Robot Item Selection Using 3D Clearance Prisms
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vision-guided robotics systems struggle to efficiently retrieve items from a stack using a lateral-motion gripper, as they need to assess clearance in 3D Point Cloud data from Depth Sensors to avoid collisions.
Innovation Solution
A retrieval controller that processes depth maps from Depth Sensors to create 3D Point Clouds, segments items, computes enclosing prisms, determines grasp points, and selects items with sufficient clearance for retrieval by a robot.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a lateral-motion gripper is used to retrieve items, then the robot can approach items from the side, but it cannot grasp items if the path is blocked by other items
Solution Approach 1:
The system segments the 3D point cloud data into individual item representations, allowing the robot to analyze clearance for each item separately. This segmentation enables the lateral-motion gripper to identify valid grasp targets by evaluating spatial relationships between items, resolving the contradiction between approach capability and grasp reliability.
2Adaptability or versatility
If the robot uses sensor readings to locate items and destinations, then it can adapt to variations in locations, but the system complexity increases
Solution Approach 1:
The system replaces complex mechanical sensing and localization mechanisms with depth map computing and 3D point cloud processing. By using depth sensors to capture spatial information and algorithmic processing to interpret it, the system achieves location flexibility while keeping the physical hardware relatively simple.
3Measurement precision
If depth sensors are used to create 3D point clouds for item location, then retrieval accuracy improves, but processing complexity increases
Solution Approach 1:
The system performs preliminary processing of depth map data into 3D point clouds and further segments them into item representations before the actual retrieval decision is made. This preliminary action organizes the complex spatial data into manageable structures, reducing the processing burden during critical retrieval operations.
Data Source
AI summary
A retrieval controller is disclosed for identifying an item to be retrieved from a flat storage surface by a robot for identifying one item, amongst items stored on a common surface that can be retrieved by a robot equipped with a lateral-motion gripper. The retrieval controller includes: a depth map computing unit configured to establish a global coordinate system, and establish an orthonormal set of basis vectors u, v and w defined in the global coordinate system, where w is approximately orthogonal to the surface that the items are stored on; a vector determination unit; an item selection unit; and a robot instructing unit configured to instruct the robot to retrieve the item based on uv coordinates of one or more w-aligned edges having associated quadrilateral-based, right prisms which do not have interiors that intersect any right, enclosing prisms selected by the item selection unit.