maragh_s
Abstract:
The purpose of my project is twofold: My major project is to continue the hydroponics automation project started this year and the other is to integrate software and robotics by creating a chess program that utilizes machine learning.
Introduction and Goals:
The hydroponics automation is an extension of the sustainable housing research and development I created last year. The goals are to continue the development of sustainable housing and to create a second hydroponic installation for the elab which will be specific to certain crops. This project is based on my desire to stop the ongoing energy, housing and agricultural crisis in my home country of Jamaica. The other project I have started was based on my desire to improve at chess. The problem today with chess computers is that they play like a computer however my program with the integration of machine learning will adopt my own playstyle and become the best chess player I could theoretically be.
Implementation and challenges:
The second installation of the the hydroponics will be almost the same as the first. Large hollow PVC towers and wooden pallets are the foundation of each installation which are based on three methods: Kraty or horizontal method, vertical or tower method and flood and drain method. Each of them work with a five gallon water supply and a pump (except for the Kratky). The main challenges of this project are growth of fungus on the plants and death of seedlings ( can be circumvented with the use of pre purchased seedlings instead of growing them from seeds.) as well as the rat lungworm disease which can make a crop completely inedible. Also weather due to Waimea’s heavy winds can knock over the larger plants with extended branches and leaves. The chess project is based on the Stockfish engine which is opened sourced and can be edited in C++ and will run on a Raspberry Pi. The motorized armature will be 3D printed and will be attached to a chessboard as well as able to play without internet connection. The challenges in this project are due to my coding limitation as my major experience is with Python as well as the integration of machine learning with the Tensor flow program. This requires me to continually play against the engine in order to teach it my playstyle and have it retain it for future matches.
Impact and Legacy:
The hydroponics installations will remain at the elab and will be continually use for future classes as by the end of the year it will be completely automated and will require very little human input other than to add additional seedlings and follow the instructions given by the computer. The project will also be able to feed more and more people as the crops improve and more installations are added. The chess project will remain in the elab to challenge anyone who wishes to test the abilities in chess. The hope is have chess tournaments in the elab with the sole opponent being the computer so that it can continually learn new styles and get better.
Resources and Possible costs:
3D printed armature- from printer in elab.
A possible 3 motors for joints on the armature
New seedlings for the hydroponics
Nutrients and Ph regulators.
Raspberry Pi
Additional PVC towers
Collaborators:
Oliver Grayson
Tommy Truong
Sneha Nair
Comments