Identification and Retrieval of Relevant Information for Instantiating Digital Twins During the Construction of Process Plants

Abstract

While volume-driven industries such as automotive are characterized by a high degree of data backflow across all production cycles, there is still a certain residue in the planning and construction of process plants. This is firstly due to the high proportion of customer-specific requirements and secondly to the significant amount of value added on site during construction. To handle recurring project-specific process plants as time- and cost-efficiently as possible, optimal information exchange among contractors of various disciplines and the plant developer is a prerequisite. For this purpose, a holistic digital representation of the plant is created, which consolidates all relevant information in one place serving as a foundation of multiple digital twins. An approach to identify and define relevant information depending on their subsequent use is developed. On this basis, a framework is proposed to enable a multipliable BOM-based automatic definition of information backflow to instantiate digital representations in parallel to the planning and construction process. Furthermore, project-specific contextual information will be captured and referenced in a structured form preventing their loss for subsequent similar projects.

Automatic knowledge graph creation from engineering standards using the example of formulas

“…Information extraction aims at capturing and extracting useful information from unstructured or semistructured data, such as text, images, audio, or video, whereas extraction techniques are divided into rule-based and learning-based approaches (Adnan and Akbar, 2019). Examples for extracting information from design relevant sources range from patents (Siddharth et al, 2022) and engineering standards -including requirements (Luttmer et al, 2023) as well as context and user specific information (Ehring et al, 2023;Layer et al, 2023) -to circular design guidelines (Gräßler and Hesse, 2023). While different research papers on extraction of valuable information from engineering standards exist, the automatic aggregation to KGs has not been investigated in detail.…”

Introducing a multipliable BOM-based automatic definition of information retrieval in plant engineering

“…Information backflow, in this context, pertains to capturing and managing essential data emanating from production, construction, commissioning phases, and plant operation to guarantee accurate representation of traceability or Digital Twins of the physical plant and its components. The proposed framework foundation identified different categories of information backflow, each necessitating predefined specific information retrieval methods based on a general architecture (Layer et al, 2023c).…”

“…Furthermore, it enables plant designers and engineers to modify and optimize information retrieval processes for different projects based on a reference, helping to reduce inefficiencies and surplus efforts. This paper builds up on a conducted systematic review of data management in process industry (Layer et al, 2023a), a generic approach of pre-defining information retrieval (Layer et al, 2023c) and a first approach to automatically identify relevant information (Layer et al, 2023b). A novel framework is investigated, that focuses on the automatic definition of information retrieval by utilizing so-called Instantiation Blocks (IB), acting as placeholders and as a definition of to-be instantiated information, both connected to the Bill of Material (BoM) of process plants as shown in Figure 1A.…”