Robot Motion Synchronization With Posture Correction Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing entertainment robots struggle to effectively synchronize their movements with those of a user, limiting their ability to provide immersive and interactive experiences.
Innovation Solution
A robot control system that includes a posture estimating section to acquire and analyze image data of a user, and a motion control section to synchronize the robot's posture and movement with that of the user, using a synchronization control section and a correction processing section to refine the motion.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a robot device simply copies user movement data, then the synchronization is computationally simple, but the movement appears unnatural and lacks adaptability to the robot's physical constraints
Solution Approach 1:
The patent introduces an intermediary model (virtual human or reference motion database) between the user's raw movement data and the robot's execution. This intermediary layer transforms natural human movement into robot-appropriate motion patterns, resolving the contradiction by adapting movements to physical constraints while maintaining natural appearance through learned motion patterns.
Solution Approach 2:
The system performs preliminary action by pre-learning and storing appropriate motion patterns in a database before actual synchronization occurs. Motion patterns are pre-adapted to the robot's physical constraints through machine learning, so that during real-time synchronization, the robot can directly retrieve and execute pre-processed motion data without complex real-time calculations.
2Measurement precision
If the robot uses complex machine learning models to generate natural movements, then the movement quality improves, but the processing time and computational resources increase
Solution Approach 1:
The system performs complex machine learning processing in advance during the training phase, where the neural network learns to map between different motion representations. The trained model and learned patterns are stored for rapid retrieval during real-time operation, achieving high precision posture estimation without real-time computational delays.
Solution Approach 2:
The patent uses copying by creating virtual human models and reference motion databases that replicate natural movement patterns. Instead of performing complex real-time calculations, the system copies pre-learned motion patterns from the virtual model or database, achieving accurate and natural-looking movements with minimal processing time during actual synchronization.
3Adaptability or versatility
If the robot synchronizes all body parts with the user, then the synchronization completeness is high, but the system becomes less adaptable to situations where full synchronization is inappropriate
Solution Approach 1:
The patent implements dynamic adaptability by allowing the synchronization degree and scope to be adjusted based on the situation. The system can dynamically select which body parts to synchronize and to what extent, switching between full-body synchronization, partial synchronization, or no synchronization as needed, thereby maintaining both completeness when appropriate and adaptability when conditions change.
Data Source
AI summary
A posture estimating section acquires image data in which images of a person are recorded, to estimate a posture of the person. A motion control section controls the motion of a robot device on the basis of an estimation result of the posture estimating section. A synchronization control section synchronizes a posture of the robot device with the posture of the person estimated by the posture estimating section. A correction processing section corrects the synchronized motion of the robot device made by the synchronization control section.