A potential solution for automating the generation of MFM models is the semantical data extraction from plant documents to MFM. Two types of documentation are regarded as major modelling resources: diagram-based (e.g. P&IDs) and text-based (e.g. SOPs). The former is usually used to extract functions that are inherent in the devices and hence independent of domains. It is necessary to establish relationships between standard plant drawing symbols (e.g. ISO 10628) and MFM semantics in order to model patterns, relations, and structures in MFM. Text-based documents can be used to identify environment-oriented functions and means-ends interactions among functions. Identifying these functions and interactions is necessary for composing complete models.
IEC 62414 - Computer Aided Engineering Exchange (CAEX) is an XML-scheme neutral data model that can provide a basic format for electronical exchange between P&IDs and MFM. It can be a tool to implement the concept of modelling library. By additionally using natural language process (NLP) to analyse text-based documents, it is possible to automatically transmit functional knowledge from heterogeneous engineering disciplines to modelling, and further to reasoning applications.