In association with the OGI Pilot, this project aims to develop flexible data transformation software to enable operators of capital facilities, such as an oil refinery, make effective use of data. Currently, necessary data is hidden in siloed information systems, causing enormous costs due to data errors and inefficiencies. Therefore, interoperability solutions are required to alleviate these costs and provide other benefits, such as reduced risk, to owner/operators, suppliers, and regulatory agencies engaged in the management of critical infrastructure. This project will further develop a software interoperability platform, the UniSA Transformation Engine, to bridge information silos by enabling the uploading of large volumes of raw data in different formats and different suppliers into the data management system and translated into a standardised format.
Researchers: Prof. Markus Stumptner, Dr. Georg Grossmann, Dr. Wolfgang Mayer, Dr. Matt Selway
The aim of the Oil & Gas Interoperability Pilot (OGI Pilot) is to demonstrate data and software interoperability through software ecosystems using open, vendor neutral standards. The result will bridge the currently siloed information systems of the Engineering/Procurement/Construction (EPC) companies (who typically use ISO 15926) and those of Operations & Maintenance organisations (using MIMOSA CCOM) currently costing an estimated 3-6% of the lifetime cost of a capital facility (such as an industrial plant, refinery or drilling rig) in overheads or inefficiencies (costs of about $15Bn across the USA in 2002). The OGI Pilot uses the example of an oil refinery in conjunction with 11 industry-defined use cases, which cover the information keeping requirements across the full range of life-cycle activities of industrial plants, platforms, and facilities. The primary use case is for Continuous Information Handover from EPC to Operator, which will alleviate the costs and errors associated with the currently manual process of transferring plant information between EPC and O&M organisations as PDF files. With interoperability, Digital Handover can occur that allows the detailed CAD models produced during design to be used across the entire plant life-cycle, thereby removing obsolescence problems when the plant undergoes maintenance and improving plant efficiency and safety. The outcomes of the OGI Pilot include:
- software tools
- demonstrations of live data exchange on real-life industrial data
- guidelines for validating the mappings
- a set of reference data
A live demonstration of the OGI Pilot work and the UniSA Transformation Engine is available on YouTube
Standards-Driven Data Transformation
As more and more complex data models are put into use, more powerful techniques to represent mappings between models are required. These mappings are complex mappings need to be represented as data structures in their own right, an approach known as Model-Driven Engineering (MDE). The UniSA Transformation Engine is built on MDE principles and research into MDE-based mappings that allow the definition, manipulation, execution, and export of fully declarative mapping patterns at the metamodel level. However, the models for which the mappings are being developed have multiple meta-levels internally, which traditional software modelling languages such as UML have restricted ability to handle. This restriction has driven the recent research in the area of Multi-Level Modelling (note: link to our MLM page).
We are applying multi-level modelling techniques in the context of interoperability with the goal of incorporating such approaches into the UniSA Transformation Engine. This will allow the improved handling of different language constructs in different domains and provide clear definitions for open standards: finally bridging the heterogeneity between ISO 15926 and MIMOSA CCOM. Moreover, these mappings require languages powerful enough to capture domain semantics such as constraints between units of measurement or on the selection of appropriate parts in a design. Therefore, the approach multi-level models will be anchored in semantic representations such as ontologies, integrating concerns that would normally be handled through different languages in MDE. The resulting unified approach will be simpler while supporting complex validation of models and their mappings.
Finally, the UniSA Transformation Engine will be enhanced with bi-directional transformation capabilities. Rather than just EPC to O&M transformations, bi-directional transformations will support the backward propagation of model changes to ensure all-round consistency of the ecosystems. Moreover, it will extend the capabilities of the system to additional use cases of the OGI Pilot.