Abstract
The design of complex engineering systems is an often long and articulatedprocess that highly relies on engineers' expertise and professional judgment.As such, the typical pitfalls of activities involving the human factor oftenmanifest themselves in terms of lack of completeness or exhaustiveness of theanalysis, inconsistencies across design choices or documentation, as well as animplicit degree of subjectivity. An approach is proposed to assist systemsengineers in the automatic generation of systems diagrams from unstructurednatural language text. Natural Language Processing (NLP) techniques are used toextract entities and their relationships from textual resources (e.g.,specifications, manuals, technical reports, maintenance reports) availablewithin an organisation, and convert them into Systems Modelling Language(SysML) diagrams, with particular focus on structure and requirement diagrams.The intention is to provide the users with a more standardised, comprehensiveand automated starting point onto which subsequently refine and adapt thediagrams according to their needs. The proposed approach is flexible andopen-domain. It consists of six steps which leverage open-access tools, and itleads to an automatic generation of SysML diagrams without intermediatemodelling requirement, but through the specification of a set of parameters bythe user. The applicability and benefits of the proposed approach are shownthrough six case studies having different textual sources as inputs, andbenchmarked against manually defined diagram elements.