PipeBeGood – An AIoT-based Smart Water Quality Monitoring System

This paper presents PipeBeGood, an in-telligent water quality monitoring (WQM) system for water pipe systems. PipeBeGood leverages real-time multimodal data from cameras, flowmeters, and temperature sensors, along with a hybrid approach for data analysis. It employs a hybrid multimodal data analysis approach with two deep learning-based computer vision models for image analysis and rule-based approaches for flowmeter and temperature data analysis to monitor the state of pipes and their contents. As demonstrated, PipeBeGood can detect irregularities or problems and notify stakeholders accordingly. However, current image recognition models require further training due to the need for live image data with real-life objects within water, such as wet wipes.