Petrozavodsk State University unveils an AI-driven fish breeding system for real-time monitoring
Researchers at Petrozavodsk State University have introduced a clever fish breeding platform that tracks water quality, fish size, and disease signals using artificial intelligence. This development was shared with socialbites.ca through the Ministry of Education and Science.
The software and hardware suite is built from four interconnected modules. Together, they gather data about the fish habitat and overall health, then compute the best possible growing conditions from that information. The result is a smarter, more efficient aquaculture operation that can respond to changing circumstances on the fly.
According to Alexey Marakhtanov, director of the Center for Artificial Intelligence at PetrSU, the system’s deployment in a fish farm enhances product quality and competitiveness. Continuous environmental monitoring, combined with a non-contact method for estimating fish weight, contributes to higher purity and more reliable stock. This approach also supports production growth by reducing fish mortality and process waste, while sharpening profitability through optimized workflows across the enterprise.
The platform continuously collects data on water quality, surrounding conditions, fish health, and routine maintenance. It also provides comprehensive video surveillance, including underwater footage, to capture a complete picture of the farming environment.
Within the analytics module, the captured data is processed to determine the optimal feeding schedules and environmental settings necessary to achieve targeted gains in biomass. A video analysis component identifies individual fish in footage, estimates their height and weight, and flags potential behavioral anomalies or signs of disease. This combination of data and video analytics enables proactive management of stock health and growth trends.
The system offers a Control module that enables remote operation of equipment such as feeders, motors, and lighting. A separate Integration module ensures seamless interaction between the platform and external information systems used by the enterprise. This connectivity supports centralized decision making and easier data sharing across teams.
Practically, the platform’s output can be presented to operators and technologists as actionable recommendations for specific tasks. In higher automation scenarios, the system can trigger automatic actions by activating on-site actuators like filters, pumps, feeders, and lighting to maintain optimal conditions without human intervention.
From a broader perspective, the integration of AI with aquaculture operations reflects a growing trend toward data-driven farming methods. The ability to monitor multiple parameters in real time and to adjust conditions automatically can lead to more consistent product quality, reduced resource use, and better overall efficiency in fish farming settings across North America.
In a landscape where precision agriculture is expanding, this Petrozavodsk initiative demonstrates how combining sensing, video analytics, and autonomous control can create more resilient and productive aquaculture systems that meet demanding market standards.
Note: The content above describes a research-driven aquaculture platform and its potential implications for modern fish farming practices.