Automated Image Registration for Composite Assembly Validation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Automated inspection of precise component placement and fiber orientation in large composite parts is challenging due to the small size of features and three-dimensional work surfaces, leading to human error and high costs with existing imaging systems.
Innovation Solution
A method and apparatus using a programmable controller, a movable imaging system with a pan/tilt device and zoom lens, and a light projector to project reference features onto a work surface, allowing for accurate validation of piece location and orientation through coordinated targeting and image comparison with design data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If conventional imaging systems are used to inspect component placement, then the inspection process can be automated, but the system cannot resolve small features on large work surfaces
Solution Approach 1:
The patent introduces reference features (such as high-contrast markers or projected patterns) as intermediary elements between the imaging system and the small fiber features. These reference features serve as mediators that the camera can easily detect and use for registration, enabling the system to locate and focus on small features indirectly through the larger, more detectable reference markers.
Solution Approach 2:
The patent adds a temporal dimension to the inspection process by using sequential imaging - first capturing reference features for registration, then capturing fiber features. This multi-stage approach in time allows the system to handle both large-scale positioning and fine-detail inspection that would be impossible in a single static image.
2Adaptability or versatility
If the work surface is three-dimensional, then the assembly can accommodate complex geometries, but the camera cannot rapidly locate and focus on small elements
Solution Approach 1:
The patent performs preliminary registration by capturing reference features and establishing a coordinate transformation before inspecting the actual fiber features. This preliminary action creates a mapping between the camera view and the work surface coordinate system, enabling rapid localization of small elements without needing to physically move or refocus the camera for each feature.
Solution Approach 2:
The patent creates a digital copy or model of the work surface geometry through the reference feature registration process. This virtual model allows the system to rapidly calculate camera positions and focus settings for three-dimensional surfaces without physical adjustment, maintaining speed while handling complex geometries.
3Area of stationary object
If galvanometer mirrors are used to redirect camera view, then inspection coverage is improved, but the cost exceeds practical limits
Solution Approach 1:
The patent replaces the mechanical galvanometer mirror system with a computational approach using reference feature detection and coordinate transformation. Instead of physically redirecting the camera view through moving mirrors, the system uses software-based registration to achieve the same effect, eliminating expensive mechanical components while maintaining inspection coverage.
4Measurement precision
If manual inspection is used to verify piece location and fiber orientation, then accurate validation is achieved, but the process is slow and prone to human error
Solution Approach 1:
The patent enables the inspection system to self-register and self-validate by automatically detecting reference features, establishing coordinate transformations, and comparing fiber orientations against design data without human intervention. This automation maintains the accuracy previously requiring manual inspection while dramatically increasing speed and eliminating human error.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables rapid and highly accurate automated inspection of large composite components, reducing human error and overcoming limitations of previous systems by ensuring precise placement and orientation validation.
Implementation Method 1
A light projector is provided for projecting reference features to known locations on the work surface. The light projector is contemplated to be a laser projector or other light source capable of projecting reference features onto the mandrel, work surface, or layer of ply already applied to the work surface.
Implementation Method 2
The imaging device generates an image of the reference features projected by the light projector and of the piece applied onto the work surface.
Data Source
AI summary
A method and apparatus for verifying location and orientation of a piece attached to a work surface for s manufacturing a workpiece is disclosed. A controller is programmed with design data for manufacturing steps. Multiple layers of ply are sequentially layered over a work surface. An imaging system capable of redirecting a field of view along the work surface is provided. A light projector is provided for projecting reference features onto the work surface. The light projector projects a reference feature to a known location relative to the piece applied to the work surface within the field of view of the imaging device. The imaging device generates an image of the reference features projected by the light projector and of the piece applied onto the work surface. The controller compares the location of the piece on the work surface with the design data validating accurate placement of the piece.