Robotic Bin Picking With Object Scoring for Secure Grasping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Robotic arms often fail to pick up objects from bins securely, leading to objects falling or breaking during transportation, especially when attempting to grasp objects by floating labels or partially attached parts.
Innovation Solution
A method employing a robotic arm with a controller, kinematic chain, scanning module, and picking-up module that captures images of the bin, recognizes object types, assigns scores based on type and position, and selectively uses a vacuum sucking disk or gripper to pick up objects with the highest score above a predetermined value, ensuring secure grasping.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the robotic arm picks up objects using a conventional picking-up module, then the picking-up process is simple, but the object may not be held firmly and may fall or break during transportation
Solution Approach 1:
The picking-up module is designed with multi-functionality, capable of switching between vacuum suction mode and gripper mode based on object characteristics. This allows a single module to handle diverse objects (bottles, boxes, bags) effectively, improving reliability without proportionally increasing system complexity
Solution Approach 2:
The picking-up module employs dynamic switching between different picking-up methods (vacuum suction vs. mechanical gripping) based on real-time object recognition results. The controller dynamically selects the appropriate mode, enabling adaptive response to different object types and ensuring secure holding
2Adaptability or versatility
If the robotic arm picks up objects without object recognition, then the operation is fast, but the picking-up method cannot be adapted to different object types
Solution Approach 1:
The system performs preliminary object recognition and analysis before the actual picking-up action. The scanning module captures images and the controller analyzes object characteristics in advance, determining the optimal picking-up method beforehand. This preliminary action enables adaptive picking-up while minimizing time loss during the critical picking-up moment
3Reliability
If the robotic arm uses a single picking-up method, then the device structure is simple, but it cannot securely hold objects with different characteristics
Solution Approach 1:
The picking-up module integrates multiple picking-up methods (vacuum suction disk and mechanical gripper) into a single universal module. The vacuum suction disk handles smooth-surfaced objects like bottles, while the mechanical gripper handles irregular or porous objects like boxes and bags. This multi-functional design ensures secure transport across different object types
Solution Approach 2:
The controller acts as an intermediary that receives object recognition data, analyzes characteristics, and selects the appropriate picking-up method. This intermediary function coordinates between the scanning module and picking-up module, ensuring the right method is applied to each object type without requiring complex mechanical switches
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The method enables the robotic arm to automatically pick up objects that can be firmly held, preventing them from falling during transportation by accurately identifying and scoring objects for secure grasping based on type and position, thus enhancing the reliability of object handling.
Implementation Method 1
a capturing step of capturing, by the scanning module, an image of an interior of the bin
Implementation Method 2
the picking-up module to pick up the one of the at least one TBT object
Data Source
AI summary
There is provided a method for picking up an object from a bin to be implemented by a robotic arm including a controller and a picking-up module. The method includes: recognizing, by the controller, at least one object in the bin based on an image of an interior of the bin so as to determine a type of each object; determining, by the controller, a score for each object based on the type thereof; determining, by the controller, whether a greatest score among the score(s) of the at least one TBT object is greater than a predetermined value; and by the picking-up module, picking up one of the at least one TBT object that has the greatest score when it is determined that the greatest score is greater than the predetermined value.