Introduction
Monitoring key physico-chemical parameters is critically important to support the expanding aquaculture industry that provides a sustainable source of seafood for people worldwide. Uncontrolled conditions may lead to severe impacts on animal health and growth rate, compromising entire cultivation systems. Although a common practice, manual assessment of physico-chemical parameters is disadvantageous to aquaculture applications in which delayed actions to control water quality account for most of animal loss cases (Lafont, 2019). Using LPWAN (Low Power Wide Area Networks) technologies may become an ally for monitoring physical-chemical parameters. This technology enables communication between radiofrequency (RF) devices with low energy consumption. Among the existing LPWAN networks, LoRa technology is a promising alternative for aquaculture farms. The LoRaWAN network layer comprises a complete solution for managing and integrating sensors into the end-user application. This work investigates the system issues associated with the practical deployment of an end-to-end cloud-based system (a) coupled with long-range low-power communication protocol (b) intended to monitor key physico-chemical parameters within an Integrated Multi Trophic Aquaculture (IMTA) farm in Brazil.
System overview
The deployed system aims to monitor key physico-chemical parameters - Dissolved Oxigen (DO), Temperature, Salinity, Conductivity, Turbidity and Ph - within a multitrophic cultivation tank from an IMTA farming facility located in Rio Grande – RS, Brazil. The system architecture (Figure 1) consists of a LoRaWAN network. The system end-node comprises four sensors, a microcontroller, and a LoRa radio with enabled LoRaWAN protocol. The AC powered end-node collects sensor measurements every 5 minutes and sends the data to a central gateway through LoRaWAN radio communication. The LoRaWAN gateway directs the received LoRaWAN packet (sensor readings) to an AI Data Analytics platform (named AIDAP), which is a cloud-based platform that integrates different sensors, allowing easy data access and visualisation for early alarm systems.
Deployment Results and Discussions
The system has been deployed in the IMTA LAB Greenhouse 5 (tank 1). The sensors were strategically positioned 1 m from the aeration tubes and 70 cm deep. The LoRaWAN gateway has been installed 35 meters away from the tank. Once operational, the end-to-end cloud-based system could presentthe collected data in real-time through the end-user application interface. Before the technology deployment, the IMTA lab farmers used to carry out manual physico-chemical measurements twice a day. The LoRaWAN system provided 240 sensor readings per day, representing a significant contribution to best support aquaculture applications. The technology deployment enabled the end user to early detect and mitigate parameter variations through an integrated cloud platform, for instance, DO levels as presented in Figure 2. The deployed system and user validation have been key assets in the farming facility in avoiding production loss and impacts on animal welfare. A series of user feedback meetings guided the technology design and optimisations to better meet IMTA’s day-to-day needs. Upcoming optimisations include local data backup, self-powering capability and a customised alarm system.
Conclusions
This case study addressed the feasibility of implementing a cloud-based end-to-end system and a LoRaWAN network architecture to support aquaculture day-to-day activities. The system monitors physicochemical parameters of the cultivation tanks in a representative IMTA facility in Brazil. The system delivered satisfactory results for quasi-real-time physico-chemical monitoring. It allowed early detection of parameter variations through a cloud-based platform. The immediate mitigation action played an important role in preventing impacts on aquaculture production and animal welfare. The end-to-end system optimised the data acquisition process, centralizing and standardizing the data collection. It also increased data acquisition frequency, best supporting aquaculture management. The user feedback provided valuable insights to guide technology optimization. A few system limitations could be established with this practical deployment, including areas of difficult access which can lead to data communication issues (e.g. packet loss). Such limitations will be addressed in future versions of the system.
Acknowledgements
This work is part of the ASTRAL (All Atlantic Ocean Sustainable, Profitable and Resilient Aquaculture) project. This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement Nº 863034.
References
M. Lafont, S. Dupont, P. Cousin, A. Vallauri, and C. Dupont. 2019. Back to the future: IoT to improve aquaculture : Real-time monitoring and algorithmic prediction of water parameters for aquaculture needs. In 2019 Global IoT Summit (GIoTS). 1–6. https://doi.org/10.1109/GIOTS.2019.8766436