Experimental Study of Real-Time Water Quality Analysis with Deep Neural Network (DNN) on Edge Computing

Water quality assurance within production systems is critical for both public health and environmental conservation. Conventional water quality assessment methodologies hinge on the collection and subsequent laboratory analysis of water samples-a process that is both time-consuming and costly. This sequential methodology hinders timely responses to potential water quality issues, demanding an innovative solution to facilitate immediate and on-site water quality monitoring. This study introduces the integration of DNN (Deep Neural Network) with Edge Computing to predict water quality in real-time based on data collected by sensors. Utilizing a training optimization algorithm, specifically Adam Optimizer, this approach promises to significantly reduce latency and enhance operational flexibility. The performance evaluation of the proposed system highlights its efficacy by comparing the response times between Edge Computing and traditional Cloud Computing setups. The findings suggest that implementing DNN on Edge Computing platforms can offer a more cost-effective, efficient, and timely solution for water quality monitoring, paving the way for more responsive and sustainable water management practices.