Advanced Integration of Discrete Line Segments in Digitized P&ID for Continuous Instrument Connectivity
This addresses the challenge of digitizing plant blueprints for engineers and operators, but it appears incremental as it builds on existing computer vision detection methods.
The paper tackles the time-consuming and error-prone manual mapping of Piping and Instrumentation Diagrams (P&IDs), which takes 3-6 months, by developing a method to merge detected line segments to create a digitized P&ID for continuous instrument connectivity.
Piping and Instrumentation Diagrams (P&IDs) constitute the foundational blueprint of a plant, depicting the interconnections among process equipment, instrumentation for process control, and the flow of fluids and control signals. In their existing setup, the manual mapping of information from P&ID sheets holds a significant challenge. This is a time-consuming process, taking around 3-6 months, and is susceptible to errors. It also depends on the expertise of the domain experts and often requires multiple rounds of review. The digitization of P&IDs entails merging detected line segments, which is essential for linking various detected instruments, thereby creating a comprehensive digitized P&ID. This paper focuses on explaining how line segments which are detected using a computer vision model are merged and eventually building the connection between equipment and merged lines. Hence presenting a digitized form of information stating the interconnection between process equipment, instrumentation, flow of fluids and control signals. Eventually, which can be stored in a knowledge graph and that information along with the help of advanced algorithms can be leveraged for tasks like finding optimal routes, detecting system cycles, computing transitive closures, and more.