3D Sensor Hypothesis Ranking for Carton Unloading
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing automated single box picking and decanting systems rely on pre-defined parameters and require extensive training, retraining, and collection of large datasets, making them inefficient and prone to errors, especially when handling varied box types and orientations.
Innovation Solution
A machine vision-based system utilizing 3-D sensors and approximation algorithms to generate and rank hypotheses about box poses, eliminating the need for training and arbitrary parameters, and enabling simultaneous processing of printed and unprinted boxes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If pre-defined parameters and extensive training are used in automated single box picking systems, then the system can handle box identification and positioning, but the system becomes inefficient and prone to errors when handling varied box types and orientations
Solution Approach 1:
The system transitions from using pre-defined parameters requiring extensive training to a machine vision-based system that uses 3-D sensors and approximation algorithms. This parameter change eliminates the need for extensive training datasets while improving the system's ability to handle varied box types and orientations efficiently and reliably
2Adaptability or versatility
If extensive training and large datasets are collected for box identification, then the system can recognize different box types, but the system requires significant time and computational resources for training and retraining
Solution Approach 1:
The system replaces the traditional machine learning training process with a machine vision-based approach using 3-D sensors and approximation algorithms. This substitution eliminates the need for collecting large datasets and performing extensive training, while maintaining the ability to recognize and adapt to different box types and orientations
3Measurement precision
If pre-defined parameters are used for box positioning, then the system can locate boxes in preset locations, but the system cannot easily adapt to new boxes without training or configuration
Solution Approach 1:
The system transitions from static pre-defined parameters to a dynamic machine vision-based system that can adapt to new box types and orientations in real-time. The 3-D sensors and approximation algorithms enable the system to dynamically adjust to varying box configurations without requiring retraining or reconfiguration, while maintaining precise location accuracy
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
The system facilitates faster and more reliable unloading of cartons by eliminating the need for extensive training, reducing errors, and allowing for quick adaptation to new box types, while unified information sources improve accuracy and efficiency.
Implementation Method 1
The at least one sensor has a set of radiation sensing elements which detect reflected, projected radiation to obtain 3-D sensor data
Data Source
AI summary
A machine vision-based method and system to facilitate the unloading of a pile of cartons within a work cell are provided. The method includes the step of providing at least one 3-D or depth sensor having a field of view at the work cell. Each sensor has a set of radiation sensing elements which detect reflected, projected radiation to obtain 3-D sensor data. The 3-D sensor data including a plurality of pixels. For each possible pixel location and each possible carton orientation, the method includes generating a hypothesis that a carton with a known structure appears at that pixel location with that container orientation to obtain a plurality of hypotheses. The method further includes ranking the plurality of hypotheses. The step of ranking includes calculating a surprisal for each of the hypotheses to obtain a plurality of surprisals. The step of ranking is based on the surprisals of the hypotheses.