Institutional Repository [SANDBOX]
Technical University of Crete
EN  |  EL

Search

Browse

My Space

Robotic bee swarm crop pollination

Mamis Konstantinos

Full record


URI: http://purl.tuc.gr/dl/dias/8C30FF6D-065B-44F9-BB15-418060421041
Year 2021
Type of Item Diploma Work
License
Details
Bibliographic Citation Konstantinos Mamis, "Robotic bee swarm crop pollination", Diploma Work, School of Production Engineering and Management, Technical University of Crete, Chania, Greece, 2021 https://doi.org/10.26233/heallink.tuc.90553
Appears in Collections

Summary

This thesis aims to determine the feasibility of a robotic bee swarm algorithm, which could pollinate a crop field effectively. This endeavor will provide a solution to the pollinator decline issue. Moreover, this thesis fills the gap in the literature about robotic bee swarms and whenever they can exist in the near future. Swarm robotics is a field of multy-robotics, inspired by the collective behavior of bug colonies, animal packs and social insects. The swarm consist of simple robots which cooperate in order to perform a complex task. Inspired by ant colonies, the Ant Colony Optimization algorithm uses a number of robot agents in order to find a good path through a graph.The algorithm was developed with Matlab. In the Matlab environment, a graph in x and y axes with random nodes is created. The graph’s acreage represents the crop field’s acreage. Nodes in the graph represents florets of Helianthus annus. A random node gets chosen as a starting point (this node represents the e-hyve). The robot bees/ agents, perform The ACO algotithm in order to locate the open florets in the field, and find a good path in the graph. Robot bees of the colony have limited fly time (battery), the same fly speed and they collect a random amount of pollen from each flower. The algorithm ends if the majority of florets have been visited, have no pollen left and all bees are back to the e-hyve).In order to verify the accuracy of the results, the variable of the algorithm have been filled with data collected from real bees. After the completion of the process of the algorithm, its results regarding the footfall per 100 flowers and the average traffic, were compared with 2 different, independent studies for sunflower cultures. The algorithms results were found to be really similar to those of the studies

Available Files

Services

Statistics