Robot Pose Compensation Using In-Machine Visual References
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional systems using automatic guided vehicles (AGVs) for robot operations at machine tools face challenges in accurately compensating the robot's pose due to tilting and varying center of gravity, leading to inaccurate positioning of the robot's hand, especially when attaching or removing workpieces, which can result in failed clamping and reduced system availability.
Innovation Solution
Compensate the robot's operating pose based on images including an identification figure and the machine tool's internal structure, using a camera to capture and adjust the robot's position accurately.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If a visual target consisting of two calibration markers is arranged on an outer surface of the machine tool for position compensation, then the position compensation can be performed with a simple setup, but the robot's hand cannot be accurately positioned when inserted inside the machine tool due to tilt variations
Solution Approach 1:
The patent introduces an identification figure as an intermediary element that is arranged inside the machine tool's machining area. This identification figure serves as a mediator between the camera and the machining area, enabling accurate position compensation by providing a reference that accounts for tilt variations when the robot arm is inserted into the machine tool.
Solution Approach 2:
The patent changes the arrangement dimension of the reference element from the outer surface (external dimension) to the internal machining area (internal dimension). By arranging the identification figure inside the machining area, the system captures position information from the actual working dimension where tilt effects occur, enabling accurate compensation in three-dimensional space.
2Adaptability or versatility
If the robot arm is overhanging to a great extent from the automatic guided vehicle to insert the hand inside the machine tool, then the robot can perform workpiece attachment and removal operations, but the tilt of the robot-mounted surface increases causing positioning errors
Solution Approach 1:
The patent implements a feedback mechanism where the camera captures images of the identification figure arranged inside the machining area, and the controller calculates the robot's actual position and orientation based on these images. This feedback loop compensates for tilt variations caused by arm overhanging, allowing the robot to maintain positioning accuracy despite the mechanical constraints of the mobile platform.
3Manufacturing precision
If the chuck has only a very small movement allowance (stroke) with very small clearance between the workpiece and the chuck, then the clamping precision is high, but the robot hand cannot be reliably positioned due to accumulated positioning errors
Solution Approach 1:
The patent replaces the conventional mechanical position compensation method (using two calibration markers on the outer surface) with an optical measurement system. By using a camera to capture images of an identification figure and calculating position based on image analysis, the system achieves higher positioning accuracy that is sufficient for chucks with very small movement allowances, thereby ensuring reliable workpiece attachment.
Data Source
AI summary
A system includes a machine tool 10, a robot 25 having a camera 31, and a transportation device 35 having the robot 25 mounted thereon, and an identification figure is arranged in a machining area of the machine tool 10.