The ARVIDA Reference Architecture
In order to be able to fulfil the increasing requirements of virtual product development with regard to functionality, consistency and ease of availability, as well as to ensure the broad availability and usability of virtual technologies (VT) in industrial work sequences, existing VT applications need to become more functional. In many cases, today's VT systems still have a closed system design. All functional modules are integrated into the software environment so that it is almost impossible to exchange any of the modules.
In order to be able to compose new adapted VT applications faster, more flexible and cost-effective, future VT environments need to have a significantly more modular set-up and be easier to scale than the existing systems. Therefore, the ARVIDA project creates an open reference architecture, based on established web technologies which will make creating distributed heterogeneous VT applications much more efficient in the future. The reference architecture uses the following, globally established web standards that have been acknowledged by the W3C (www.w3c.org):
- RESTful (https://en.wikipedia.org/wiki/Representational_state_transfer
- RDF (http://www.w3.org/RDF/), OWL (http://www.w3.org/2001/sw/wiki/OWL), and
- Linked Open Data (http://linkeddata.org/).
The key development targets of the reference architecture are:
- Introduction of a service-oriented structure
- Use of the Representational State Transfer (REST) principle for describing services and resources
- Use of semantic web technologies (RDF, OWL, Linked Open Data) for distributed applications
- Modularity due to open semantic resource descriptions with RDF/OWL
- Provision of VT-specific RDF vocabularies for the VT application development
- Easy creation of resources
- Uniform description of resources
- Standardised resources behaviour
- Basic compatibility with an Industry 4.0 reference architecture
Figure 1 shows an overview of the architecture concepts that have been implemented up to date.
So far, the results show that the ARVIDA reference architecture will initially save very little implementation effort, but will make it much easier to connect different systems with each other. This also applies if these systems have been created using different programming languages. It was shown that the reference architecture makes it much easier to replace functional modules with other modules, without a high implementation effort being required. The course of the project to date has shown that it requires less effort to maintain and standardise an RDF vocabulary than an entire Application Programming Interface (API).
The ARVIDA consortium is convinced that considerable progress with regard to the refinement and maturity of the reference architecture will have been made by the end of the project in August 2016. This conviction has become more deeply rooted based on the numerous excellent project examples from the 1st status meeting in April 2015.