Mixed-Reality Welding Robot Teaching for Precise Torch Posture
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing robot teaching systems face challenges in accurately teaching the posture of a welding torch due to difficulties in aligning the worker's view with the workpiece and welding robot using head-mounted displays, leading to inefficiencies in teaching point posture adjustments.
Innovation Solution
A robot teaching system utilizing a mixed reality device that superimposes a virtual teaching tool on the real-world environment, allowing workers to perform posture adjustments through aerial operations, which are then recorded and applied to the actual welding robot.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If a head-mounted display is used to show the three-dimensional model and operation screen, then the worker can view the virtual teaching tool overlaid on the real workpiece, but it becomes difficult to align the worker's view with the workpiece and welding robot
Solution Approach 1:
The patent introduces a camera as an intermediary device to capture images of the workpiece and welding robot, and a processing device to process these images and determine spatial relationships. This intermediary system bridges the gap between the worker's head-mounted display view and the actual workpiece position, enabling accurate alignment without directly observing the real scene.
Solution Approach 2:
The patent creates a virtual copy of the workpiece and welding robot in the three-dimensional model displayed on the head-mounted display. By capturing images with a camera and generating corresponding three-dimensional models, the system replicates the real scene virtually, allowing the worker to interact with an accurate digital representation that maintains precise spatial relationships.
2Productivity
If traditional teaching methods are used for welding robot posture adjustment, then the process is straightforward, but the teaching time is excessive
Solution Approach 1:
The system captures the real workpiece and welding robot with a camera and creates a three-dimensional virtual model that replicates the actual scene. This virtual copy allows the worker to perform aerial operations in the virtual space, which are then automatically transferred to control the real welding robot, dramatically reducing the time required for teaching while maintaining accuracy.
Solution Approach 2:
The patent replaces traditional mechanical teaching methods (manual positioning and adjustment) with an aerial operation system. Instead of physically moving the welding robot to teach positions, the worker performs gestures in the air that are detected by the camera and processed to control the robot's movements, substituting mechanical manipulation with optical detection and digital processing.
3Ease of operation
If aerial operations are detected to control the three-dimensional model, then the worker can perform operations separated from the display device, but the detection complexity increases
Solution Approach 1:
The patent combines multiple functions into integrated components: the camera serves both to capture images of the workpiece and to detect aerial operations; the processing device simultaneously generates three-dimensional models, detects hand gestures, and determines spatial relationships. This merging reduces the number of separate devices needed while maintaining the capability for easy aerial operations.
Solution Approach 2:
The system uses the camera's image capture capability to serve dual purposes: creating the three-dimensional model for display and detecting the worker's aerial operations. The same optical input that builds the virtual environment also provides the data needed to interpret and respond to user gestures, eliminating the need for separate detection sensors.
Data Source
AI summary
A robot teaching system stores a three-dimensional model corresponding to at least a part of a robot or a welding torch, or a teaching member, outputs a display image for displaying the three-dimensional model and an operation screen on a display device configured to be mountable to a worker based on the three-dimensional model and operation screen data and displaying an image to be superimposed on an image of an actual environment or the actual environment itself, and generates a display image for displaying the post-change three-dimensional model after changing a posture of the three-dimensional model based on an aerial operation of the worker in the air separated from the display device.