Open-source low-cost design of a buoy for remote water quality monitoring in fish farming.

In this paper we present the design of an open-source and low-cost buoy prototype for remote monitoring of water quality variables in fish farming. The designed battery-powered system periodically measures temperature, pH and dissolved oxygen, transmitting the information locally through a low-power wide-area network protocol to a gateway connected to a cloud service for data storage and visualization. We provide a novel buoy design that can be easily constructed with off-the-shelf materials, delivering a stable anchored float for the IoT device and the probes immersed in the water pond. The prototype was tested at an operating fish farm, showing promising results for a low-cost remote monitoring tool that enables automatic data acquisition and storage in fish farming scenarios. All the elements of this design, including hardware and software designs, are freely available under permissive licenses as an open-source project.

Machine learning for manually-measured water quality prediction in fish farming.

Monitoring variables such as dissolved oxygen, pH, and pond temperature is a key aspect of high-quality fish farming. Machine learning (ML) techniques have been proposed to model the dynamics of such variables to improve the fish farmer's decision-making. Most of the research on ML in aquaculture has focused on scenarios where devices for real-time data acquisition, storage, and remote monitoring are available, making it easy to develop accurate ML techniques. However, fish farmers do not necessarily have access to such devices. Many of them prefer to use equipment to manually measure these variables limiting the amount of available data to process. In this work, we study the use of random forests, multivariate linear regression, and artificial neural networks in scenarios with limited amount of measurements to analyze data from water-quality variables that are commonly measured in fish farming. We propose a methodology to build models in two scenarios: i) estimation of unobserved variables based on the observed ones, and ii) forecasting when a low amount of data is available for training. We show that random forests can be used to forecast dissolved oxygen, pond temperature, pH, ammonia, and ammonium when the water pond variables are measured only twice per day. Moreover, we showed that these prediction models can be implemented on a mobile-based information system and run in an average smartphone that fish farmers can afford.