Robot Teaching Method Using Relative Error Compensation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional robot systems require individual teaching work for each hand, which is inefficient due to mechanical and assembly errors, necessitating separate teaching processes for hands positioned around a common axis.
Innovation Solution
A robot system with a first and second hand connected to a robot arm, where teaching values for the second hand are generated based on relative errors detected between the hands, allowing efficient teaching by using one hand as a reference and adjusting the other hand's values accordingly.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If separate teaching work is performed for each hand, then teaching accuracy can be ensured, but teaching time and operational complexity increase significantly
Solution Approach 1:
The patent uses the teaching values obtained from teaching the first hand as a reference, and generates teaching values for the second hand by copying and adjusting the reference values based on measured relative errors. This eliminates the need to perform separate teaching work for each hand while maintaining teaching accuracy through error compensation.
Solution Approach 2:
The patent measures the relative error in rotation amount between the first hand and second hand, and uses this parameter to adjust the teaching values for the second hand. By changing the teaching values based on the measured error parameter, the system achieves accurate teaching without repeating the entire teaching process for each hand.
2Measurement precision
If separate teaching work is performed for each hand, then individual hand accuracy is improved, but device complexity and operational difficulty increase
Solution Approach 1:
The patent merges the teaching processes for multiple hands into a single reference-based teaching workflow. By teaching one hand as reference and generating teaching values for other hands through error compensation, the system reduces teaching process complexity while maintaining individual hand positioning accuracy.
Solution Approach 2:
The patent introduces relative error measurement as an intermediary step between teaching the reference hand and generating teaching values for other hands. This intermediary process simplifies the overall teaching complexity by providing a systematic method to transfer teaching values across multiple hands with minimal additional complexity.
3Stability of the object's composition
If reference unit teaching is performed for all hands, then positioning consistency is achieved, but productivity decreases due to repeated teaching requirements
Solution Approach 1:
The patent performs preliminary teaching of one hand as reference and stores its teaching values. These pre-obtained teaching values are then used as the basis for generating teaching values for other hands through error compensation, eliminating the need for repeated teaching and improving productivity while maintaining positioning consistency.
Solution Approach 2:
The patent measures the relative error between hands and uses this feedback information to adjust and generate teaching values for non-reference hands. This feedback mechanism ensures positioning consistency across all hands while avoiding the need for repeated teaching, thereby improving teaching efficiency and productivity.
Data Source
AI summary
A robot system includes a robot including a robot arm, and a first hand and a second hand which are connected to the robot arm and which are provided to independently rotate about an axis on the robot arm; and a controller configured to control an operation of the robot. When the robot arm and the first hand are operated so that the first hand reaches a predetermined target position, teaching values for the first hand in the target position is generated. When the first hand and the second hand are rotated based on the teaching values for the first hand, a relative error in rotation amount around the axis between the first hand and the second hand is acquired and stored in a memory. Teaching values for the second hand is generated from the teaching values for the first hand based on the acquired relative error.