Metal loss determinations based on thermography machine learning approach for insulated structures

Patent metadata

Field	Value
Publication number	US-11112349-B2
Application number	US-201916513168-A
Country	US
Kind code	B2
Filing date	Jul 16, 2019
Priority date	Jul 16, 2019
Publication date	Sep 7, 2021
Grant date	Sep 7, 2021

Abstract

A method for identifying metal wall loss in an insulated metal structure is provided. The method includes receiving thermograms of the outer surface of the structure using an infrared camera, applying filters to the thermograms using a first machine learning (ML) system, determining wall loss classifications based on outputs from the filters, validating the wall loss classifications by inspecting the structure, and training the first ML system using the validation results. Outputs of the first ML system and additional structural and environmental data are input to a second ML system that incorporates information from earlier states into current states. The second ML system is trained to estimate wall loss according to changes in the outputs of the first ML system and the additional data over time until a wall loss classification accuracy is reached. The metal wall loss is thereafter estimated using the first and second ML systems in coordination.

First claim

What is claimed is:

1. A method for inspection of a first insulated structure, the method comprising:
acquiring, by an infrared camera, first thermograms of a first outer surface of the first insulated structure;
determining, by a prediction circuit, a first wall thickness loss of a first defect in a first metal wall of the first insulated structure using the first thermograms, the first metal wall being underneath a first insulation layer of the first insulated structure, the first insulation layer corresponding to the first outer surface; and
outputting, by the prediction circuit, the first wall thickness loss, wherein
the prediction circuit is built from training data using a machine learning process,
the training data comprises second thermograms of a second outer surface of a second insulated structure having at least three distinct known second wall thickness losses of a corresponding at least three second defects in a second metal wall of the second insulated structure, the second metal being underneath a second insulation layer of the second insulated structure, the second insulation layer corresponding to the second outer surface, and the method further comprises:
determining, by a prediction circuit, a first size of the first defect using the first thermograms, the first size comprising one or more of a length, a width, and an area; and
outputting, by the prediction circuit, the first size, wherein
the at least three second defects have a corresponding at least three distinct known second sizes, the second sizes each comprising one or more of a length, a width, and an area.

2. The method of claim 1 , further comprising:
moving, by gripper wheels, the infrared camera from a first location on the first outer surface corresponding to the first thermograms, to a second location on the first outer surface different from the first location, the gripper wheels being attached to the infrared camera and in adhesive contact with the first outer surface; and
acquiring, by the infrared camera, third thermograms of the second location.

3. The method of claim 2 , further comprising:
determining, by the prediction circuit, a third wall thickness loss of a third defect in the first metal wall using the third thermograms, the third wall thickness loss being different from the first wall thickness loss; and
outputting, by the prediction circuit, the third wall thickness loss.

4. The method of claim 2 , wherein moving the infrared camera comprises:
autonomously moving the infrared camera along the first outer surface in a circumferential or longitudinal direction.

5. The method of claim 1 , further comprising:
heating, by a heat source, some of the first outer surface prior to acquiring some of the first thermograms, the heat source being attached to the infrared camera.

6. The method of claim 1 , wherein the first insulated structure comprises:
a pipe including the first metal wall; and
a cladding layer coinciding with or underneath the first outer surface, wherein
the first insulating layer is underneath the cladding layer.

7. A method for inspection of a first insulated structure, the method comprising
acquiring, by an infrared camera, first thermograms of a first outer surface of the first insulated structure;
determining, by a prediction circuit, a first wall thickness loss of a first defect in a first metal wall of the first insulated structure using the first thermograms, the first metal wall being underneath a first insulation layer of the first insulated structure, the first insulation layer corresponding to the first outer surface; and
outputting, by the prediction circuit, the first wall thickness loss, wherein
the prediction circuit is built from training data using a machine learning process,
the training data comprises second thermograms of a second outer surface of a second insulated structure having at least three distinct known second wall thickness losses of a corresponding at least three second defects in a second metal wall of the second insulated structure, the second metal being underneath a second insulation layer of the second insulated structure, the second insulation layer corresponding to the second outer surface, and the machine learning process comprises:
applying, by a training circuit, filters to the second thermograms;
determining, by the training circuit, wall thickness loss classifications of the at least three second defects based on output from the filters; and
validating, by the training circuit, the wall thickness loss classifications using the known second wall thickness losses.

8. The method of claim 7 , wherein the machine learning process further comprises:
outputting, by the training circuit, the validated wall thickness loss classifications; and
building, by a machine learning circuit, the prediction circuit from the validated wall thickness loss classifications.

9. A system for inspection of a first insulated structure, the system comprising:
an infrared camera for acquiring first thermograms of a first outer surface of the first insulated structure; and
a prediction circuit for
determining a first wall thickness loss of a first defect in a first metal wall of the first insulated structure using the first thermograms, the first metal wall being underneath a first insulation layer of the first insulated structure, the first insulation layer corresponding to the first outer surface, and
outputting the first wall thickness loss, wherein
the prediction circuit is built from training data using a machine learning process, and
the training data comprises second thermograms of a second outer surface of a second insulated structure having at least three distinct known second wall thickness losses of a corresponding at least three second defects in a second metal wall of the second insulated structure, the second metal being underneath a second insulation layer of the second insulated structure, the second insulation layer corresponding to the second outer surface,

wherein the system further comprises a heat source for heating some of the first outer surface prior to acquiring some of the first thermograms, the heat source being attached to the infrared camera.

10. The system of claim 9 , further comprising:
gripper wheels for moving the infrared camera from a first location on the first outer surface corresponding to the first thermograms, to a second location on the first outer surface different from the first location, the gripper wheels being attached to the infrared camera and in adhesive contact with the first outer surface, wherein
the infrared camera is further for acquiring third thermograms of the second location.

11. The system of claim 10 , wherein the prediction circuit is further for:
determining a third wall thickness loss of a third defect in the first metal wall using the third thermograms, the third wall thickness loss being different from the first wall thickness loss; and
outputting the third wall thickness loss.

12. The system of claim 10 , wherein the gripper wheels are further for:
autonomously moving the infrared camera along the first outer surface in a circumferential or longitudinal direction.

13. The system of claim 10 , further comprising:
a sliding carrier for attaching the infrared camera to the gripper wheels;
slide guides for guiding movement of the sliding carrier along the first outer surface in a circumferential or longitudinal direction; and
suction pads for stabilizing the slide guides with respect to the first outer surface.

14. The system of claim 9 , wherein the prediction circuit is further for:
determin

Assignees

• Saudi Arabian Oil Co

Inventors

• Amer Ayman
• Shehri Ali Al
• Cunningham Vincent

Classifications

• G01N17/04 Primary

  Corrosion probes · CPC title

• G01N25/72 Primary

  Investigating presence of flaws · CPC title

• G06F18/24

  Classification techniques · CPC title

• G01N17/006

  of metals · CPC title

• G01N17/00

  Investigating resistance of materials to the weather, to corrosion, or to light · CPC title