Dual-Camera Robot Articulation for Faster Visual Servoing

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional visual servoing systems for robot apparatuses lack improved speed and accuracy, and are not economically viable.

Innovation Solution

A system and method utilizing two cameras coupled to a robot arm to minimize pixel-wise distance between a member and a target by alternating input from each camera, allowing the robot to articulate the member in multiple directions, with image processing for precise alignment, enabling robust convergence without camera calibration or point triangulation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional visual servoing systems are used, then the system is economically viable, but the speed and accuracy of servoing are insufficient

Engineering Contradiction:
Improveservoing accuracyVSAvoidservoing speed
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The visual servoing system is segmented into two independent camera channels, each responsible for measuring distance in a specific direction. This segmentation allows parallel processing of visual information from different perspectives, improving both measurement precision and processing speed without requiring complex calibration procedures.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system transitions from single-dimension visual feedback to multi-dimensional measurement by using two cameras oriented in different directions. Each camera provides distance measurements along its optical axis, creating a multi-dimensional measurement space that improves accuracy while maintaining computational efficiency through alternating input processing.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Measurement precision

If multiple cameras are used to improve measurement accuracy, then servoing precision improves, but system complexity increases

Engineering Contradiction:
Improvedistance measurement accuracyVSAvoidcamera system complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The control device alternates between receiving input from the first camera and the second camera in periodic cycles. This periodic action allows the system to process visual information from multiple cameras sequentially rather than simultaneously, reducing computational complexity while maintaining the measurement accuracy benefits of multi-camera input.

Inventive Principle:
Principle #19Periodic action

Solution Approach 2:

The control device acts as an intermediary that selectively processes input from either the first or second camera at any given time. This intermediary approach simplifies the control architecture by avoiding the need to process all camera inputs simultaneously, reducing system complexity while preserving measurement precision through alternating measurements.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Speed

If conventional single-camera visual servoing is used, then the system is simpler, but speed and accuracy of position control are limited

Engineering Contradiction:
Improverobot motion speedVSAvoidmotion control stability
Core Design Contradiction:
SpeedVSReliability

Solution Approach 1:

The system implements enhanced feedback by alternating between two camera inputs, each providing distance measurements in different directions. This multi-directional feedback mechanism improves motion control stability and reliability while enabling faster convergence to target positions, as the robot receives corrective information from multiple perspectives rather than a single viewpoint.

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP3871842A1System and method for member articulation
Publication Date: 2021.09.01 HAMILTON SUNDSTRAND CORP
  • EP3871842A1 patent drawingFigure 1
  • EP3871842A1 patent drawing

AI summary

A system (100) including a robot (102) mean to move a member (104) by using a first camera (106) coupled to the robot (102), a second camera (108) coupled to the robot (102), a control device (110) configured to control position of the robot (102) in order to minimize a pixel-wise distance between the member (104) and a target (112) based on alternating input from the first camera (106) and the second camera (108).