Looking at Process and Automation Engineering (P&AE) today, for the technically adept engineer, there are many different tools available to support the engineering work from translation of engineering intentions into module and plant descriptions, to definition and parametrization of entire process plant setups, for export to a control system. However, still today, in the very early engineering phases, engineering intentions either need to be entered already in a structured and controlled expert language or require a human expert's manual efforts for translation from unstructured language into formalized representations, in order for thereon-based consistent further processing in the existing tools. This process is time-consuming, fuzzy, and error-prone due to potential misconceptions and ambiguities, even for domain experts. In this work, we therefore present our NL2IBE Tool, which makes use of modern Natural Language Processing in combination with Ontology Mining, and which, based on and controlled by an underlying ontology, allows for the deterministic transformation of natural language intentions into structured and consistent engineering artefacts. We describe the overall tool architecture as well as crucial functionalities and implementation features, followed by an evaluation by the example of a hydrogen generation and CCSU use case. We conclude with a discussion of the proposed tool and give an outlook on future research. (Abstract)
