Relating machine and business data for added business value

Connecting shop floor resources with enterprise resources is a major benefit for manufacturing enterprises as it provides advanced real-time monitoring and analytic capabilities. The mobile maintenance use case of the ComVantage project, for instance, leverages interlinked machine and business data to accelerate the prediction and diagnosis of machine defects. Interlinked data sets provide access to case-relevant information like defects history, similar maintenance reports and manufacturer specifications about operating thresholds or provide a comprehensive base for statistical evaluations. Increased machine up-times and lower maintenance costs due to optimized preparation of on-site operations are direct business benefits (refer to demo #1 for more details).

ComVantage uses a semantic data harmonization approach based on Linked Data that is capable of publishing and interlinking data from various distributed sources in the web, even across organizational boundaries. Moreover, the use of standardized semantics (e.g. from well-known ontologies such as vCard or GoodRelations) allow for automated processing and interpretation of information or enable non-experts to deal with complex problems.

Challenges of the overall approach are mainly connected with the extraction and integration of data from different data sources. ComVantage utilizes technology-specific Linked Data adapters, e.g. for relational databases, Excel spreadsheets or machine middleware. These adapters perform a mapping of arbitrary interfaces, data formats and metamodels to a uniform model that is the basis for interlinking business entities among different data sources.

The Data Harmonization Middleware (DHM) adapter was developed to integrate various machine types from the shop floor. It leverages the capabilities of the GAMMA machine middleware that provides a uniform interface to machines. The DHM adapter allows for live access to current machine data (read and write), access to information about local machine set ups, access to aggregated machine history data (e.g. sensor time series) and access to machine self-diagnosis functionalities (e.g. test scripts that are deployed on the machine). It offers a SPARQL endpoint to query for information and delivers Linked Data to the requesting client. The concept of the DHM adapter can be reused to integrate also other middleware solutions, e.g. a proof-of-concept adapter for the common OPC-UA standard was developed as well.

The results of the ComVantage project enable the evolvement of linked machine data networks to be leveraged in business contexts, e.g. manufacturing execution or predictive mobile maintenance systems.