Robot Joint Teaching Calibration in Sealed Substrate Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In electronic device manufacturing systems, robots face challenges in achieving precise and rapid transport of substrates due to joint errors during joint coordinate teaching, leading to misaligned substrate handoffs and potential damage to substrates and system components, with conventional manual processes disrupting the sealed environment and requiring lengthy requalification processes.
Innovation Solution
A method and system that position a robot in various postures relative to a fixed location within a substrate processing system, generate sensor data to determine error values, and perform corrective actions automatically, without opening the sealed environment, using machine learning models to improve joint coordinate teaching accuracy and reduce errors in robot alignment and calibration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual joint coordinate teaching is performed to calibrate robot joints, then robot positioning accuracy can be improved, but the sealed environment must be opened causing contamination risk and lengthy requalification processes
Solution Approach 1:
The patent replaces manual mechanical teaching operations with an automated optical measurement system. Sensors mounted on the robot end effector automatically capture images of calibration targets at multiple joint positions, eliminating the need for manual intervention that would require opening the sealed environment. The system processes images computationally to determine joint coordinates, substituting mechanical teaching with automated sensing and computation.
Solution Approach 2:
The robot performs self-calibration by automatically positioning itself at multiple joint angles and capturing images of calibration targets. The system autonomously collects sensor data, processes images to extract coordinate information, and computes joint calibration parameters without external intervention. This self-service capability maintains the sealed environment while achieving accurate joint coordinate teaching.
2Measurement precision
If multiple joint positions are measured to improve calibration accuracy, then teaching precision is enhanced, but measurement time and system downtime increase
Solution Approach 1:
The patent implements continuous automated measurement across multiple joint positions without interruption to the sealed environment or production workflow. The robot systematically moves through calibration positions and captures images continuously, maximizing measurement efficiency. The automated image processing occurs in parallel, eliminating idle time between measurements and minimizing total calibration time.
3Ease of manufacture
If conventional manual teaching methods are used, then implementation simplicity is maintained, but productivity and line yield are reduced due to disruptions
Solution Approach 1:
The patent replaces complex manual teaching procedures with an automated optical measurement and computation system. The system requires minimal setup with calibration targets and sensors, then autonomously performs all measurement and calculation operations. This substitution maintains ease of implementation while eliminating production disruptions, thereby improving line yield and overall productivity.
Data Source
AI summary
A method includes positioning a robot in a plurality of postures in a substrate processing system relative to a fixed location in the substrate processing system and generating sensor data identifying a fixed location relative to the robot in the plurality of postures. The method further includes determining, based on the sensor data, a plurality of error values corresponding to one or more components of the substrate processing system and causing, based on the plurality of error values, performance of one or more corrective actions associated with the one or more components of the substrate processing system.