Field of the invention

The present invention relates to a method for breeding/farming one or more aquatic animals in an aquatic farming plant while using a simulation model. Simulation of the breeding process will give possibilities to predict one or more parameters in the breeding process, such as expected quality of the end product can be obtained based on collected data from previous monitored breeding processes. This will enable design and production of aquatic animals in accordance with the end users expectations and/or specifications. Further will this invention enable delivery of specially adapted/tailormade aquatic products with specified target qualities to market expectations as well as designing productions systems.

Background of the invention

The aquaculture industry today produce a standard product that are sold on the market as a commodity, and “a fish is a fish” with very few parameters.

Aquaculture is a diverse industry, ranging from simple and small land ponds to big sea-based or land-based fish farms that can produce large quantities of fish.

There are quite few international, regional and local standards and certification schemes regulating how the fish should be treated, fed and delivered. However, we know no standard for the end product, e.g. a salmon fish fillet that is optimised for a specific market or products such as sushi or smoking, a target quality.

The aquaculture industry is faced with many issues that cause fish suffering and production problems. Such issues are commonly addressed by distinct measures that address the issue at hand and may cause other unintended issues/ side effects. In addition the aquaculture industry is faced with increasing cost, environmental issues, and complexity of operations. Most of these problems are due to insufficient knowledge, lack of data gathering, no ability to understand complicated casual relationships, no utilisation of large datasets in assembled simulators and production models.

In order to address complicated matters, it's often better to understand the fundamental problems and how these are caused rather than focusing on treating the symptoms.

There are simulation models on the market, which are used in the fish farming industry today, most of them are created for feeding and with an intention to obtain a faster growth rate of the fish, from smolt to slaughter size. None of them are related to the end product quality, and how to obtain a specific quality of the end product.

International patent publication WO2019/00288 A1 relates to a method and apparatus for collecting and/or pre- processing data related to feeding animals in water. More particularly, the invention relates to a method and apparatus for minimising wasted feed used in a fish farm. A provided a computer-implemented method for detecting motion in relation to aquatic animals, the method comprising the steps of: receiving sensor data; determining from the sensor data moving objects using learned functions; and generating output data in relation to the determined moving objects.

There is also a project that that introduces a method for simulation with a holistic digital representation, of the fish farm and fish farming operations that can serve farmers' needs to enhance available single measurements and other datapoints for providing insight into parameters relate to the structure, biomass, and environmental conditions (https://prosiektbanken.forskninqsradet.no/en/proiect/FORISS /3Q9524?Kilde=FORIS S&distribution=Ar&chart=bar&calcType=fundinq& ;Sprak=no&sortBy=date&sortOrder =desc&resultCount=30&offset=0&Faq.3=Marin+teknol oqi).

The use of the holistic digital representation concepts is focused and developed to ensure more suitable maintenance procedures and to create more efficient farming processes.

The present invention is generally concerned with solving at least one, but preferably several, of the problems which exist with the prior art. More particularly, it has been an object of the invention to develop a method that can assist users, like companies such as technology developers/ farmers/ producers, to design and manufacture aquatic products, like fish, shellfish or algae, supported by use of information from sources such as biological-Zenvironmental-/ technical data collection, sensor technology, environmental forecast, genetic information, measurement technology in a computing based system that by use of advanced processing, such as machine learning, analytics/ artificial Intelligence and simulations that can tailor make better product produced in a more efficient process with more streamlined production facilities. Example task: Design and produce a fish that have qualities that are better for producing sushi for the Chinese middle class market by selecting the best match genetic properties, grow the fish with feed and environmental factors managed to obtain a good right grow rate, a good trim factor, fat content/ distribution, a desired taste and other quality parameters important for the particular use/ market as well as control the slaughtering, processing and logistics to deliver a near optimal product at right time. Measure the product and process in all steps and use this data to further improve the next batch of products produced in the same facilities or similar facilities.

Brief summary of the invention

The invention is defined by the independent claims 1 and 17.

Further, inventive embodiments of the invention are set out in the dependent claims.

Figure captions

The attached figures illustrate some embodiments of the claimed invention.

