Human-Inspired Robot Gripper Control for Grasping Precision
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current robot systems face challenges in effectively mimicking human-like grasping of objects using multi-fingered grippers, as they struggle to align and orient their fingers to minimize object penetration and maintain a grasp similar to that of a human hand, leading to inefficiencies in grasping and manipulation tasks.
Innovation Solution
A robot system that includes a hand module to determine human hand vectors and positions, a gripper module to align and position its own gripper vectors, and an actuation module to minimize object penetration loss by adjusting finger positions and orientations based on image and sensor data, using a total loss function that considers object penetration, contact heatmap, orientation, and gripper self-penetration losses.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If robot grippers use conventional grasping methods, then the grasping process is simple, but the grasping precision and object damage risk increase
Solution Approach 1:
The patent copies the human hand grasping posture and finger configuration to control the robot gripper. By mapping human hand joint angles and finger positions directly to gripper actuation, the system achieves human-like grasping precision without requiring complex custom control algorithms, thus resolving the contradiction between grasping precision and control complexity
Solution Approach 2:
The system uses depth information from the RGB-D camera to calculate 3D contact points and verify gripper-object interaction. This feedback mechanism allows real-time adjustment of grasping force and finger positioning, improving grasping precision while maintaining manageable control complexity through automated adjustment
2Reliability
If robot grippers apply high grasping force, then the grasp stability improves, but the object penetration loss increases
Solution Approach 1:
The patent applies different grasping forces to different fingers based on their specific contact points and object geometry. Each finger adjusts its force locally according to the calculated optimal contact region, achieving overall grasp stability while minimizing localized penetration damage to the object
Solution Approach 2:
The system dynamically adjusts grasping force parameters based on object properties detected by the depth camera. By changing force magnitude and distribution parameters according to object characteristics, the system maintains reliable grasps while minimizing penetration loss
3Productivity
If robot grippers use human-like grasping postures, then the manipulation effectiveness improves, but the system complexity increases
Solution Approach 1:
The patent creates a universal control framework that maps human hand postures to various gripper configurations. This multi-functional approach allows the same control system to handle different objects and grasping scenarios by simply changing the input human hand demonstration, improving manipulation effectiveness without proportionally increasing system complexity
Solution Approach 2:
The system uses an intermediary computational layer that translates human hand posture data into gripper actuation commands. This intermediary processing layer simplifies the overall system architecture by decoupling the human interaction interface from the specific gripper mechanics, making the system more manageable despite its advanced capabilities
Data Source
AI summary
A system includes: a hand module to, based on a demonstration of a human hand grasping an object, determine first and second vectors that are normal to and parallel to a palm of the human hand, respectively, and a position of the human hand; a gripper module to determine third and fourth vectors that are normal to and parallel to a palm of a gripper of a robot, respectively, and a present position of the gripper; and an actuation module to: move the gripper when open such that the present position of the gripper is at the position of the human hand, the third and first vectors are aligned, and the fourth and second vectors are aligned; close fingers of the gripper based on minimizing a first loss; and actuate the fingers of the gripper to minimize a second loss determined based on the first loss and a third loss.