Robot Handover Using Image Recognition to Replace Wireless Tags
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing robot apparatuses face challenges in safely and efficiently handing over objects to users, particularly due to difficulties in grasping and receiving objects with complex shapes or hazardous materials, and require additional devices like wireless tags for positional alignment.
Innovation Solution
A robot apparatus with a grasping section, recognition section, grasp planning section, and grasp control section that identifies graspable and handing-over areas of an object using image recognition, plans a path for safe handover, and controls the grasp operation to adapt to different postures and user receptivity, reducing the need for external devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If the robot hands over an object without additional devices, then the operation is simpler, but it becomes difficult to recognize graspable parts and handing-over areas
Solution Approach 1:
The patent replaces mechanical or electronic tagging systems with image recognition technology. The recognition section uses imaging to automatically identify graspable parts and handing-over areas of the object, eliminating the need for wireless tags or other additional devices while maintaining accurate region detection.
2Device complexity
If the robot uses image recognition to identify object regions, then additional devices like wireless tags are not needed, but the recognition precision may be insufficient for complex shapes
Solution Approach 1:
The grasp planning section dynamically adjusts the grasping plan based on recognition results, object properties, and user characteristics. This dynamic planning allows the system to handle complex shapes effectively by selecting appropriate graspable regions and handing-over regions, improving recognition precision without adding physical tags or markers.
Data Source
AI summary
A robot apparatus includes: a grasping section configured to grasp an object; a recognition section configured to recognize a graspable part and a handing-over area part of the object; a grasp planning section configured to plan a path of the grasping section for handing over the object to a recipient by the handing-over area part; and a grasp control section configured to control grasp operation of the object by the grasping section in accordance with the planned path.