Deformation Modeling for Deformable Workpieces Using Edge-to-Edge Constraints
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current robotic systems face challenges in accurately manipulating deformable workpieces due to their flexible nature, which leads to unreliable pose determination and increased complexity in fixture design, especially for workpieces with highly planar shapes.
Innovation Solution
The implementation of enhanced deformation models using edge-to-edge constraints between the nominal representation of a workpiece and detected edges in 3D point cloud data, which provides a more stable and accurate deformation modeling technique, allowing for precise robotic manipulation without the need for custom fixtures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If point cloud data matching is used to determine workpiece pose, then the system can operate without fixtures, but the pose determination becomes unreliable for deformable workpieces with highly planar shapes
Solution Approach 1:
The workpiece surface is segmented into multiple planar regions, each represented by a plane equation with parameters (a, b, c, d). This segmentation allows the system to distinguish between different planar areas and track their deformations independently, preventing the tablecloth effect where the entire surface slides without constraint.
Solution Approach 2:
The system transitions from point-based matching (2D/3D point cloud) to plane-based matching (2D plane equations). By representing surfaces as planes with explicit equations and matching these planes to canonical plane representations, the system gains additional dimensional constraints that prevent ambiguous matching on highly planar surfaces.
2Manufacturing precision
If custom fixtures are designed for each workpiece, then manipulation precision is improved, but the time and complexity of building fixtures increases
Solution Approach 1:
Instead of building physical fixtures for each workpiece, the system creates a digital copy (canonical representation) of the workpiece geometry. This canonical model, stored as a set of planes with known spatial relationships, serves as a reference that can be rapidly matched to sensor data without requiring physical fixture construction or customization.
Solution Approach 2:
The system replaces the mechanical fixture system with a computational plane-matching system. Rather than using physical clamps, vises, or fixtures to constrain workpieces, the system uses algorithmic plane fitting and deformation modeling to achieve precise manipulation, eliminating the time-consuming process of mechanical fixture design and fabrication.
3Adaptability or versatility
If point-to-plane constraints are used for deformation modeling, then the system can handle deformable workpieces, but the optimization becomes slow and less stable
Solution Approach 1:
The deformation model is segmented into independent plane components, each with its own parameters (a, b, c, d) and deformation characteristics. This segmentation allows the optimization to proceed independently for each plane region, reducing the complexity of the overall optimization problem and improving convergence speed and stability.
Solution Approach 2:
The system changes the optimization parameters from point-based coordinates to plane equation parameters (a, b, c, d). By optimizing these plane parameters directly rather than adjusting individual point positions, the system achieves faster and more stable convergence because the plane parameters provide more direct constraints on the deformation geometry.
Data Source
AI summary
Methods, systems, and apparatus, including computer programs encoded on computer storage media, for performing deformation modeling with edge-to-edge constraints. One of the methods includes computing edge-to-edge constraints that match one or more edges in point cloud data to one or more edges in a nominal representation of a workpiece. A current estimate of the deformation data is repeatedly updated according to the computed edge-to-edge constraints. A robotic manipulation task is then performed according to the generated deformation data.