The Challenge of Process Data Warehousing

The management of organizational knowledge is becoming a key requirement in many engineering organizations. In many cases, it is difficult to capture this knowledge directly, as it is hidden in the way-of-working followed by networks of highly qualified specialists. Moreover, much of this knowledge is strongly context-dependent, so rules to be followed must be augmented by adequate situation analysis. Hardware and software tools used to support these processes are strongly heterogeneous, involving significant effort of usage and very different kinds of data. In this paper, we propose SURFHVV GDWD ZDUH KRXVHV as a means to remedy these problems. A process data warehouse, according to our approach, is centered around a knowledge-based metadata repository which records and drives a heterogeneous engineering process, supported by selected materialized instance data. We follow a concept-centered approach expanding ideas from the European DWQ project and illustrate our solution with a prototypical process data warehouse for chemical engineering design developed within the Collaborative Research Centre IMPROVE at Aachen University of Technology.