Figure 1 shows an overview the method steps from start to end.

Figure 2 shows an illustration of step a1 )-a3) of the method.

Figure 3 shows an illustration of step b1 ) of the method.

Figure 4 shows an illustration of step b1 ) through c1 )-c3) of the method.

Figure 5 shows an illustration of step from c) through d1 )-d4) of the method. Figure 6 shows an illustration of step from d4) through e1 )-e3) of the method. Figure 7 shows an illustration of step from e3) through f1 )-f2)-g 1 ) of the method.

Figure 8 shows an illustration of some environmental parameters for an aquatic animal.

Figure 9 shows the process of the invention, claim 1 , first part.

Figure 10 shows the process of the invention, claim 1 , continued.

Figure 11 shows an illustration of Liebig's law of minimum.

Figure 12 shows an illustration of Shelford's law of tolerance.

Figure 13 shows an illustration of example of Shelford's law of tolerance.

Figure 14 shows an illustration of data collection.

Figure 15 shows an illustration of Oceanfront Aquaculture System (OAS).

Figure 16 shows an illustration of "from genes to fork".

Figure 17 shows an illustration of the method process .

Figure 18 shows an illustration of a data collection structure. Embodiments of the present invention will now be described, by way of example only, with reference to the above mentioned Figures.

Embodiments of the invention

The present invention provides a method for breeding/farming one or more aquatic animals by utilizing a computer-implemented simulation model (10SM), wherein said computer-based simulation model (10SM) can be interpreted as a computer-based process of creating and analyzing a digital prototype of a physical model to predict its performance in the real world, as presented in independent claim 1.

We have interchangeably used terms breeding and farming.

The present invention provides an apparatus for one or more aquatic animals by utilizing a computer-implemented simulation model (10SM), as presented in independent claim 17.

In an embodiment of the invention, the first steps, a1-a3, are setting the assumptions by defining the system that simulation model (10SM) is intended to describe. This means that the relevant parameters, as factors and variables, have to be chosen and their interaction has to be built into the structure of the simulation model. See especially Figure 2. a1) Establishing a model quality [QM] with a set of selectable model quality parameters [QSMP].

The model quality [QM] is representing real world qualities in the simulation model (10SM), and the basis for the model quality [QM] is the set of selectable model quality parameters [QSMP], and the parameters are defined as variables and factors.

The selectable quality model parameters [QSMP] may be considered as a matrix QSMP[ ] reserved in the algorithm, wherein all significant qualities of the aquatic animal are listed up. The matrix QSMP[ ] is multidimensional, with some linear parameters such as weight, length, width, and with some parameters subjective and selectable among discrete text values such as taste, smell, and some discrete such as boilable Y/N, sushi Y/N, aquatic Y/N, freezable Y/N, and some which may be discrete or linear such as texture.

Said selectable quality model parameters [QSMP] are configured for updates and tuning. a2) Establishing a model aquatic population (200M) with a set of selectable model aquatic population parameters (210SMP).

The model aquatic population (200M) is representing real world aquatic population in in the simulation model (10SM), and the basis for the model aquatic population (200M) is the set of selectable model aquatic population parameters (210SMP), and the parameters are defined as variables and factors.

The set of selectable model aquatic population parameters (21 0SMP) may be considered as a matrix 210SMP[ ] reserved in the algorithm, wherein all significant parameters of the aquatic population are listed up. The matrix 21 0SMP [ ] is multidimensional, with some linear parameters such as weight, length, width, and with some parameters subjective and selectable among discrete text values such as taste, smell, and some discrete such as boilable Y/N, sushi Y/N, aquatic Y/N, freezable Y/N, and some which may be discrete or linear such as texture.

Said set of selectable model aquatic population parameters (21 0SMP) are configured for updates and tuning. a3) Establishing a model aquatic farming plant (300M) with a set of selectable model aquatic farming plant parameters (310SMP).

The model aquatic farming plant (300M) is representing real world aquatic farming plant in the simulation model (10SM), and the basis for the model aquatic farming plant (300M) is the set of selectable model aquatic farming plant parameters (31 0SMP).

