Autonomous Piece-Goods Picking With Defined 3D Scan Positioning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current automated order picking systems face challenges in achieving reliable and precise x-y positioning of sensors relative to source pallets due to inherent tolerances in global localization and translational and rotational errors, leading to inconsistent scan positions and reduced accuracy.
Innovation Solution
A method involving sensor systems, such as 3D cameras or laser scanners, for detecting and evaluating the relative positioning of source load carriers in the x-y plane and determining a defined scan position in a three-dimensional space, allowing for consistent and repeatable detection of package heights and precise alignment of the robot arm, decoupling sensor movement from the robot arm's vertical movement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If global localization and navigation based on a global map is used for positioning the autonomous industrial truck, then the system can operate autonomously without additional aids, but inherent tolerances cause deviation in horizontal x-y plane positioning relative to source pallets
Solution Approach 1:
The patent introduces an intermediary measurement step between global localization and package detection. A sensor system (e.g., laser scanner, camera) scans the source pallet to detect its actual position and orientation in the x-y plane. This intermediate measurement compensates for the tolerances of global localization, providing precise relative positioning information for subsequent package scanning and picking operations.
2Device complexity
If the sensor system is rigidly attached to the robot arm, then the structure is simplified, but vertical movement of the robot arm causes inconsistent scan positions relative to the source pallet
Solution Approach 1:
The patent merges the sensor system with the robot arm structure, attaching sensors (e.g., laser scanners, cameras) directly to the robot arm or its end effector. This integration allows the sensors to move with the robot arm while maintaining a defined geometric relationship. The system compensates for position changes through real-time calculation of the scan position based on robot arm pose data, ensuring consistent measurement accuracy despite the dynamic attachment.
3Adaptability or versatility
If the scan position is not precisely determined, then the system can operate more flexibly, but accuracy in detecting and localizing packages is reduced
Solution Approach 1:
The patent implements a feedback mechanism where the actual scan position is continuously determined and used to adjust subsequent scanning operations. The system calculates the precise position of the sensor system relative to the source pallet based on robot arm pose data and performs real-time corrections. This feedback loop maintains high measurement accuracy while allowing the system to adapt to different picking scenarios and configurations.
Data Source
Figure 1
Figure 2~3
Figure 4~5
AI summary
The invention relates to a method for picking individual items (P) that are stored on at least one source load carrier (Q) and individually placed on at least one target load carrier (Z), wherein the individual items (P) are detected by means of sensors (S) arranged on an autonomous industrial truck (1) and picked up from the source load carrier (Q) and placed on the target load carrier (Z) by means of a handling device (R) arranged on the industrial truck (1). It is proposed that, in a first step, load carrier detection is carried out by detecting the source load carrier (Q) by means of the sensors (S) and identifying it by evaluating the sensor data in a data processing unit (D), and determining its relative position to the industrial truck (1) in a horizontal xy-plane; in a second step, height detection is carried out.by scanning a vertical z-coordinate using the sensor (S) in a vertical scanning process, wherein the scanning process begins in a lower starting position, in which the source load carrier (Q) was detected in the first step, and ends in an upper end position, in which a top level E of the individual items (P) is detected by evaluating the sensor data in the data processing unit (D), and in a third step a defined scan position of the sensor (S) in a three-dimensional space is determined from the evaluations of the sensor data of the load carrier detection and the height detection. The invention further relates to a system for carrying out the method.