2024 ASEE Annual Conference & Exposition

Eastern North Carolina farmlands are often below the standing water level, requiring constant drainage to avoid flooding. Irrigation canals collect water from multiple farms, and they must be emptied regularly to avoid crop losses and damage to houses and equipment. Maintaining a low water level in these canals is a critical component of the daily tasks farmers must manage. In most cases, this process requires a visit to the location of the pumps and turning them on for a given amount of time, a manual, time-consuming activity. In this paper we discuss the development of an automated solution using ultrasound sensors to measure surface water levels networked over LoRaWAN with actuators that can remotely turn the pump on and off. Machine Learning based automation algorithms provide workflow optimization and necessary redundancy is built in. The solution can be customized to the specific performance characteristics of the pump being utilized. Farmers are provided with water level visualization tools accessible on mobile devices as well as automatic, intelligent notifications to help address failures and unusual circumstances. Future expansion options for this solution, such as integration of weather forecast and live weather data, are discussed.