Search

EP-4738175-A1 - METHOD FOR OBTAINING A TOPOLOGIC MODEL OF ONE OR MORE COMPONENTS OF AN INDUSTRIAL PLANT

EP4738175A1EP 4738175 A1EP4738175 A1EP 4738175A1EP-4738175-A1

Abstract

The disclosure relates to a method (100) for obtaining a topologic model of one or more components of an industrial plant, the method comprising: - (102) obtaining image data of one or more components of the industrial plant; - (103) providing the image data of the one or more components of the industrial plant to an AI processing component (44) for processing; - (104) receiving geometry data of the one or more components of the industrial plant based on the processed image data; and - (105) obtaining by the AI processing component (44) the topologic model of one or more components of the industrial plant based on the geometry data of the one or more components, wherein the topologic model comprises information about a physical arrangement of the one or more components within the industrial plant.

Inventors

  • SCHOCH, Nicolai
  • PRETLOVE, JOHN
  • Gruener, Sten
  • HOERNICKE, MARIO
  • BORRISON, Reuben

Assignees

  • ABB SCHWEIZ AG

Dates

Publication Date
20260506
Application Date
20241104

Claims (15)

  1. Method (100) for obtaining a topologic model of one or more components of an industrial plant, the method comprising: - (102) obtaining image data of one or more components of the industrial plant; - (103) providing the image data of the one or more components of the industrial plant to an AI processing component (44) for processing; - (104) receiving geometry data of the one or more components of the industrial plant based on the processed image data; and - (105) obtaining by the AI processing component (44) the topologic model of one or more components of the industrial plant based on the geometry data of the one or more components, wherein the topologic model comprises information about a physical arrangement of the one or more components within the industrial plant.
  2. The method (100) of claim 1, wherein the topologic model further comprises information about physical connections and/or input/outputs of the one or more components within the industrial plant.
  3. The method (100) according to according claim 1 or 2, wherein the image data comprises geo-spatial information in the form of one or more of GPS positions, rotation angle, acceleration, and speed of movement.
  4. The method (100) according to claim 3, wherein the image data further comprises geo-location tags and the method further comprises: - determining a distance between the one or more components of the industrial plant with at least another component of the industrial plant.
  5. The method (100) according to claim 4, wherein the method further comprises: - determining a transport time of a fluid or gas between at least two components of the industrial plant based on the determined distance between the components.
  6. The method (100) according to claim 4, wherein the method further comprises: - assembling the topologic model of the one or more components to a topologic plant model of the industrial plant.
  7. The method (100) according to according to any of the preceding claims, wherein a Generative AI model, is used to output the generated topologic model into a standardized format.
  8. The method (100) according to claim 7, wherein the topologic model is stored in a database.
  9. The method (100) according to claim 8, wherein the topologic model is compared to a previous topologic model.
  10. The method (100) according to claim 9, wherein the AI processing component (44) uses graph algorithms to identify similarities between the topologic model and the previous topologic model.
  11. The method (100) according to any of the preceding claims, wherein the image data is associated with existing plant model data, wherein the existing plant model data comprises at least one or more of: P&ID diagrams, wiring diagrams, Human-Machine Interface (HMI) data, or legacy data models.
  12. The method (100) according to any of the preceding claims, wherein the AI processing component (44) generates an uncertainty score for the topologic model, indicating the confidence level of the generated model.
  13. The method (100) according to any of the preceding claims, wherein the method further comprises contextual verification of the topologic model with an identified adjacent component.
  14. One or more computer program products (34, 40) comprising instructions which, when executed by one or more data processing apparatuses (30), cause the one or more data processing apparatuses (30) to carry out the method (100) of any one of the previous claims.
  15. A data processing system (50) configured to carry out the method (100) of any one of claims 1 to 13.

Description

TECHNICAL FIELD The present invention relates to a method for obtaining a topologic model of one or more components of an industrial plant, one or more computer program products and a data processing system. BACKGROUND Industrial plants, such as those in the oil and gas, chemical, mining, and water treatment industries, comprise numerous interconnected components, including tanks, pipes, valves, controllers, and other critical equipment. Accurate documentation of the physical layout and interconnections of these components is essential for design, maintenance, safety compliance, and operational efficiency. Traditionally, the creation and maintenance of such documentation rely on detailed engineering drawings, including Piping and Instrumentation Diagrams (P&IDs), wiring diagrams, and control system Human-Machine Interfaces (HMI). In existing practices, obtaining an accurate "as-built" digital model of an industrial plant, especially in brownfield facilities (existing plants with potential undocumented modifications), poses significant challenges. Legacy documentation may be incomplete, outdated, or stored in inaccessible formats, complicating efforts to generate comprehensive and up-to-date digital representations. Moreover, the manual process of updating and verifying these models is labor-intensive, time-consuming, and prone to human error, exacerbated by the demographic shift where experienced maintenance and operations personnel are retiring, and there is a shortage of skilled replacements. Advanced technologies such as 3D laser scanning and Building Information Modeling (BIM) have been employed to create digital representations of industrial plants. However, these solutions are often expensive, require specialized equipment, and may not integrate seamlessly with existing engineering documentation or control systems. SUMMARY The above problem or need is at least partially solved or alleviated by the subject matters of the independent claims of the present disclosure, wherein further examples are incorporated in the dependent claims. According to an aspect of the present disclosure, there is provided a method for obtaining a topologic model of one or more components of an industrial plant, the method comprising: obtaining image data of one or more components of the industrial plant;providing the image data of the one or more components of the industrial plant to an AI processing component for processing;receiving geometry data of the one or more one or more components of the industrial plant based on the processed image data; andobtaining by the AI processing component the topologic model of one or more components of the industrial plant based on the geometry data of the one or more components, wherein the topologic model comprises information about a physical arrangement of the one or more components within the industrial plant. A topologic model in the context of industrial plant engineering may refer to a digital representation that captures and visualizes the physical arrangement of one or more components within an industrial plant. The topologic model may depict the spatial positions of individual components such as but not limited to machinery, tanks, pipes, valves, and control units within the industrial plant. The topologic model of the one or more components may e.g. comprise an orientation and alignment of the one or more components relative to each other, ensuring an accurate representation of the plant's layout. Thus, the topologic model may provide a foundational structure that illustrates how various components are distributed and oriented within the industrial plant, facilitating a clear understanding of the overall physical setup. Image data may comprise a broad spectrum of visual information of the one or more components or parts of an industrial plant. For example, image data may comprise 2D visual information such as but not limited to sequential photographs, video frames, panoramic shots, scans, or annotated images of the one or more components. These images may be acquired from imaging sensors, such as but not limited to cameras, wearable devices, drones, or stationary cameras. The image data may be acquired from different perspectives. By acquiring different or multiple perspectives, more comprehensive image data to accurately reconstruct depth, shape, and structure of the one or more components may be generated, for example. For example, sequential photographs or video frames may be captured as a person or a drone navigates through the industrial plant. The acquisition of image data may be facilitated through the integration of an AI assistance system. For example, the acquisition process by a person may be guided. Thus, a person may be provided with real-time instructions on which areas or components of the industrial plant should be covered and from which perspectives the images or videos should be taken. For generating a basic topologic model showing primary process elements, su