AI Stud Welding Discontinuity Detection and Corrective Action
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for identifying discontinuities and anomalies in joinings, such as stud welding and riveting, are unreliable due to human error and subjective interpretation, leading to delays and increased costs in manufacturing.
Innovation Solution
A computer-based system utilizing artificial intelligence models to analyze real-time data from joining machines, including voltage, current, and lift values, to detect and classify discontinuities and anomalies, and generate corrective or preventative actions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If visual inspection by technicians is used to identify discontinuities and anomalies, then the method is simple and requires minimal equipment, but the reliability is low due to human error and subjective interpretation
Solution Approach 1:
The patent replaces the mechanical visual inspection system with an automated computer-based vision system that captures images of joinings and uses machine learning models to detect discontinuities and anomalies. This substitution eliminates human error and subjective interpretation while maintaining detection capability through automated image analysis algorithms.
Solution Approach 2:
The system enables self-inspection by the joining machine itself through integrated imaging sensors and automated analysis. The computer model automatically processes images captured during or after the joining process, eliminating the need for separate manual inspection steps and providing immediate feedback on joining quality.
2Measurement precision
If visual inspection is performed at rapid pace during assembly, then productivity is maintained, but the detection precision deteriorates due to human limitations
Solution Approach 1:
The automated vision system operates continuously during the joining process, capturing and analyzing images in real-time without interruption. The system processes multiple images per minute at speeds matching or exceeding assembly rates, maintaining continuous detection capability that does not slow down productivity while achieving superior precision through automated analysis.
Solution Approach 2:
The system creates digital copies of the joining through high-speed imaging, allowing detailed analysis of the copied image data without physically interfering with the joining process. This enables precise anomaly detection in the digital domain while the physical assembly continues at full speed.
3Loss of time
If conventional inspection methods are used, then the process is simple to implement, but the loss of time occurs due to delays in manufacturing processes and vehicle recalls
Solution Approach 1:
The system performs inspection during or immediately after the joining process itself, rather than as a separate subsequent step. Images are captured and analyzed in real-time, allowing immediate detection of defects before they propagate through the manufacturing process or reach the customer, thereby eliminating delays and recall risks.
Solution Approach 2:
The automated system provides immediate feedback on joining quality by analyzing images and detecting anomalies in real-time. This feedback loop allows for instant corrective action if defects are detected, preventing defective parts from advancing to subsequent manufacturing stages and reducing overall cycle time through proactive quality control.
Data Source
AI summary
Disclosed herein are systems and methods for identifying welding anomalies and discontinuities in stud welding using AI models. Instead of conventional welding accuracy methods (e.g. destructive and/or image generation methods) a processor may communicate with one or more sensors associated with a joining machine to retrieve joining data and attributes. The processor may then execute an AI model that is trained based on previously performed stud welding, their corresponding welding attributes, and their corresponding discontinuities and/or anomalies. The processor may execute the AI model using data retrieved from the sensors and may calculate a likelihood of a discontinuity and discontinuity attributes, such as, location, depth, and the like. The processor may also execute a second AI model to identify an appropriate course of action to remedy the identified/predicted discontinuity.