Weld Torch Angle Correction Using 3D Seam Geometry
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Programming motion trajectories for collaborative robots in welding or cutting is complex, particularly in setting and programming torch angles and orientations along weld joints.
Innovation Solution
A robotic welding system equipped with a weld angle correction tool that uses a depth camera to acquire stereoscopic depth image data, allowing for the calculation and correction of torch angles based on user-recorded positions and pre-stored ideal angles.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a human user manually positions the torch at various points along the joint to train the robot, then the robot can learn the weld trajectory, but the user must be particularly careful about positioning the torch angles (push angle and work angle) precisely, which increases the complexity and difficulty of the training process
Solution Approach 1:
The patent replaces manual mechanical positioning of the torch with an automated vision-based system. A depth camera captures stereoscopic images of the weld joint, and computer vision algorithms automatically calculate the joint geometry and determine the correct torch angles. This substitutes the mechanical skill of manual angle positioning with automated optical measurement and computational geometry processing.
Solution Approach 2:
The system enables self-service by allowing the robot to automatically determine its own torch angles through vision-based measurement. The depth camera and processing algorithms work together to autonomously calculate the push angle and work angle without requiring human expertise in angle positioning, making the training process more accessible and less complex.
2Reliability
If the robot is programmed to traverse the weld joint with corrected torch angles, then weld quality consistency is improved, but additional equipment (depth camera) and processing steps are required during the training phase
Solution Approach 1:
The depth camera serves as an intermediary device that bridges the gap between the robot's positional data and the weld joint's geometric requirements. It captures depth information that is processed to determine the joint geometry, which then informs the torch angle corrections. This intermediary enables automatic angle calculation without requiring complex mechanical measurement devices.
Solution Approach 2:
The system changes the parameter representation from direct manual angle input to vision-based geometric parameter extraction. By capturing depth images and processing them into 3D point clouds, the system automatically derives joint geometry parameters (planes, intersections) that are then used to calculate optimal torch angles, transforming the training process from manual angle setting to automated geometric analysis.
Data Source
AI summary
A method of correcting angles of a welding torch positioned by a user while training a robot of a robotic welding system is provided. Weldment depth data of a weldment and a corresponding weld seam is acquired and 3D point cloud data is generated. 3D plane and intersection data is generated from the 3D point cloud data, representing the weldment and weld seam. User-placed 3D torch position and orientation data for a recorded weld point along the weld seam is imported. A torch push angle and a torch work angle are calculated for the recorded weld point, with respect to the weldment and weld seam, based on the user-placed torch position and orientation data and the 3D plane and intersection data. The torch push angle and the torch work angle are corrected for the recorded weld point based on pre-stored ideal angles for the weld seam.