Ensuring Asset Optimization through Ontology-based Modelling of Technical Documentation

The physical or engineering assets within modern industries have consistently been considered as the core elements of significant value that constitute the backbone for success and overall growth. Ensuring the optimal performance of an asset is a crucial and challenging task, since any breakdown or downtime may have a large impact on product quality, cost of production, operation effectiveness, or even cause health, safety and environmental issues. The performance of any asset within an industrial environment is greatly dependent on the overall usage and maintenance instructions provided by the asset manufacturer to the user. These instructions have always been in the form of a technical documents compilation that accompanies the physical asset. These documents describe the product in a complete manner and often constitute the first line of support when asset operators or maintenance engineers need help with a problem or when they seek to advance their use of an asset. The documentation inherently dictates how the asset should be overall managed in order to achieve the optimal performance. The motivation for this work lies in the fact that traditional technical documentations mainly consist of textual and graphical documents with information that is not being properly put to effective use. Although there has been great progress in newer electronic versions, essentially it is still an arduous and time-consuming process for the users to identify specific answers to their problems or questions. It often requires considerable effort to locate much needed information, especially in case of unexpected emergencies. The documentation is often inaccessible, outdated, has a poor structure and language, and may assume knowledge that the readers don't essentially possess. Therefore, the information included in the documentation is not always being properly and fully consumed by the asset users. Overall, there is a gap between the original documentation instructions and the actual usage that may prevent the assets’ overall effectiveness from reaching its full potential and may have great repercussions such as damage to the assets, the environment, or even injury to the personnel[6]. This dissertation proposes an ontology-based approach to represent the technical documentation content concerning the asset’s overall usage and maintenance. Ontologies possess high expressive power and degree of formality and most importantly they can describe the content in a machine understandable manner. The approach proposes using ontologies towards efficiently processing and consuming the documentation content in order to confirm that the asset is operated and maintained according to the manufacturer’s instructions. The Technical Documentation Ontology model is developed to represent all the important information contained in the documentation concerning how the asset should be managed. The approach subsequently proposes the implementation of a rule-based mechanism to check the real asset information that is collected during the actual asset usage against the documentation instructions. By comparing the real asset data and the respective instructions, it is possible to identify abnormal behaviors or deviations and recommend appropriate usage and maintenance actions towards ensuring or even optimizing the asset performance. Finally, the approach is validated and evaluated within a case study implemented on a packaging machine from a major Swiss company.