The goal of FishAI is to provide an ecologically and economically sustainable supply of fish products through recirculating aquaculture systems.
1
2
An intelligent, data-based feeding system is designed to make feed use more efficient and thus increase plant productivity while saving operating costs.
3
The basis is formed by autonomous sensors of the smart image recognition, with which a biomass determination and evaluation of the feed intake should be possible.
in a nutshell
What?
The FishAI research project aims to increase the economic efficiency of RAS plants and thus establish a sustainable supply of fish products in Germany in accordance with the NASTAQ goals for 2030. The focus is on achieving the following goals:
- Increasing the degree of automation and reducing the operating costs of RAS systems.
- Improvement of animal welfare by optical, non-invasive measuring methods to reduce stress & risk of injury and by optimal feed quantities adapted to environmental conditions (e.g. water temperature)
- optimization of production & feed supply and thereby reduction of resource input (here: feed)
- intelligent data evaluation to provide forecasts regarding production quantities and product quality
How?
To achieve these goals, FishAI is developing an intelligent, data-based feeding system that can be integrated into existing RAS facilities. The system is based on four core components: FishSizer (sizing), FeedWatch (feed intake), Dynamic Growth Models, and Stocking Management Dashboard.
The basis is formed by autonomous sensors of the smart image recognition with which a biomass determination and evaluation of the feed intake are to be made possible. By combining this with metadata on husbandry conditions, dynamic growth models will be created that take species- and husbandry-specific characteristics into account. With this database, reliable forecasts regarding feed intake, production quantity and product quality will be possible, with which the economic efficiency of the plants can be increased.
Initial situation
Aquacultures are becoming increasingly important for feeding the population, as the rising global demand for fish and marine products can no longer be met via wild catch without serious consequential costs for the environment.
In this context, recirculating aquaculture systems, as a location-independent solution in rural areas, offer a resource-conserving, transparent and decentralized alternative to the previous supply.
FishSizer
AI-based determination
of height and weight
FeedWatch
AI-based assessment
of the feed intake
The August-Wilhelm Scheer Institute develops the following components:
We develop the modules to size the fish and determine the feed intake in the water tanks. In addition, the underlying database, as well as the trained AI models, are integrated into a cloud solution developed by us. With this system, users can access all functionalities.
Partner
FUNDING NOTICE
The FishAI project is funded by the German Federal Ministry of Education and Research.
Funding code: 031B1252B
Duration: 01.08.2022 – 31.07.2025