Vision-Guided Workpiece Transfer for Flexible Pick-and-Place

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing automated workpiece transfer systems are costly and time-consuming to deploy due to customization requirements and have limited processing rates, as they often require significant manual reconfiguration to handle different workpieces and orientations.

Innovation Solution

An automated system utilizing an imaging device and a machine-learning model to identify pickable workpieces, defining operating parameters for an autonomous pick-and-place robot to retrieve and transfer workpieces based on position and orientation, allowing for efficient handling of varied workpieces without extensive reconfiguration.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Extent of automation

If traditional automated workpiece transfer systems are used, then workpiece transfer can be automated, but the system is costly and time-consuming to deploy due to customization requirements

Engineering Contradiction:
Improveworkpiece transfer automationVSAvoidsystem customization complexity
Core Design Contradiction:
Extent of automationVSDevice complexity

Solution Approach 1:

The system employs a universal robotic arm with interchangeable end effectors that can handle multiple workpiece types and orientations without requiring custom-built transfer mechanisms for each application, thereby reducing deployment costs and time while maintaining full automation capability

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

Solution Approach 2:

The patent replaces traditional mechanical transfer systems with vision-guided robotic manipulation, where machine learning models process imaging data to determine workpiece positions and orientations, and robotic arms execute transfer operations based on software control rather than hardwired mechanical linkages

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Productivity

If traditional feeders are used to transfer workpieces, then workpieces can be sorted and fed to assembly lines, but the processing rate is limited due to manual reconfiguration requirements

Engineering Contradiction:
Improveworkpiece processing rateVSAvoidmanual reconfiguration effort
Core Design Contradiction:
ProductivityVSEase of operation

Solution Approach 1:

The system enables robots to autonomously adapt to different workpieces by capturing images, processing them through machine learning models to identify positions and orientations, and automatically adjusting transfer parameters without requiring manual reconfiguration, thereby maintaining high processing rates while eliminating manual intervention

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent dynamically adjusts operational parameters such as robotic arm trajectories, end effector positions, and transfer timing based on real-time image analysis and machine learning predictions, allowing the system to optimize processing rates for each workpiece configuration without physical reconfiguration

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP4344834A1Automated workpiece transfer systems and methods of implementing thereof
Publication Date: 2024.04.03 ATS CORPORATION
  • EP4344834A1 patent drawingFigure 1
  • EP4344834A1 patent drawingFigure 2
  • EP4344834A1 patent drawingFigure 3

AI summary

Automated workpiece transfer systems and methods of implementing thereof are disclosed. The system can include an imaging device operable to capture an initial image of workpieces loaded onto a loading area; an autonomous pick-and-place robot, and a processor in communication with the imaging device and the pick-and-place robot. The autonomous pick-and-place robot can include an end-of-arm-tooling component operable to retrieve pickable workpieces from the loading area and transfer the pickable workpieces to a receiving area according to a set of operating parameters. The processor can be operable to apply a machine-learning model to the initial image to identify pickable workpieces; identify a region of interest within the initial image; and based on the initial image, define the set of operating parameters for operating the end-of-arm-tooling component to retrieve the one or more pickable workpieces.