Haptic Robot Teaching for Force Feedback Task Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional robot learning methods struggle to effectively handle tasks requiring force information, particularly in dynamic environments, due to limitations in standardizing processes and adapting to various situations, leading to inefficiencies in automation across industries like electrical, electronic, and logistics.
Innovation Solution
A robot system utilizing a haptic manipulating device to transfer force information, with a sensing unit and learning unit that detect and learn from force data, allowing the robot to autonomously perform tasks by reflecting the user's intention and adapting to changing environments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional robot learning methods (teaching pendant, kinesthetic teaching, teleoperation, hard coding) are used, then the robot can perform simple and repetitive tasks rapidly and precisely, but the robot cannot effectively handle tasks requiring force information and cannot adapt to various situations in new industry fields
Solution Approach 1:
The patent introduces a haptic manipulating device that provides force feedback to the user during robot teaching. The device transmits force information from the robot's interaction with objects back to the user's hands, enabling the user to sense and control the force applied by the robot. This feedback mechanism allows the robot to learn tasks requiring force information through demonstration, resolving the contradiction between maintaining high productivity for simple tasks and gaining adaptability for complex force-sensitive tasks.
Solution Approach 2:
The haptic manipulating device serves as an intermediary between the user and the robot. It translates the user's manual manipulation forces into robot control commands while simultaneously providing haptic feedback about the robot's interaction forces to the user. This intermediary device enables effective communication of force information that neither conventional teaching methods nor direct robot programming can achieve, allowing the robot to learn diverse tasks across different industry fields.
2Ease of manufacture
If conventional teaching methods are used, then the programming process is simple for standardized tasks, but extensive programming is required for each task and flexibility for complicated tasks is low
Solution Approach 1:
The patent enables task copying through physical demonstration. Instead of programming each task individually, a user can demonstrate a task by manually manipulating the haptic device, and the robot copies this demonstrated behavior. The haptic manipulating device captures the force information and motion patterns during demonstration, storing them as learnable data. This copying approach dramatically reduces programming complexity while maintaining ease of teaching, as users can demonstrate complex tasks intuitively without writing code.
Solution Approach 2:
The robot performs self-learning by processing the force information and demonstration data captured during teaching. The learning unit automatically processes the demonstrated tasks and generates executable programs without requiring manual programming intervention. This self-service capability eliminates the need for extensive programming for each task while maintaining simplicity in the teaching process, as the robot learns autonomously from demonstrations.
3Productivity
If standardization is applied to robot tasks, then simple repetitive tasks can be performed efficiently, but tasks in new industry fields with short production line lifspans and non-standardized processes cannot be handled flexibly
Solution Approach 1:
The patent introduces dynamic adaptability through force information-based learning. The robot system can dynamically adjust its behavior based on real-time force feedback during task execution. For standardized tasks, the robot maintains efficient automated operation, while for non-standardized tasks in new industry fields, the robot can adapt by learning new force patterns through demonstration with the haptic device. This dynamic capability allows the same robot to efficiently handle both standardized high-volume tasks and customized low-volume tasks without requiring complete reprogramming.
Solution Approach 2:
The patent enables parameter changes in robot behavior through force information. By capturing force magnitudes, directions, and patterns during demonstration, the system learns to adjust operational parameters dynamically. For example, the robot can learn to apply different forces for different materials, adjust grip strength based on object properties, or modify motion profiles based on interaction forces. This parameter adaptability allows efficient handling of both standardized tasks with fixed parameters and non-standardized tasks requiring dynamic parameter adjustment.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables robots to learn and perform complex tasks efficiently, reducing the need for extensive programming and allowing flexible handling of various situations, thereby improving automation capabilities in diverse industries.
Implementation Method 1
an interface device formed to be manipulable to teach the robot device, and configured to output a haptic feedback based on the force information while teaching of the task to the robot device is performed
Implementation Method 2
a robot device configured to perform a task, and to detect force information related to the task
Data Source
Figure 1A
Figure 1B
Figure 2
AI summary
The present invention relates to a method for learning a robot task, and a robot system using the same. The robot system includes: a robot device configured to perform a task, and to detect force information related to the task; an interface device formed to be manipulable to teach the robot device, and configured to output a haptic feedback based on the force information while teaching of the task to the robot device is performed; a sensing unit configured to sense first information related to a task environment of the robot device, and second information related to a driving state of the robot device, while the teaching is performed by the interface device for outputting the haptic feedback; and a learning unit configured to learn a motion of the robot device related to the task, by using the first information and the second information, such that the robot device autonomously performs the task.