Chaotic Pick-and-Place Item Detection With Upstream Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Robotic gripper systems face challenges in precisely identifying and handling products that are touching or overlapping on a conveyor belt, leading to slower picks, incorrect picks, or potential damage, and struggle with variability in product size, shape, and weight, requiring adaptable gripping mechanisms.
Innovation Solution
A computer-implemented method separates object detection and tracking logic, using machine learning constructs for efficient object discrimination and tracking, allowing the robotic arm to grasp objects with improved precision by determining their location, orientation, and adjusting the gripper's opening based on object width and occlusion, employing a multi-headed model for real-time coordination.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional object detection logic is used to identify products on a conveyor belt, then the system can detect objects, but it struggles with products that are touching or overlapping, leading to incorrect picks and slower operation
Solution Approach 1:
The patent segments the detection task into two distinct components: object detection logic that identifies what objects are present, and object tracking logic that monitors object positions over time. This segmentation allows each component to be optimized independently, with tracking providing continuous position updates that improve precision for touching or overlapping products without requiring complete re-detection, thereby maintaining high productivity.
Solution Approach 2:
The system performs preliminary object detection to identify products on the conveyor belt, then uses object tracking logic to continuously monitor and update their positions. This preliminary detection followed by continuous tracking allows the system to anticipate object positions and adjust gripper timing accordingly, improving pick accuracy for overlapping products while maintaining high-speed operation.
2Productivity
If the robotic system attempts to grasp products quickly to maintain high throughput, then productivity increases, but precision and accuracy of grasping decrease
Solution Approach 1:
The patent implements a feedback mechanism where object tracking logic continuously monitors product positions and provides real-time updates to the control system. This feedback loop allows the system to adjust gripper timing and positioning based on actual object locations, ensuring high grasping accuracy even at high speeds. The tracking data feeds back into the pick execution to compensate for conveyor variations and maintain precision.
Solution Approach 2:
The system uses dynamic object tracking that adapts to changing object positions on the conveyor belt. Rather than relying on static detection, the tracking logic continuously updates object locations and predicts future positions, allowing the robotic system to dynamically adjust its grasping timing and position. This dynamic approach maintains high accuracy while enabling fast operation.
3Reliability
If the system uses comprehensive object detection to identify all products, then detection completeness improves, but the system becomes slower and less efficient
Solution Approach 1:
The patent divides the detection process into initial object detection that identifies all products, followed by efficient object tracking that monitors their positions. This segmentation allows comprehensive detection to occur once, while subsequent tracking provides continuous position updates with minimal processing time, maintaining detection completeness while reducing overall time loss.
Solution Approach 2:
The system performs comprehensive object detection as a preliminary action to identify all products on the conveyor belt, then uses this information to initialize tracking for each detected object. This preliminary comprehensive detection ensures no objects are missed, while the subsequent tracking phase maintains reliability with much faster, continuous position updates.
4Reliability
If the gripper applies strong force to secure products, then product handling reliability improves, but products may be damaged
Solution Approach 1:
The system performs preliminary tracking to determine the exact position, orientation, and dimensions of each target product before the gripper engages. This advance information allows the control system to calculate optimal gripper force and positioning, securing products reliably while applying only the necessary force to avoid damage.
Solution Approach 2:
The object tracking logic provides real-time feedback on product position and characteristics, allowing the control system to adjust gripper force dynamically. This feedback ensures that sufficient force is applied to secure the product reliably while avoiding excessive force that could cause damage, adapting to each product's specific properties.
Data Source
AI summary
Exemplary embodiments relate to a machine-learning based approach to detecting individual items in a chaotic moving pick-and-place environment. In such an environment, objects may move relative to a robotic arm. As the objects move through the environment, their locations may change. A relatively more-processing-intensive procedure is employed once on an upstream side of the pick and place station in order to identify or initially segment objects in the environment. Identified items are then tracked using less intensive methods as the object moves through the environment. Detection is performed once on an upstream side of the pick and place station and then identified items are tracked using less intensive methods as the object moves through the pick-and-place station.