Robot Arm Surface Input Mapping for Fast Contact Commands
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for interacting with robot arms suffer from significant delays and low accuracy in detecting contact positions, limiting their application to single contact points, making them inefficient for intuitive operation, especially in complex interactions.
Innovation Solution
A method where predefined input points or areas on the robot arm's surface are marked, allowing operators to easily recognize and interact with specific command-generating areas, using two-dimensional images applied to the three-dimensional surface, enabling intuitive operation and expanding functional capabilities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If tactile sensors and machine learning algorithms are used to detect contact forces and estimate contact positions, then the robot arm can interact with objects, but the system experiences significant delay due to algorithm execution time
Solution Approach 1:
The patent pre-calculates and stores mapping relationships between contact forces and contact positions in a lookup table before operation. When a contact occurs, the system directly queries the pre-computed table rather than executing complex algorithms in real-time, thereby eliminating algorithm execution delay while maintaining interaction capability
Solution Approach 2:
The patent creates a simplified virtual model (lookup table) that copies the essential mapping between contact forces and positions from the complex physical system. This virtual representation allows rapid query-based determination of contact positions without replicating the computational complexity of the original sensing and analysis algorithms
2Adaptability or versatility
If machine learning algorithms are used to estimate contact position, then the robot can detect contact forces, but the accuracy in estimating contact position on the surface is low
Solution Approach 1:
The patent replaces complex machine learning algorithms with a deterministic physics-based calculation approach. By using the known robot arm geometry and force sensor measurements, the system calculates contact position through geometric relationships and force moment analysis, providing higher precision and repeatability compared to probabilistic machine learning methods
Solution Approach 2:
The patent transforms the problem from estimating contact position directly from sensor data to calculating it from known geometric parameters and force measurements. By changing the approach from data-driven estimation to physics-driven calculation using precise geometric parameters, the system achieves higher measurement accuracy
3Ease of operation
If operators interact with the robot arm by applying force at undesignated points, then the robot can receive input, but the operation lacks intuitiveness and requires understanding of complex force-position relationships
Solution Approach 1:
The patent assigns different command functions to different locations on the robot arm surface by marking specific regions. Each marked region corresponds to a predetermined command, allowing operators to intuitively select functions by touching specific locations rather than understanding complex force application requirements. This creates local functional differentiation that simplifies operation
Solution Approach 2:
The patent introduces marked regions on the robot arm surface as intermediaries between the operator and the control system. These markings serve as visual guides that mediate the interaction, translating the operator's intuitive touch actions into specific commands without requiring the operator to understand the underlying force sensing and calculation mechanisms
Data Source
Figure 1~2
Figure 3a~3b
Figure 4~5
AI summary
The invention relates to a method for optimizing the function of a robot arm, comprising the steps of: marking an input point or an input surface on the surface of the robot arm, when this point or surface is pressed by an operator, a force and/or torque is detected in at least one joint, and deriving the position of the contact point on the surface of the robot arm, so that a command is derived which is to be triggered by pressing this input point and/or the input surface.