A Low-Cost Sensor Network System for Reliable Small Stream Monitoring

Resilient monitoring of streams with dense sense sensor networks requires a system that manages and monitors everything from sensor node deployment through data collection, deployed equipment health, data quality analysis, and visualization.  The Clemson University Intelligent River® project has developed a range of technologies to monitor water in small urban streams and beyond using internet-connected devices that can stream data in near real-time. By leveraging the latest Internet of Things (IoT) advances, these sensor systems can increase the density of water measurements in both the waterways and connected stormwater pipes to help monitor the storm and drought response of these streams to help understand their function. These Intelligent River® sensor nodes have the added advantage of being able to interface into the latest sensor technologies as they are developed and to the best-in-class commercial water sensors. Spatially-dense measurements of water quantity and quality utilizing newly developed Intelligent River® real-time monitoring technology includes integrated low-cost sensors, embedded computers, and metadata management and visualization system. These technologies have the potential to significantly improve how water can be monitored in a range of situations and environments (e.g., river water monitoring and modeling and well water level). The recent advances in cellular Internet-of-Things (IoT) and Low Power Wide Area Network (LPWAN) technologies have increased the communication range and optimized the operating time for the wireless sensor network nodes. We will discuss a low-cost IoT system with near real-time hydrologic reporting using a sensor network of power-optimized embedded computers linked to a cloud-based back end system that serves as a data repository and interfaces to commercial IoT platforms for machine-learning-based anomaly and event detection. The designed sensor network is an end-to-end power-efficient system capable of adaptive sensing, remote data-logging which uses either LoRaWAN or low-power LTE cellular networks to send observations to a cloud-based data repository. We will discuss the design installation, and data results for deployments that measure stream level, flow, and water quality.  Overall network and data reliability is examined through data analysis of over a year of system deployment.