Robot Grinding Control Using Image-Learned Weld Adaptation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Automation of grinding work by robots is challenging due to the non-uniform state of welded parts, such as unevenness, which varies for each part and requires manual adjustment.
Innovation Solution
A control device for robots that includes an autonomous command generation unit, a manual command generation unit, an operation control unit, a storage unit, and a learning unit. The learning unit performs machine learning using image data and operation data to generate operation correspondence commands, enabling the autonomous command generation unit to produce improved grinding accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If manual operation is used to adjust grinding parameters for each welded part, then grinding accuracy can be maintained, but productivity decreases due to repeated manual adjustments
Solution Approach 1:
The system enables the robot to autonomously adjust grinding parameters by capturing images of welded parts, analyzing their states through image processing, and automatically modifying grinding commands without manual intervention. This self-service capability maintains grinding accuracy while eliminating repeated manual adjustments, thereby improving productivity
Solution Approach 2:
The system implements a feedback mechanism where the robot captures images of welded parts during grinding, analyzes the actual state (such as unevenness), and uses this information to adjust subsequent grinding operations. This closed-loop feedback ensures grinding accuracy is maintained while automating the adjustment process to improve productivity
2Productivity
If autonomous operation is implemented without machine learning, then productivity increases, but manufacturing precision decreases due to inability to adapt to varying welded part states
Solution Approach 1:
The system uses machine learning to automatically adjust grinding parameters (such as grinding depth, speed, and path) based on the captured image data of welded parts. By dynamically changing these parameters according to the actual part state, the system maintains grinding accuracy while operating autonomously, thus improving productivity without sacrificing precision
Solution Approach 2:
The system applies different grinding parameters to different regions of the welded part based on local characteristics detected through image analysis. The machine learning model identifies specific features (such as unevenness locations and magnitudes) and adjusts grinding parameters locally, enabling autonomous operation to achieve both high productivity and maintained grinding accuracy
3Manufacturing precision
If machine learning is used to adapt to different welded part states, then manufacturing precision is maintained, but device complexity increases
Solution Approach 1:
The system replaces complex mechanical adjustment mechanisms with an information-processing approach using machine learning. Instead of using complex mechanical systems to manually adjust grinding parameters for each part variation, the system uses image capture, processing, and AI-based decision-making to automatically determine appropriate parameters, maintaining grinding accuracy while avoiding mechanical complexity
Solution Approach 2:
The system introduces an intermediary information-processing layer (image processing and machine learning algorithms) between the welded part and the grinding operation. This intermediary analyzes the part characteristics and translates them into appropriate grinding commands, maintaining precision while keeping the physical system relatively simple and modular
Data Source
AI summary
A control device includes: first circuitry that generates a command to cause a robot to autonomously grind a grinding target portion; second circuitry that generates a command to cause the robot to grind a grinding target portion according to manipulation information from an operation device; third circuitry that controls operation of the robot according to the command; storage that stores image data of a grinding target portion and operation data of the robot corresponding to the command; and forth circuitry that performs machine learning by using image data of a grinding target portion and the operation data for the grinding target portion, receives the image data as input data, and outputs an operation correspondence command corresponding to the operation data as output data. The first circuitry generates the command, based on the operation correspondence command.