Adaptive path finding algorithm in dynamic environment for warehouse robot

Warehouse robots have been widely used by manufacturers and online retailer to automate good delivery process. One of the fundamental components when designing a warehouse robot is path finding algorithm. In the past, many path finding algorithms had been proposed to identify the optimal path and im...

Full description

Saved in:
Bibliographic Details
Published in:Neural computing & applications 2020-09, Vol.32 (17), p.13155-13171
Main Authors: Ng, Mun-Kit, Chong, Yung-Wey, Ko, Kwang-man, Park, Young-Hoon, Leau, Yu-Beng
Format: Article
Language:English
Subjects:
Adaptive algorithms
Algorithms
Artificial Intelligence
Barriers
Computational Biology/Bioinformatics
Computational Science and Engineering
Computer Science
Data Mining and Knowledge Discovery
Image Processing and Computer Vision
Industrial robots
Path predictors
Probability and Statistics in Computer Science
Robots
S.I. : Green and Human Information Technology 2019
Subtraction
Warehouses
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Warehouse robots have been widely used by manufacturers and online retailer to automate good delivery process. One of the fundamental components when designing a warehouse robot is path finding algorithm. In the past, many path finding algorithms had been proposed to identify the optimal path and improve the efficiency in different conditions. For example, A* path finding algorithm is developed to obtain the shortest path, while D* obtains a complete coverage path from source to destination. Although these algorithms improved the efficiency in path finding, dynamic obstacle that may exist in warehouse environment was not considered. This paper presents AD* algorithm, a path finding algorithm that works in dynamic environment for warehouse robot. AD* algorithm is able to detect not only static obstacle but also dynamic obstacles while operating in warehouse environment. In dynamic obstacle path prediction, image of the warehouse environment is processed to identify and track obstacles in the path. The image is pre-processed using perspective transformation, dilation and erosion. Once obstacle has been identified using background subtraction, the server will track and predict future path of the dynamic object to avoid the obstacle.
ISSN:0941-0643
1433-3058
DOI:10.1007/s00521-020-04764-3