Authors: Klaus Weidenhaupt, Matthias Jarke, Thomas List
Tags: 1999, conceptual modeling
The process industries (chemicals, food, oil,…) are characterized by — continuous or batch — processes of material transformation. The design of such processes, and their mapping to the available equipment (plants composed of production units in which reactions take place), is a complex process that determines the competitiveness of these industries, as well as their environmental impact. In cooperation with researchers and industry from chemical engineering, we have developed the idea to capture and evaluate the experiences gained about process designs in so-called process data warehouses. The data sources for such process data warehouses are highly heterogeneous tools, e.g. for conceptual design (termed flowsheeting in chemical engineering), for mathematical simulations of large non-linear differential equation systems, for measurements gained with experimental usage of equipment at small scale or in original size, or even from molecular modeling. The clients of a data warehouse are interested in operational data transfer as well as experience analysis (pattern detection, process mining) and reuse. Starting from an empirical analysis of the requirements for a process data warehouse, the paper describes the solution architecture we are pursuing, the models driving the approach, and the status of a prototypical implementation we are undertaking. The prototype links commercial components operationally through advanced wrapping techniques, where the workflow is determined by constraint analysis in a logic-based meta model and executed through a process-integrated modeling environment. In the conclusions, we point out what can be learned from this work for conceptual modeling in general.Read the full paper here: https://link.springer.com/chapter/10.1007/3-540-47866-3_35
