D6.1 Initial Architecture for Fast Small-Scale Deployment

To read it download it from this link

Executive Summary

The main goal of WP6 – “Architecture and Integration” – in Learning Layers is to bridge the divide between fast and flexible deployment of Layers tools and technology in the application partner environments by providing a unified, configurable, lightweight and distributed infrastructure. To this end WP6 work during the first project year was focused on the deployment of a small-scale architecture for fast deployment, which is also the subject of this deliverable. As outlined in Section 2, the Layers architecture vision is built on three cornerstones. First, research orientation to be able to align the engineering processes with the challenges of scaling in informal learning scenarios. Second, open source development with the aim of involving and integrating as many user and developer groups as possible—also externally. Third, continuous integration­ on the co-design and infrastructure levels to ensure a cohesive and connected user experience based on an integrated Layers architecture. Realizing these cornerstones, the core part of the deliverable comprises Sections 3 to 5, reporting respective work streams and successes during Year 1 as summarized in the following.

A survey of existing architectural models and infrastructure-level technologies was conducted during the first six months of the project. With the main first-year objective being fast deployment it was clear that we would not be able to adopt technologies that are too complicated to setup, deploy, manage, or integrate. The attention was therefore put on technologies that were brought to the project by the technology partners, and to further develop those technologies in tight integration with the co-design work pursued by the four Layers design teams. The House of Quality (HoQ) approach was adopted to offer a consulting tool for assessing architectural options and comparing available tools for informal learning support. The end-user requirements were elicited through co-design activities and captured in Requirements Bazaar, a social requirements engineering toolkit that was initially developed by RWTH in the ROLE project and was awarded the best demo paper award at the 2013 IEEE International Conference on Requirements Engineering.

A major driver for developing and integrating Layers technology during the first year was the Layers Developer Task Force (LDTF), an association of developers from the technical consortium partners who met regularly and acted largely autonomously. To facilitate this distributed development workforce we followed the vision of open (source) development by setting up the Open Developer Library (ODevL), a development infrastructure that is open to external developers and enables eliciting and managing user requirements, hosting the source code, tracking issues, and managing a continuous integration process. The 1st Layers Developer Camp organized by LDTF was a full success in involving external developers by bringing together 26 developers from the technical partners in the Learning Layers project and the external developer community from the Aachen region. As a further measure for involving external developers, two student teams from the “High-tech Entrepreneurship and New Media” lab course at RWTH were recruited to support development in the CAPTUS design team and in integrating TUG’s Social Semantic Server with RWTH’s Semantic Video Annotation technology.

The Architecture Board was established with the responsibility of taking project-wide decisions on the overall Layers architecture. It signed off the year-one small-scale Layers architecture for fast deployment, which lays the foundation for the continuous integration vision on the co-design and infrastructure levels. The architecture is rooted in the design team and development activities during Year 1 and the use cases stemming from these co-design activities. It integrates components contributed by WP6 technical partners on three layers: the cloud infrastructure layer based on i5Cloud and m-learning platform; the application services layer featuring social-semantic services, community support, semantic video annotation, mobile Web, and formative evaluation services; and the service access layer, which will serve as a means for increasing flexibility during deployment in the application clusters in the coming project years.

Leave a Reply