Automation of Hydroponic Installations using a Robot with Position Based Visual Feedback - Robotics Institute Carnegie Mellon University

Automation of Hydroponic Installations using a Robot with Position Based Visual Feedback

Niels Tanke, Guoming Alex Long, Dhruv Agrawal, Abhinav Valada, and George A. Kantor
Conference Paper, Proceedings of 3rd International Conference of Agricultural Engineering (CIGR '12), July, 2012

Abstract

This paper presents a robot for the automation of hydroponic farms using Position Based Visual Feedback. In recent years advances in agricultural engineering have resulted in higher crop yields and more environmentally friendly practices. To meet the increasing world food demand even more advances in agriculture have to be made. A way to do this is by intensifying agriculture systems. Hydroponics and robotics are proven areas that contribute to intensifying agricultural practices, but automating these hydroponics systems using current automation techniques requires a large capital investments. The designed system is a low cost and simple addition to existing Nutrient Film Technique infrastructures. This robot manipulates seedlings and plants without human intervention and can be used as a monitoring system to give parameters of the crop and environment to the grower. A low cost alternative to stereo cameras, a Microsoft Kinect and a Position Based Visual Feedback algorithm were used to detect the plants and position the robot. It is demonstrated that by using a Microsoft Kinect the positioning is accurate enough to manipulate plants on an hydroponic system.

BibTeX

@conference{Tanke-2012-7525,
author = {Niels Tanke and Guoming Alex Long and Dhruv Agrawal and Abhinav Valada and George A. Kantor},
title = {Automation of Hydroponic Installations using a Robot with Position Based Visual Feedback},
booktitle = {Proceedings of 3rd International Conference of Agricultural Engineering (CIGR '12)},
year = {2012},
month = {July},
keywords = {Hydroponics, Nutrient Film Technique, Kinect, Position Based Visual Feedback},
}