C2 - Integration of Logistics Data to support Autonomous Cooperating Logistics Processes (new)

Motivation

Autonomously controlled process landscapes comprise of a plethora of individual entities. Autonomous control calls for access to data which not only describes the entities themselves but also the environments in which they move in order to provide a basis for the decision making which consequently influences their behaviour. The sources of that data are as heterogeneous as the logistic entities themselves – they comprise central IT systems as well as highly distributed, embedded systems, RFID tags and software agents. They communicate with each other across existing channels using various protocols. The challenge with respect to providing access to data lies in the resolution of structural and semantic heterogeneity conflicts between the data sources and sinks – that is, data integration.

Objectives Phase 2 (2008-2011)

The objective of the subproject is the conception and implementation of a service-oriented, system independent approach to data integration within the context of autonomous control. On the basis of the existing communication channels and protocols, the proposed approach will provide support for arbitrary logical views upon the entirety of data required by autonomous control in logistics processes. Regardless of the querying entity, the approach will provide transparent access to the physical data sources whilst delivering the data in a native form. Necessary conversions and transformations will be carried out automatically and without the querying entity’s involvement.


Approach Phase 2 (2008-2011)

In this period, both the theoretical foundations for the development of a service-oriented, semantic data integration mechanism for autonomous logistics processes were researched and a prototypical mechanism were realised. A number of methods and tools were utilised to achieve these goals. These include methods of data modelling, the representation of semantic information and the model-driven design of service architectures. The first step was the analysis of relevant data sources, which contributed to the specification of a requirements analysis. Subsequently, a service-oriented, semantic concept for data integration was specified. Semantic descriptions of data sources and corresponding transformation mechanisms were developed. Consequently, the concepts and specifications were realised as a service-oriented, semantic approach to data integration and evaluated.

Results Phase 2 (2008-2011)

Sub-project C2 developed a concept for a service-oriented approach to data integration. An ontology-based mediator for the integration of heterogeneous data sources in autonomous logistics processes was realised. Semantic descriptions of widespread data exchange formats in logistics, such as EDIFACT EANCOM and EPCIS, were specified using the ontology description language OWL-DL. Corresponding rule-based and algorithmic transformation mechanisms were researched and realised. A development model was designed which applies methods of model-driven, service-oriented software architecture to the systematic provision of logical views as services. A hardware abstraction layer was developed for the connection of RFID systems to the mediator. The layer was adapted and implemented for the demonstrator “Factory of Autonomous Products”.