Robot Operation Control Using Segmented Learning Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing learning devices for robots require re-learning when new operations are added, leading to increased learning data and time, which is inefficient.
Innovation Solution
The use of multiple learning devices, each trained on specific processing operations, allows for efficient learning by generating a new device for each additional operation, reducing the burden on existing learning devices and calculating command values based on evaluation outputs from multiple learning devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a single learning device is used to control all robot operations, then the device can handle multiple processing operations, but the learning data becomes enormous and re-learning takes excessive time when new operations are added
Solution Approach 1:
The patent divides the single learning device into multiple specialized learning devices, where each learning device is responsible for a specific processing operation (e.g., first learning device for first processing operation, second learning device for second processing operation). This segmentation allows each device to learn only its designated operation independently, so adding a new operation requires creating only a new learning device for that operation rather than re-learning all operations in a single device.
2Adaptability or versatility
If a single learning device is used to control all robot operations, then the device can handle multiple processing operations, but the learning data becomes enormous
Solution Approach 1:
The patent segments the learning data into separate datasets for each processing operation. Each learning device receives and processes only the learning data relevant to its specific operation, rather than a single device handling all learning data for all operations. This reduces the data volume each device must manage and process.
3Productivity
If multiple learning devices are created for different processing operations, then the learning burden is reduced, but the device complexity increases
Solution Approach 1:
The patent implements a control device with universal functionality that can selectively activate appropriate learning devices based on the required processing operation. The control device serves multiple functions: it manages multiple specialized learning devices, selects which device to use based on the operation type, and integrates their outputs. This universality at the control level manages the complexity of having multiple learning devices while maintaining system coherence.
Data Source
AI summary
An operation control device for a robot comprises: an input part inputting at least one operation candidate, and a captured image including an object to be processed; a first learning device that has finished learning performed according to first learning data to output a first evaluation value indicating evaluation of each operation candidate when the robot performs a first processing operation upon input of the captured image and the operation candidate; a second learning device that has finished learning performed according to second learning data which differs from the first learning data, to output a second evaluation value indicating evaluation of each operation candidate when the robot performs a second processing operation upon input of the captured image and the operation candidate; and an evaluation part that, based on at least one of the first evaluation value and the second evaluation value, calculates a command value.