Robotic Picking Analytics for Overlap-Aware Gripper Control
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 method for logging and analyzing robotic picking data, using machine learning constructs for object discrimination, grasp quality detection, and real-time tracking, with adaptable grippers like soft robotic members that can re-image the pile as objects move, and adjust rules and filters based on analytics.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional robotic gripper systems are used to pick products from a moving conveyor belt, then the system can operate automatically, but it struggles to precisely identify and handle products that are touching or overlapping, leading to slower picks and incorrect picks
Solution Approach 1:
The system performs preliminary imaging and object discrimination before the pick action occurs. The vision system captures images of products on the conveyor belt and identifies target objects in advance, allowing the controller to plan the pick trajectory and gripper positioning before the actual pick, thereby improving both accuracy and speed
Solution Approach 2:
The system uses real-time feedback from vision systems and sensors to monitor product positions and adjust pick parameters dynamically. The controller receives feedback about product location, orientation, and overlap conditions, and adjusts the pick trajectory and gripper positioning accordingly to maintain high accuracy even when products are touching or overlapping
2Productivity
If the robotic system attempts to handle products quickly to increase throughput, then productivity improves, but the likelihood of incorrect picks and product damage increases
Solution Approach 1:
The system performs preliminary imaging and trajectory planning before the pick action. The vision system captures images and the controller calculates the optimal pick trajectory and gripper positioning in advance, allowing high-speed operation without sacrificing accuracy because all critical decisions are made before the rapid pick execution
Solution Approach 2:
The system dynamically adjusts pick parameters based on real-time conditions. The controller modifies trajectory, speed, and gripper positioning dynamically during the pick cycle based on feedback from vision systems and sensors, enabling high productivity while maintaining reliability by adapting to each specific pick scenario
3Adaptability or versatility
If the robotic gripper uses fixed gripping parameters to handle products, then the control system is simple, but it cannot adapt to variability in product size, shape, and weight
Solution Approach 1:
The system uses dynamic, adjustable gripping parameters instead of fixed values. The controller modifies gripper force, positioning, and trajectory dynamically based on real-time feedback about product size, shape, and weight detected by vision systems and sensors, enabling high adaptability to product variability
Solution Approach 2:
The system performs self-adjustment of gripping parameters based on sensor feedback. The vision system and sensors automatically detect product characteristics and the controller autonomously adjusts gripper parameters without human intervention, providing adaptability while keeping the control system manageable through automated decision-making
4Measurement precision
If the vision system is finely tuned to cope with various product characteristics and environmental conditions, then measurement precision improves, but the system becomes more complex and requires extensive calibration
Solution Approach 1:
The vision system is designed with multi-functional capabilities to handle various product characteristics and environmental conditions through a unified platform. The system uses general-purpose image processing algorithms and sensors that can adapt to different products and lighting conditions, improving measurement precision without proportionally increasing complexity through standardized, versatile components
Data Source
AI summary
Exemplary embodiments pertain to an intelligence module for a robotic pick-and-place system. As the intelligence module identifies the next pick for the robotic gripper, data is logged and presented to a user. The user can evaluate why the system decided to select a particular target object, and adjust parameters of the system (such as filtering and sorting rules, gripper opening amounts, etc.) to achieve desired outcomes and/or load balancing. The system may optionally simulate proposed changes to estimate how the changes would affect system throughput and display projected and historical throughput on an analytics interface so that the user can evaluate the effects of their proposed changes.