Robotic Wire Insertion Using Connector Segmentation Masks

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current methods for automating the insertion of wires into connectors face challenges in achieving high accuracy and speed, particularly due to sensitivity to ambient lighting conditions and the need for manual tuning of vision parameters, which limits adaptability to different connector types and environments.

Innovation Solution

A robotic system equipped with a camera system and a controller that generates images of connectors, creates segmentation masks, and applies thresholding to accurately insert wires into connectors, reducing the impact of ambient lighting and enabling faster, more adaptable wire insertion by using a final mask for multiple insertions without regenerating it each time.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional vision systems are used for wire insertion automation, then the system can operate with standard imaging, but the accuracy is compromised due to sensitivity to ambient lighting conditions

Engineering Contradiction:
Improveconnector detection accuracyVSAvoidambient lighting sensitivity
Core Design Contradiction:
Measurement precisionVSObject-affected harmful factors

Solution Approach 1:

The patent applies color space transformation by converting images from RGB color space to LAB color space, where the L channel represents lightness and the AB channels represent color information. This transformation separates luminance from chrominance, making the connector detection invariant to ambient lighting conditions while maintaining detection accuracy across varying light environments.

Inventive Principle:
Principle #32Color changes

2Measurement precision

If manual tuning of vision parameters is performed for each connector type, then detection accuracy can be optimized, but the device complexity and time consumption increase significantly

Engineering Contradiction:
Improvevision parameter accuracyVSAvoidparameter tuning complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent implements an automated parameter selection mechanism that uses unsupervised clustering (K-means algorithm) to automatically determine optimal vision parameters for different connector types. The system self-adjusts by clustering image data and selecting parameters based on cluster characteristics, eliminating the need for manual tuning while maintaining high detection accuracy across various connector types.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent dynamically changes imaging parameters based on detected connector characteristics. After initial connector detection, the system adjusts exposure time, gain, and other camera parameters according to the specific connector type and lighting conditions, enabling adaptive optimization without manual intervention for each connector variant.

Inventive Principle:
Principle #35Parameter changes

3Measurement precision

If vision parameters are optimized for specific connector types, then detection accuracy improves, but the adaptability to different connector types decreases

Engineering Contradiction:
Improvedetection accuracyVSAvoidconnector type adaptability
Core Design Contradiction:
Measurement precisionVSAdaptability or versatility

Solution Approach 1:

The patent creates a universal detection framework that can handle multiple connector types using the same core algorithm. By employing unsupervised clustering and automatic parameter selection, the system achieves both high accuracy for specific connector types and broad adaptability to new connector variants without requiring retraining or manual parameter adjustment, making the system multi-functional across diverse connector types.

Inventive Principle:
Principle #6Universality (Multi-functionality)

4Measurement precision

If image processing is performed for every wire insertion operation, then real-time accuracy is maintained, but the productivity decreases due to computational redundancy

Engineering Contradiction:
Improvereal-time detection accuracyVSAvoidwire insertion speed
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The patent performs comprehensive image processing and parameter optimization in advance before the wire insertion sequence begins. The system captures an initial image, performs connector detection, determines optimal parameters, and prepares the insertion path beforehand. During the actual wire insertion operations, the pre-computed parameters and paths are reused, eliminating redundant processing while maintaining real-time accuracy for each insertion.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP4471715A1Robotic wire insertion using connector segmentation
Publication Date: 2024.12.04 THE BOEING CO
  • EP4471715A1 patent drawingFigure 1
  • EP4471715A1 patent drawingFigure 2
  • EP4471715A1 patent drawingFigure 3

AI summary

A method for inserting a wire. An image of a connector is generated using a camera system connected to an end effector. The end effector is at a connector pose facing the connector. A region in the image encompassing the connector is selected. An initial mask is created using the image and the region. The initial mask comprises a background region and a connector region. A segmentation mask is created using the initial mask and the image, wherein the segmentation mask includes edges for the connector. Thresholding is applied to the segmentation mask to recover pixels for the connector and generate a final mask. The final mask is used to insert the wire into the connector.