Climate Smart IoT Based Agriculture Monitoring System : client side monitoring system for improving climate change adaptation
Aliche, kelvin (2024)
2024
All rights reserved. This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-2024121736829
https://urn.fi/URN:NBN:fi:amk-2024121736829
Tiivistelmä
The aim of this thesis focuses on the development of a client-side monitoring system for the climate-smart agriculture system aimed at addressing the challenges posed by climate change using visualiza-tion tools and sustainable technology. This study wasn’t part of the pilot implementation of the SMAGRI project but it played a crucial role in the development of the client side monitoring platform for monitoring water management and data handling on farms. Utilizing modern data-driven techniques for monitoring critical environmental parameters such as soil moisture, temperature, and humidity using an interactive interface for visualizing, controlling and monitoring the wireless sensors buried 15cm to 60cm under-ground with a battery life of over 20 years without maintenance, the system enables farmers to make informed decisions in response to climate variability. These sensors communicated with a dedicated Base station tower, The weather tower is connected to the Raspberry Pi via RS-485, relaying its data to Savonia ThingsBoard for efficient storage and analysis.
The ThingsBoard API was used to extract real-time sensor data, which was integrated into a React-based web interface supported by Node.js. Geographic Information System (GIS) mapping with Leaflet was ap-plied to visualize the locations of the sensors, facilitating farm zone management. Predictive analysis was done using the random forest regression to analyze historical data and predict future soil moisture levels, supporting informed water management decisions.
The stored sensor data was tested in different regression models with the aim to make future predictions of the moisture content, this proved possible using statistical prediction resulting to an error rate of 0.5% off the accurate readings during monitoring and validation. Future improvements would include the inte-gration of more accurate machine learning models for predictive analytics and weather forecasting tools. In conclusion, this thesis offers a scalable monitoring solution for sustainable water management, sup-porting agricultural adaptation to climate change.
The ThingsBoard API was used to extract real-time sensor data, which was integrated into a React-based web interface supported by Node.js. Geographic Information System (GIS) mapping with Leaflet was ap-plied to visualize the locations of the sensors, facilitating farm zone management. Predictive analysis was done using the random forest regression to analyze historical data and predict future soil moisture levels, supporting informed water management decisions.
The stored sensor data was tested in different regression models with the aim to make future predictions of the moisture content, this proved possible using statistical prediction resulting to an error rate of 0.5% off the accurate readings during monitoring and validation. Future improvements would include the inte-gration of more accurate machine learning models for predictive analytics and weather forecasting tools. In conclusion, this thesis offers a scalable monitoring solution for sustainable water management, sup-porting agricultural adaptation to climate change.