Solar energy, water, algae and plankton – that’s all the system from the startup Blue Planet Ecosystems needs to farm fish fully automatically. It has container size and can be placed in any location. With the Internet of Things, it is remotely monitored, controlled and continuously optimized.
The challenge: Fish farming is costly and labor-intensive
The solution: The networked bioreactor for fish farming
Sensors regularly measure temperature, light, oxygen content of the water and other parameters. Cameras monitor the school of fish and by means of artificial intelligence data about their condition. Software uses this information to regulate light and water temperature, as well as control the water pumps and automatic feeding. In addition, all values are recorded in order to optimize the bioreactor by means of data analyses. Manual control is even possible via a web interface to be able to intervene in case of problems.
The overall system consists of two subsystems: Regulation and control. The control system is responsible for reading the sensors, interpreting the data and controlling actuators for feeding, for example. The prototype uses a Raspberry Pi for this purpose. In mass production, this device is replaced by a custom-built controller that is precisely adapted to its tasks. The regulatory system is internal and its primary function is to regulate fish farming.
The control system connects to the Internet via a router using WLAN or mobile communications. At the same time, it is connected to the controller via the MQTT(Message Queuing Telemetry Transport) network protocol and exchanges telemetry data with it. All data is sent to the cloud. There are two storage locations there: a database for its visualization and a data lake for the analysis of the aggregated data. The platform also sends commands to the system that can be used to override the internal control. The connection is protected with encryption in both directions.
The components of the solution, whether microcontroller or single board computer can also be remotely maintained / updated. This allows changes to be easily and quickly transmitted to test systems or to customers.
The result: Simple fish farming for any location
The modular design allows the entire system to be used for service with very little effort. The system works autonomously, but regularly sends status information to the cloud. In addition to being simple and straightforward to grow fish at any location, the technology is easy to maintain and defective components can be quickly removed. The MQTT network protocol and the selected architecture also allow easy upgrading with additional sensors and actuators. This makes it possible to permanently update the prototype and later the series version and adapt them to new requirements.