The set of selectable model aquatic farming plant parameters (31 0SMP) may be considered as a matrix 310SMP[ ] reserved in the algorithm, wherein all significant parameters of the aquatic farming plant are listed up. The matrix 31 0SMP [ ] is multidimensional, with some linear parameters such as salinity, flow, turbulence, and oxygen content some parameters selectable among discrete text values such as depth, length and location, and some discrete such as fish plant Y/N, enclosed farming plant Y/N and some which may be discrete or linear such as salt water.

Said set of selectable model aquatic farming plant parameters (31 0SMP) are configured for updates and tuning. b1) Determining a target quality [QT] by specifying target quality parameters [QTP].

The target quality [QT] is representing the real end users preferences given by e.g. an end user of the product, a designer of aquatic animals farming plant, a chef, a fish farmer, a fish exporter etc, by specifying target quality parameters [QTP].

The real target quality parameters [QRP] must be within the of selectable model quality parameters [QSMP] and its limited values.

If not within the selectable model quality parameters [QSMP] and its limited values, then either update the set of selectable model quality parameters [QSMP] and/or its limited values, or the simulation model (10SM) should indicate that the of selectable model quality parameters [QSMP] is out of range or similar.

Figure 3 illustrates method step b1 ), and that the target quality [QT] has to correspond with the predefined model quality [QM] iwo that the target quality parameters [QTP] has to correspond within the limited values given from the selectable model quality parameters [QSMP]. c1 ) Running said simulation model (10SM) with said target quality parameters [QTP] on said model quality [QM] to specify quality model parameters [QMP].

The target qualities parameters [QTP] redefined in said simulation model (10SM) to said specified quality model parameter [QMP]. c2) Running said simulation model (10SM) using said model quality parameters [QMP] on said model aquatic population (200M) to specify a set of model aquatic population parameters (210MP) in order to reasonably approach said target quality [QT].

The simulation model (10SM) will detect and specify the model aquatic parameters (210MP) which are highest graded for breeding the aquatic population into the finished product as defined by the target quality [QT]. c3) Running said simulation model (10SM) using said model quality parameters [QMP] and said set of model aquatic population parameters (210MP) on said model aquatic farming plant (300M) to specify a set of model aquatic farming plant parameters (310MP) in order to reasonably approach said target quality [QT].

The simulation model (10SM) will detect and specify the set of model aquatic farming plant parameters (310MP) which are highest graded for breeding the aquatic population into the finished product as defined by the real target quality [QR],

Figure 4 further illustrates the steps from b1 through step c1)-c2)-c3). The Figure illustrates that the target quality [QT] defines the model quality parameters [QMP], which is further used in the simulation (10SM) to select the model aquatic population parameters (210MP). Both the model quality parameters [QMP] and the set of model aquatic population parameters (210MP) are then used in the next simulation (10SM) to define the set of model aquatic farming plant parameters (310MP). d1) Establishing a real aquatic population (20R) with a set of real aquatic population parameters (20RP) based on said set of selected model aquatic population parameters (210MP) from c1 ).

The real aquatic population (20R) will be selected to reasonably approach to the model aquatic population (200M) with the set of selected model aquatic population parameters (210MP) and can come from original genes (clean species) or modified genes in any form as broodstock, breeding/genomics, genotype and DNA. d2) Establishing a real aquatic farming plant (30R) with real aquatic farming plant parameters (31 RP) and real aquatic farming plant parameter sensors (32RPS) based on said set of specified model aquatic farming plant parameters (310MP) from c2).

The real aquatic farming plant (30R) can either chosen from a stock of already established/built farming plants, or give design parameters to design and establish a new real aquatic farming plant (30R). In order to get feedback, especially, from the breeding process it is beneficial to establish real aquatic farming plant parameters sensors (32RPS). These real aquatic farming plant parameters sensors (32RPS) may be different from the real aquatic farming plant parameters (31 RP), and the main purpose with the sensors (32RPS) is to give feedback from the process, but they may also have other purposes as directly controls processes such as feeding, illumination, turbulence, waves, current, temperature, salinity, oxygen level in water etc.

