Visual analysis spanning multiple data sources usually requires the integration of multiple specialized applications to handle their heterogeneity. This is also true in manufacturing, where data about orders, personnel, workloads, maintenance, etc. must be analyzed together to make well-founded management decisions. Yet, the orchestration of multiple data sources and applications poses challenges to the software infrastructure and to the analyst. We present a 3-tiered approach to cope with these challenges. In a first step, we assume a domain-dependent analysis workflow as the mental model of the analyst. Based on the novel concept of contextualization, we then align the different applications with this model in order to provide their meaningful integration. As a third step, we incorporate the data according to its use in the aligned applications by means of a service-based architecture. By starting the integration process on the user level, we are able to pragmatically target and streamline the required integration to a degree that is technically achievable and interactively manageable. We exemplify our approach with the Plant@Hand system for integrating manufacturing data and applications.
Aehnelt, M., Schulz, H. J., & Urban, B. (2013). Towards a contextualized visual analysis of heterogeneous manufacturing data. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 8034 LNCS, pp. 76–85). https://doi.org/10.1007/978-3-642-41939-3_8