Welding Motion Data Correlation for Quality Control and Skill Transfer
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing welding quality control methods struggle to accurately quantify and instruct welding skills due to the tacit nature of skilled technicians' knowledge, leading to inefficiencies in skill transfer and potential loss of expertise.
Innovation Solution
A welding work data accumulation device that measures and analyzes welding motions and phenomena, creating a database to extract feature amounts and correlate them with welding quality parameters, enabling precise quality control and skill instruction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If skilled technicians directly instruct skills to unskilled operators through observation and imitation, then skill transfer occurs, but the tacit knowledge is difficult to quantify and verify, leading to imprecise instruction
Solution Approach 1:
The patent replaces the mechanical observation and imitation system with an optical measurement system. Cameras and sensors capture welding motions and phenomena, converting tacit knowledge into digital data that can be precisely analyzed and instructed, eliminating the imprecision of human observation
Solution Approach 2:
The patent introduces measurement instruments and data processing systems as intermediaries between the skilled technician's actions and the unskilled operator's learning. These intermediaries objectively capture and transmit skill information, preventing information loss while maintaining measurement precision
2Adaptability or versatility
If multiple skilled technicians work with different motion indicators, then diverse expertise is utilized, but the unskilled operator cannot learn effectively because there is no unified standard to imitate
Solution Approach 1:
The patent changes the parameter representation of welding skills from subjective motion indicators to objective measured parameters. By standardizing skill description in terms of measurable quantities (positions, speeds, temperatures), the system maintains adaptability to different technicians while ensuring consistent welding quality through precise parameter control
3Productivity
If skill instruction relies on observation and imitation without systematic classification, then learning occurs through experience, but it takes excessive time for unskilled operators to become skilled
Solution Approach 1:
The patent segments welding skills into discrete, measurable components such as torch position, movement speed, and welding parameters. This segmentation allows unskilled operators to learn specific skills independently rather than observing entire complex motions, dramatically reducing training time while maintaining productivity
4Reliability
If measurement data is used for quality control, then objective evaluation is achieved, but the system cannot provide targeted correction motions for skill improvement
Solution Approach 1:
The patent implements a feedback system where measurement data not only evaluates welding quality but also generates correction motions. The system compares actual welding motions with standard motions, identifies deviations, and provides targeted correction instructions, making skill improvement as easy as following guided corrections while maintaining reliable quality control
Data Source
Figure 1
Figure 2~3
Figure 4~6
AI summary
An object of the present invention is to appropriately perform a welding quality control. Therefore, a welding work data accumulation device (100) includes: a measurement unit (4, 5, 7, 9, 11, and 16) that measures a welding motion and a welding phenomenon when a welding operator (1) grips a welding torch (2) and performs welding on a welded body (3); a data analysis unit (14) that extracts an appropriate combination of a welding motion feature amount (Tw, Ht, and Sp) and a welding phenomenon feature amount (Iw, S, and Ssym) in correction with time or coordinates based on data acquired by the measurement unit (4, 5, 7, 9, 11, and 16); and a data accumulation unit (15) that creates a database (70) based on an extraction result of the data analysis unit (14).