Paper Info
Title
Development of an environment characterization methodology for optimal design of an agricultural robot.
Abstract
PurposeThe purpose of this paper is to describe a methodology for characterization of the robot environment to help solve such problem as designing an optimal agricultural robot for a specific agricultural task.Design/methodology/approachDefining and characterizing a task is a crucial step in the optimization of a task-specific robot. It is especially difficult in the agricultural domain because of the complexity and unstructured nature of the environment. In this research, trees are modeled from orchards and are used as the robot working environment, the geometrical features of an agricultural task are investigated and a method for designing an optimal agricultural robot is developed. Using this method, a simplified characteristic environment, representing the actual environment, is developed and used.FindingsCase studies showing that the optimal robot, which is designed based on the characteristic environment, is similar to the optimal robot, which is designed based on the actual environment (less than 4 per cent error), is presented, while the optimization run time is significantly shorter (up to 22 times) when using the characteristic environment.Originality/valueThis paper proposes a new concept for solving the robot task-based optimization by the analysis of the task environment and characterizing it by a simpler artificial task environment. The methodology decreases the time of the optimal robot design, allowing to take into account more details in an acceptable time.
Year
Venue
Field
2017
Industrial Robot
Robot learning,Agricultural robot,Simulation,Computer science,Robot design,Optimal design,Control engineering,Robot
DocType
Volume
Issue
Journal
44
1
Citations 
PageRank 
References 
0
0.34
4
Authors
3
Name
Order
Citations
PageRank
Victor Bloch100.34
Avital Bechar2334.82
Amir Degani33010.15