River water quality is crucial for the environment, biodiversity and human health. To improve the decision making process and, in turn, enhance the river health, the Artificial Intelligence (AI) framework for water quality monitoring, prediction and alarming is proposed. This framework relies on the existing water quality and quantity measuring system capable of providing only daily water quality data at several sites in Serbia. It will augment and improve the existing measuring system by providing the real time water quality data from multiple sensors at the sub-daily (or end-user desired) scale which will be installed on smart floating buoys. Therefore, compact smart IoT (Internet of Things) buoys with water quality sensors will improve temporal resolution of the water quality measurement. In addition, this project enables the assessment of water quality alongside
the river stream by using water sampler equipped drones. The drones will collect the information from different river sections, at designated GPS points between the installed buoy based sensors capturing the water quality transformation, thus improving spatial resolution of the water quality measurement. Another added value of this project is that buoy and drone measurements will be coupled with satellite observations. The Danube river in Serbia is selected for the framework implementation considering the highly populated environment and its vulnerable ecological status. Existing river environmental monitoring systems in Serbia are spatially, temporally, and information sparse and do not provide full insight on the pollutants in specific hazardous areas (known as “hot spots”). The demonstration of new tools using smart sensing, IoT, and AI predictive models for data analysis and visualization and, through various dissemination events, connecting government institutions, researchers, and stakeholders to scale the effort, will have a clear impact of raising awareness on environmental pollution in Serbia.