Image-Guided Automated Welding for Precise Defect Repair

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Many existing welding systems, particularly those used for airfoil repair, are not fully automated, leading to inefficiencies and excessive process scrap due to the need for frequent re-training of welders and lack of precision in identifying areas to be welded.

Innovation Solution

An automated welding system that uses an imaging device to acquire images of the object, analyze pixels to identify defects and predefined patterns, and generate control instructions for the welding tool and mounting platform to accurately weld damaged areas while excluding as-built features from the welding process.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If manual welding operations are used with trained welders, then flexibility and adaptability are maintained, but productivity is reduced and process scrap increases

Engineering Contradiction:
Improvewelder adaptabilityVSAvoidwelding productivity
Core Design Contradiction:
Adaptability or versatilityVSProductivity

Solution Approach 1:

The welding system performs self-inspection and self-positioning through the imaging device and controller, automatically identifying defects and determining weld parameters without human intervention, thereby eliminating the need for re-training while maintaining adaptability

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent replaces manual visual inspection and decision-making with an automated imaging and control system that captures images, processes them through a controller, and automatically generates welding instructions, substituting human cognitive functions with mechanical/optical systems

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Productivity

If automated welding is implemented without image analysis, then productivity increases, but manufacturing precision deteriorates due to inability to identify specific defect areas

Engineering Contradiction:
Improvewelding automationVSAvoiddefect identification accuracy
Core Design Contradiction:
ProductivityVSManufacturing precision

Solution Approach 1:

The imaging device serves as an intermediary between the welding system and the workpiece, capturing visual information about defects and transmitting it to the controller, which then processes the images to precisely identify defect locations and characteristics before generating welding instructions

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The system performs preliminary image capture and analysis before the welding operation begins, identifying all defect areas and preparing welding instructions in advance, which enables precise automated welding without real-time human intervention

Inventive Principle:
Principle #10Preliminary action

3Reliability

If comprehensive welding of all areas is performed, then reliability improves by addressing all potential defects, but loss of substance increases due to unnecessary welding of non-defect areas

Engineering Contradiction:
Improvewelding reliabilityVSAvoidmaterial waste
Core Design Contradiction:
ReliabilityVSLoss of substance

Solution Approach 1:

The imaging and control system enables localized welding by precisely identifying defect locations and generating welding instructions only for those specific areas, allowing different parts of the workpiece to receive different treatments (welded or non-welded) based on their actual condition rather than applying uniform welding across all areas

Inventive Principle:
Principle #3Local quality

Data Source

PatentEP3417974B1System, article and method for automated welding
Publication Date: 2024.03.20 GENERAL ELECTRIC CO
  • EP3417974B1 patent drawingFigure 1
  • EP3417974B1 patent drawingFigure 2
  • EP3417974B1 patent drawingFigure 3

AI summary

An automated welding system 100 includes a mounting platform 104 configured to receive an object 112, a welding tool 106, an imaging device 108 configured to acquire at least one image 308 associated with the object 112, and a controller 110. The controller 110 is configured to receive the at least one acquired image 308, analyze at least one pixel in the at least one acquired image 308, identify, based upon the analyzing, an area to be welded 126 in the at least one acquired image 308, wherein the area to be welded 126 includes a defect, and generate, based upon the identifying, control instructions for controlling at least one of the mounting platform 104 and the welding tool 106 to weld the area to be welded 106.