Smart Urban Gardening Companion System

What it does

The earth's carbon dioxide concentration has been significantly increased about 10 times from the past decades. most citizens living in busy cities where they don't have the time and efforts to do gardening even they have the interest to save the earth.

Your inspiration

The inspiration came from the growing number of asthma patient among my friends and relative due to bad environmental condition and the lost of interest in gardening as its my mom's all time favorite activity. I rarely seen household doing gardening around the city and most of them bought the fake plants to beautify their homes. From there, i have some multiple analysis from ground surveying to questioning each gardening businesses. Then, i found out most of the reasons are due to lack of efforts and time in handling the gardening habits and requirement tons of commitment to water and fertilize the plant regularly.

How it works

Based on the problem statement, I have designed a complete flow of a autonomous gardening companion system where it involved in Machine Learning, Internet of Things, Android Apps, Remote-Monitoring and Beautification technologies. The project consist of personalized hardware and software where multiple electronic toolkit such as multiple sensors, remote-reservoir for watering plant, remote-camera for all-time monitoring, auto-fertilizing mode. The hardware is controlled by a control system made with Machine Learning that able to understand and adapt the environmental behavior which suit the most for the plant's growth rate. The adaptation is based on a huge amount of data as the input for our Machine Learning system for training. Once the training process is matured and the plant can be self-maintain without requiring any effort or commitment to take care of it. I also created a custom mobile application that allow users to visit their plants remotely.

Design process

Concept Design: - List out the required sensors and other electronic components - Draw state diagram for software progress - Understanding the inputs for the Machine Learning model Hardware: - Reconfigure all the electronic components such as The light sensors, Temperature & Humility sensors, Arduino controller, IR infrared obstacle sensor, PH meter sensor, Raspberry Pi 3 Model B controller and other electronic peripherals. - Compile all the electronic components and attached them together into one single module - Test and troubleshoot possible electronic flaws before reaching out the main software framework - Finalized the hardware by attaching the software into the system and run a complete flow of the progress. Software: - First we collect a huge pile of data from the hardware sensors that we implemented and allow the machine learning to digest and find the pattern of the follow dataset where how it could allow the plant to self-adapt the environment and maintain the growth rate - The mobile application is attached with IoT feature that consist of a centralized cloud system that link between the hardware and software

How it is different

I yet to see any product that come close to this as the following design offers a complete package from head to toe in handling gardening autonomously without require any further effort and commitment to own a beautiful and fresh plants around your household. The current products we have seen are still requiring human monitoring and manual fertilizing.

Future plans

I have reached numerous housewife and households who are interested in gardening in city but don't have the time and effort to maintain it and most of them are waiting to see the real product so that they could implement it at their home. I have started to restructure and redesign the hardware to look like a real product with lower price tag compared with the material we used for prototyping in mass production. I believed that the product can attract many youngsters on how fun is gardening.

Awards

Best of the Best project award in CAPSTONE - Universiti Teknologi Malaysia