AR Virtual Component Positioning for Markerless Assembly Guidance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional Augmented Reality (AR) systems face challenges in accurately positioning virtual objects within a user's field of view, particularly in complex and changing physical environments, such as during product manufacturing and assembly, due to the inefficiency of marker placement and limited processing capabilities of AR headsets.
Innovation Solution
The system generates and displays virtual representations of components based on three-dimensional surface data of partially assembled products, calculating their location, orientation, size, and shape without relying on markers, suitable for use with modest computing resources like AR headsets, allowing for efficient assembly guidance in AR environments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If markers are placed on physical objects for virtual object positioning, then the positioning accuracy is improved, but the manufacturing complexity and time increase due to the need to place markers on all parts including multi-part products
Solution Approach 1:
The patent extracts the positioning function from physical markers and implements it through virtual markers generated by the AR system. Instead of placing physical markers on each component, the system uses computational geometry and surface scanning to create virtual reference markers that are overlaid on the AR display, eliminating the need for physical marker placement during manufacturing and assembly.
Solution Approach 2:
The patent creates virtual copies of reference markers that are displayed through the AR system's transparent display. These virtual markers replicate the positioning function of physical markers without requiring physical placement. The system generates these virtual markers by processing three-dimensional surface data of the workpiece, allowing multiple markers to be displayed simultaneously without additional manufacturing steps.
2Ease of operation
If object recognition is used for environment tracking instead of markers, then the ease of operation is improved, but the processing requirements increase beyond the capabilities of AR headsets
Solution Approach 1:
The patent applies partial action by focusing the processing effort on specific geometric features of the workpiece surface rather than performing comprehensive object recognition. The system scans the surface and identifies key geometric elements (edges, corners, surfaces) to generate virtual markers, rather than attempting to recognize and process the entire object geometry. This reduces the computational load to levels suitable for AR headset processors.
Solution Approach 2:
The patent segments the processing task into distinct stages: surface scanning, geometric feature extraction, virtual marker generation, and AR display. By dividing the complex object recognition process into these smaller, more manageable segments, the system can perform positioning functions with reduced computational requirements at each stage, making it feasible for AR headset processors.
3Adaptability or versatility
If comprehensive object recognition is implemented for complex geometries, then the adaptability is improved, but the device complexity increases beyond AR headset capabilities
Solution Approach 1:
The patent applies partial action by focusing the processing effort on specific geometric features of the workpiece surface rather than performing comprehensive object recognition. The system scans the surface and identifies key geometric elements (edges, corners, surfaces) to generate virtual markers, rather than attempting to recognize and process the entire object geometry. This reduces the computational load to levels suitable for AR headset processors.
Solution Approach 2:
The patent replaces complex mechanical/m computational object recognition systems with a simplified approach using three-dimensional surface scanning and geometric feature extraction. Instead of implementing full object recognition algorithms that would require powerful processors, the system uses surface data to directly generate virtual markers based on identifiable geometric features, substituting a simpler computational approach for the more complex object recognition task.
Data Source
AI summary
Systems and methods include determination of a first component of a set of components under assembly in a physical environment, determination of a first physical position of a user with respect to the first component in the physical environment, determination of a second component of the set of components under assembly to be installed at least partially on the first component based on assembly information associated with the set of components, determination of three-dimensional surface data of the second component, determination of a physical relationship in which the second component is to be installed at least partially on the first component based on a model associated with the set of components, determination of a graphical representation of the second component based on the first physical position of the user with respect to the first component, the physical relationship, and the three-dimensional surface data of the second component, and presentation of the graphical representation to the user in a view including the first component in the physical environment, wherein the presented graphical representation appears to the user to be in the physical relationship with respect to the first component.