in the database layer. Web is used as a medium to convey information about DSS products or to distribute BDSS software. The COMPOSELECTOR BDSS will facilitate a wide variety of decision tasks including information gathering and analysis, model building, sensitivity analysis, collaboration, alternative evaluation, and decision implementation. The web is an enabling technology for delivering computerized decision support. Integration Platform (MuPIF) based on interacting, distributed objects. Instead of standardizing data structures, MuPIF is focused on identification and standardization of services on data and other components of the system, such as models or workflows. Focus is on services provided by objects (interfaces) and not on underlying data themselves.
The component-based design encapsulates data, metadata and services (algorithms) in one object. Data between models are exchanged in terms of (data) components, representing properties, spatial fields, or microstructures, for example. This way, models will get data, related metadata and operations in one consistent package and do not have to interpret data themselves.
The metadata on each component can be structured and defined by a schema (available in MuPIF 2.0). The component metadata can be validated against the schema. The metadata schemata have been defined for models, workflows and data components and are used to track the pedigree, versioning, and execution details.
MuPIF components can be local as well as remote objects allowing to naturally develop and execute distributed workflows, handle distributed data, or even enabling business model based on software or data as service and marketplaces integration. MuPIF platform supports SSL or VPN based secure communication and data exchange.
The individual models can be connected to the platform by implementing API, representing the interface of Model class. The platform supports interfacing to commercial, closed source as well as open source simulation tools and databases.
A prototype of Graphical workflow editor has been developed allowing more convenient and user-friendly graphical workflow composition without requiring programming skills in Python and knowledge of MuPIF internals.

Final results

The main innovation relies on the integration of material and process modeling in business decision system that takes into account: (i) the complexity of composite materials and (ii) the level of industrial expectation through the different end-users and their special requirements inherent to their respective field (KPIs). All this parameters will be considered to produce an adapted-to-end-users procedure, which maximizes the effectiveness of the material integration in the selection process chain and offers to the industrial world a powerful tool to optimize its productivity.

The market-oriented character of this proposal is justified by the fact that software developers involved in this project can base their work on their mature analysis platform for advanced multi-scale and multi-physic analyses, including pre- and post-processing and interfaces with other commercial codes. Therefore, the economic impact of the project is expected to be considerable such as: i) reduction of company costs and increased performance and commercial impact based on effective materials models driven business decisions: ii) guidance to companies in developing their strategies with an effective, user friendly materials models driven business decision system, iii) increased industrial use of existing materials knowledge and effective materials models iv) improved trust of industrial decision makers in materials modeling and their commercial advantage and v) essential company savings in time and money, especially via the elimination of the need for (some) plant trials. These all will lead to a substantial gain of time, shortening the design and pre-production of composite structures and decreasing the time/cost-to-market and therefore increase the competitiveness of the European industries.