S-Bar Finger Joint Coupling for Smaller Robot Learning State Spaces
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The challenge of training robots using reinforcement learning is exacerbated by the large state space resulting from the increased degrees of freedom in robotic systems, making it difficult to effectively learn tasks in the physical world.
Innovation Solution
Implementing an S-bar mechanism that mechanically couples two joints of a robotic hand, reducing the number of degrees of freedom and simplifying mechanical and control complexity, allowing for more efficient machine learning training.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If reinforcement learning is used to train robots, then the robot can learn to complete tasks through trial-and-error, but the large state space increases with degrees of freedom makes training difficult
Solution Approach 1:
The patent merges the control of multiple finger joints by mechanically coupling them through S-bar linkages. The proximal and distal interphalangeal joints are connected such that they move together as a unified unit, reducing the number of independent degrees of freedom from six to three while preserving the ability to perform grasping tasks
Solution Approach 2:
The S-bar mechanism serves multiple functions: it mechanically couples joints to reduce state space, provides the necessary mechanical advantage for force transmission, and enables the robot to perform complex grasping motions with reduced actuation complexity
2Adaptability or versatility
If the number of degrees of freedom is increased to improve dexterity, then the robot can grasp objects more effectively, but the mechanical and control complexity increases
Solution Approach 1:
The patent combines multiple degrees of freedom into fewer independent motions through the S-bar linkage system. Three independent actuators control six finger joints by coupling proximal and distal joints, achieving the same grasping capability with half the actuation complexity
Solution Approach 2:
The finger is segmented into proximal and distal segments that are mechanically linked. The S-bar divides the motion control into two stages: proximal joint actuation drives distal joint motion through the linkage, creating a hierarchical control structure that simplifies the overall system
Data Source
AI summary
Provided are mechanisms that mechanically couple finger joint motion in robots.