Development of Trash Aggregating and Water Quality Monitoring System for Aquatic Habitat Using Machine Learning

EasyChair Preprint 6325

Abstract

Floating trash/waste present in the water bodies emerge as a major environmental problem that endangers the lives of its inhabitants and makes the water unsuitable for drinking by altering its properties. The existing method for monitoring and collecting trash relies on the manual inspection by dispatching inspectors to the field, periodically. This process is time consuming and involves a lot of human intervention. The system proposed in here uses Unmanned Robotic Boats (URBs) mainly used for many real-time monitoring applications. In here, we propose an automated river trash aggregator system, also being able to check the quality of water, consisting of a remote processing unit, Arduino UNO, Wi-Fi module, temperature, pH and Conductivity Sensor. The values from these sensors are sent to Linode cloud, where machine learning algorithm is applied to analyze the aquatic habitat. The robot includes a MQTT architecture and machine learning model to predict whether the conditions of water is Good or Bad. A data set is generated for training and testing the machine learning network, specifically for floating trash detection application. The system finally enables the URB to communicate wirelessly with a remote computer in a real-time manner using IoT.

Keyphrases: Linode Cloud, MQTT Architecture, Unmanned Robotic Boats (URBs)

@booklet{EasyChair:6325,
  author    = {J J Jijesh and Shivashankar and S M Akanksha and S Anitha and H S Manasa and Kusuma Santhosh},
  title     = {Development of Trash Aggregating and Water Quality Monitoring System for Aquatic Habitat Using Machine Learning},
  howpublished = {EasyChair Preprint 6325},
  year      = {EasyChair, 2021}}