Russia (bbabo.net), - In 2021, scientists from the Far East conducted an experiment using artificial intelligence (AI) to account for the approaches of salmon. I learned about what is expected from digital technologies in industry science from the director of the Pacific branch of VNIRO Alexey Baitalyuk.

Aleksey Anatolyevich, what is the peculiarity of the use of new generation drones in aerial surveys, which science has been conducting for decades?

Aleksey Baytalyuk: With traditional methods of aerial census, it takes quite a long time after the completion of the flight to calculate the number of salmon approaches and the degree of filling of spawning grounds. Therefore, the main scientific and applied interest for us is the acceleration of the process of obtaining the final grade. Unmanned aerial vehicles (UAVs) with great autonomy can provide such an opportunity. They are able to collect information about the number of herds and, most importantly, to recognize the species composition. Now artificial intelligence is "learning", and when it "learns", the whole process - from fixation to inclusion in calculations - will ideally take only a few hours. That is, the members of the commission on anadromous fish species will have the most up-to-date data, and in the midst of a fishing season, it is difficult to overestimate the importance of up-to-date information.

Another task is to simplify the methods of aerial photography. Usually, observations from helicopters and airplanes are carried out by people. Of course, an experienced specialist quite easily determines the number and types of salmon. But, if you use UAVs, you can significantly reduce the cost of research and cover a much larger area. At the same time, the quality of the resulting material will improve.

Last summer, we performed a series of experimental drone flights in Chukotka, and the first results were encouraging. It is important that there are no subjective factors with this assessment method, since the AI ​​will do the calculations automatically, according to the proven software algorithms.

A new information processing scheme has also shown itself. Khabarovsk colleagues carried out the first work on GIS-systems, which make it possible to quickly transfer data on the accounting of Pacific salmon. In Putin 2021, part of the decisions of the regional commission on anadromous species were already made on the basis of this technology.

Of course, there are nuances, since each river has its own geomorphology, water and coastal vegetation. And it is impossible to apply this system all over the Far East overnight, we still cannot do without traditional methods.

Where else can smart information systems be used?

Aleksey Baytalyuk: We live in an era of global changes: the temperature of the World Ocean is rising, in the waters of the Sea of ​​Japan we are also observing such a process both in the surface and in the bottom layers. Under certain conditions, this can lead to large-scale blooms of microalgae, including toxic ones. Such phenomena provoke high mortality of hydrobionts, mariculture enterprises lose crops and incur losses.

Now artificial intelligence is "learning", and when it "learns", the traditionally long process of aerial survey of fish stocks will take only a few hours

The problem has long been in the focus of attention of branch science. We are aware of the need to expand monitoring of the habitat of aquatic biological resources, but it is not rational to send employees to take samples for six to nine months at each significant fish breeding area.

Here, devices for automatic collection of information come to the rescue - temperature, salinity, concentration of suspended particles (this is an indicator of the development of microalgae). In this way, the suitability of water for the survival and development of mariculture objects, the safety of the latter for human consumption is assessed.

In the near future, we plan to start creating such a system, consisting of several base buoy stations. They will be located in Primorye near Popov Island, in Preobrazhenie and De Livron bays.

The places were not chosen by chance, they differ in hydrological characteristics, there is a mass vegetation of microalgae at different times and various objects are cultivated. Therefore, these three test sites can be used to monitor the state and quality of the environment for larger areas.

What will this mean for mariculture farms in practice?

bbabo.Net