An ontology design enhancing the information transfer between project delivery and asset management

Abstract

As digitization moves on, more data is produced and the value of this data increases. However, data is still transferred manually between different digital systems. Reason to manually transfer information is lacking interoperability between these systems. Huge costs are related to this lacking interoperability as they cause high data losses. Issues with interoperability and information losses are present as well in the field of asset lifecycle information management. Especially in the transition of project delivery to asset management.
This research investigated the stated problem within the Engineering office of the municipality of Rotterdam. The research followed a design science research approach consisting of five steps. First the scientific knowledge base was defined, second the local practice is researched. Third design requirements were listed based on the pre-taken steps. Fourth an ontology was designed based on these requirements. Fifth, based on the investigated scientific literature into the topic the following objective was constructed:
“Design an ontology linking assets and their related information to BIM data-schemes and domain ontologies in a general way.”

Based on the analyses of the scientific knowledge base and the in-depth research cases a list of design requirements was established. To optimize the value of the ontology additional requirements from the FAIR principles were derived. Based on the derived requirements an ontology was designed with the main features as showed below. A main limitation of the research is that the design is not tested or evaluated extensively.
• Objects are central in the design, meaning, all information should be linked, eventually indirectly, to an object.
• Information type is included as a specific entity to provide structure in the information.
• Information related to an information type can be expressed in attributes, documents, and activities.
• The information inside a document can be made explicit in attributes and activities.
• The modelling references can refer to a specific object in a specific BIM-model or to meta-object of a specific BIM data-scheme.
• The domain ontologies refer always to a specific instance within such an ontology.
• Findability and accessibility are incorporated by the means of a separate regulation entity to provide flexibility on the abstraction level of regulation.
The research has some clear contribution to as well the scientific knowledge base as the local practice. Scientific contributions of this research are the designed ontology and the analysing made of the FAIR principles in relation to their application on an ontology for governmental data instead of scientific data. It should be mentioned that no extensive literature research is performed into an earlier application of this kind. Regarding the local practice, the main recommendation, next to the designed ontology, is to state explicit what information is in what documentation. This makes it easier to align the information available and needed and thereby enhances an early consideration on the information transfer process.
To conclude some recommendations for future research are presented. Future research could validate the findings of this research by investigating the information transfer withing the engineering of Rotterdam or strengthen the generalizability by doing the investigation within other major Dutch municipalities or other organizations dealing with the same problem. In addition, the ontology design could be tested and validated. Finally, it is recommended to do research in the application of software analysing documentation and able to extract explicit attributes and activities within these documents. The current developments regarding artificial intelligence and language models may offer opportunities to derive explicit knowledge from documents and transfer it in an explicit way solving a part of the problem at hand.