Intelligent authoring system based on DEXPI for the digitalization of P&IDs [P&ID 디지털화를 위한 DEXPI 기반 지능형 저작 시스템 개발]

Abstract

The digitization of an image format piping and instrumentation diagram (P&ID) entails three steps: Recognizing information objects constituting the diagram, identifying the connection relationships between the recognized objects, and generating a digital P&ID. In this study, a method of generating a digital P&ID by receiving an object adjacency graph (recognized objects and connection relationship between the objects), storing it according to the DEXPI de-facto standard, and calculating various data required in the field is suggested. An intelligent authoring system that supports P&ID digitization was designed based on the proposed method, and major automation functions were developed. An experiment was performed to generate digital P&IDs from object association graphs for two sample diagrams to verify the developed system prototype. As a result of the experiment, digital P&IDs could be generated within 50 min for both samples. © 2021 The Korean Society of Mechanical Engineers.