Robot Behavior Estimation With Bilateral Feedback for Variable Conditions

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing robotic systems for learning object operation skills face challenges in adapting to varying conditions due to the lack of environmental detection mechanisms, leading to difficulties in performing tasks when environmental conditions change.

Innovation Solution

A behavior estimation apparatus and method that includes a bilateral system with a master and slave robot, capable of bidirectional control, equipped with image, acoustic, and haptic sensors, which estimates position and force command values to enable adaptive operation by recognizing reaction information from the environment, allowing the slave robot to learn and reproduce complex tasks robustly across different conditions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If the bilateral system simply stores data and directly reproduces the data without environmental detection, then the system complexity is reduced, but the robot cannot perform object operations when environmental conditions vary

Engineering Contradiction:
Improvesystem complexityVSAvoidadaptability to varying conditions
Core Design Contradiction:
Device complexityVSAdaptability or versatility

Solution Approach 1:

The patent introduces environmental detection mechanisms that provide feedback about current conditions to the behavior estimation device. The detection unit monitors environmental parameters, and this information is fed back to adjust the reproduction of stored data, enabling the system to adapt to varying conditions while maintaining manageable complexity through selective environmental monitoring.

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The system changes operational parameters based on detected environmental conditions. When environmental parameters vary beyond predetermined thresholds, the system modifies how stored data is reproduced, allowing adaptation to different conditions without requiring complete system redesign for each scenario.

Inventive Principle:
Principle #35Parameter changes

2Extent of automation

If reinforcement learning is used to learn object operation skills, then the robot can determine behaviors through observation and reward mechanisms, but a great number of attempts are required which increases time consumption

Engineering Contradiction:
Improveautonomous behavior determinationVSAvoidtime required for attempts
Core Design Contradiction:
Extent of automationVSLoss of time

Solution Approach 1:

The patent applies preliminary action by pre-storing operation data collected during teaching phases before actual autonomous operation. The behavior estimation device uses this pre-collected data to determine appropriate behaviors without requiring the robot to learn through numerous trial attempts during execution, significantly reducing time consumption while maintaining autonomous operation capability.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system creates copies of successful human operations through data collection and storage during teaching phases. These copied operation patterns are then reproduced by the robot during autonomous operation, eliminating the need for repeated trial-and-error learning and reducing time requirements while preserving autonomous behavior determination.

Inventive Principle:
Principle #26Copying

3Loss of time

If imitation learning is used to collect data from operator operations, then the number of attempts can be reduced, but bidirectional property between operator and robot is not considered leading to insufficient object operation skills

Engineering Contradiction:
Improvenumber of attemptsVSAvoidsuccess rate of object operations
Core Design Contradiction:
Loss of timeVSReliability

Solution Approach 1:

The patent introduces bidirectional feedback mechanisms where the robot's actions and the operator's reactions are both captured during data collection. This feedback loop allows the system to learn not just the operator's commands but also the dynamic interaction between operator and robot, improving the success rate of object operations while maintaining reduced training time.

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The system merges the collection of operator commands with the detection of operator reactions into a unified data collection process. By combining these previously separate data streams into a single integrated teaching phase, the system captures bidirectional interaction information efficiently, improving operation success rates without increasing the number of attempts required.

Inventive Principle:
Principle #5Merging (Combining)

4Device complexity

If the bilateral system assumes no variation in conditions between data storage and reproduction, then the control mechanism is simplified, but it is difficult to perform object operations when conditions vary

Engineering Contradiction:
Improvecontrol mechanism complexityVSAvoidperformance under varying conditions
Core Design Contradiction:
Device complexityVSAdaptability or versatility

Solution Approach 1:

The patent implements parameter changes by establishing predetermined thresholds for environmental parameters. When detected parameters exceed these thresholds, the system automatically adjusts its operation mode to account for condition variations. This approach maintains relatively simple control mechanisms while enabling the system to adapt to varying conditions through threshold-based parameter monitoring and adjustment.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP3782768B1Behavior estimation device, behavior estimation method, and behavior estimation program
Publication Date: 2024.10.09 UNIV OF TSUKUBA
  • EP3782768B1 patent drawingFigure 1
  • EP3782768B1 patent drawingFigure 2
  • EP3782768B1 patent drawingFigure 3

AI summary

[Problem] Learning of object operation skills robust against variation of conditions is implemented. [Solution] A behavior estimation apparatus 100 includes a collection unit 200 configured to collect skill data obtained when a slave robot is operated under a plurality of different conditions by using a bilateral system capable of operating the slave robot via a master robot through bidirectional control between the master robot and the slave robot. The behavior estimation apparatus 100 further includes a behavior estimation device 300 configured to estimate a command value for causing the slave robot 520 to automatically behave, based on the skill data collected by the collection unit 200 and a response output from the slave robot 520.