Engineering Data Management Using Artificial Intelligence

Abstract

A lot of data in the engineering world exists in the form of paper drawings and documents. Technically, these are considered as "unstructured data [1]"as it is difficult to extract content from the drawings using traditional programs as compared to data stored in databases. These drawings are often used for design and maintenance activities in both greenfield and brownfield projects. Today, digital is a key enabler in oil and gas to increase workforce efficiency. Hence there is a growing need to get the dumb drawings digitized. However, the only means of converting these drawings into digital format is to manually re-draw them. With the emergence of technologies like Computer Vision, Optical Character Recognition(OCR) and Natural Language Processing(NLP), we no longer need to depend on human cognitive capabilities to process information from a drawing. Artificial Intelligence(AI) systems can be trained to recognize the visual content in drawings and provide a simplified context. AI based algorithms can read a scanned Process and Instrumentation Diagram (P&ID) to recognize the graphical content of the drawing like instruments, tags, pipelines, text etc. The information extract that AI generates from a dumb drawing can later be passed to an automation script to create a new digital version. This paper emphasizes the use of Artificial Intelligence in processing a scanned drawing and automatically redraw it on a digital platform. Adapting this approach can bring considerable advantage in the pursuit of going digital.