Figure 8 illustrates a selection of environmental parameters for which a fish subject to and they are all integrated, partly or completely, with each other. d3) Establishing said real aquatic population (20R) in said real aquatic farming plant (30R) for breeding said real aquatic population (20R). d4) Allowing said real aquatic population (20R) to grow or grow up in said real aquatic farming plant (30R) subject to measuring and monitoring said real aquatic farming plant parameters (31 RP) and said set of real aquatic population parameters (21 RP) while logging real aquatic farming plant parameter data (33RPD,) and real aquatic population parameter data (23RPD) automatically by means of said real aquatic farming plant parameter sensors (30 RPS) and/or manually.

While the real aquatic population (20R) is growing up a valuable part is to perform logging, and by logging means record and store empirical data for analysis and to improve the simulation model (10SM). The logging may appear directly from the real aquatic farming plant parameters sensors (32RPS) or by manually gathering from the process or from a specimen from the real aquatic population (20R), either by none- destructive testing on an aquatic animal or destructive testing on an aquatic animal.

Figure 5 shows where a regular aquatic breeding process starts, establishing some aquatic animals in an aquatic farming plant. The Figure illustrates the steps from d1 ) to d-4) as described above. e1) Slaughtering all or part of said real aquatic population (20R) and measuring real quality obtained [Qo] with a set of real quality obtained parameters [QOP] of said slaughtered real aquatic population (20R).

The real quality obtained [Qo] is the real world qualities and may have real quality obtained parameters [QOP] that can be measured directly at the real aquatic farming plant (30R), typical objective parameters that can be measured by some equipment such as a scale and meter, but also subjective parameters such as observation by visual or smell can be measured. Other parameters needs to be measured after some treatment such as freezing, cooking, sushi etc. at the end customer, and the end customer can be all from a chef to an aquatic animal trader in the international market to a marked analysis.

Further the real obtained quality parameters [QOP] may be considered as a matrix QOP[ ] reserved in the algorithm, wherein all significant qualities of the aquatic animal are listed up. The matrix QOP[ ] is multidimensional, with some linear parameters such as weight, length, width, and with some parameters subjective and selectable among discrete text values such as taste, smell, and some discrete such as boilable Y/N, sushi Y/N, aquatic Y/N, freezable Y/N, and some which may be discrete or linear such as texture. e2) Determining a quality gap [QG] with a set of quality gap parameters [QGP] between said real quality obtained [Qo] and said target quality [QT].

Identify the differences between the real quality obtained parameters matrix QOP[ ] and target quality parameters matrix QTP[ ] to settle the set of quality gap parameters QGPP[ ], in order to determining the quality gap [QG]. e3) Empirical analyzing said automatically and/or manually real aquatic farming plant parameter data (33RPD,) and real aquatic population parameter data (23RPD) and said set of quality gap parameters [QGP],

Empirical analysis can be identified in this subject as an evidence-based approach to the study and interpretation of information from real world experiences. The empirical approach relies on real-world data, metrics and results rather than theories and concepts.

All the logged data from step d4) needs to be processed in order to become useful for the simulation model (10SM). Empirical analyzing of the results from the logged data will contribute to identify the bias in the observation of the relevant parameters, factors and variables, and better understand how their interaction has to be built into the structure of the simulation model (10SM).

Figure 6 illustrates the steps from e1 ) to e3). f 1 ) Empirical calibrating of said set of selectable model aquatic population parameters (21 0SMP) and running said simulation model (10SM) on said model aquatic population (200M) in order to reasonably approach said set of obtained quality parameters [QOP]. f2) Empirical calibration of said set of selectable model aquatic farming plant parameters (31 0SMP) and running said simulation model (10SM) with said set of empirical calibrated model aquatic population parameters (21 0CMP) on said model aquatic farming plant (300M) in order to reasonably approach said set of real quality obtained parameters [QOP].

Empirical calibration deals with bias in observational research estimates, which is the bias that remains after any measures taken to adjust for bias and confounding inherent in observational data. Eg. calibrate and tune the model and model parameters, factors and variables, to achieve more unified results while simulating with the real world results. g1 ) Repeat from step b1 ).

This can be repeated until said quality gap [QG] with said quality gap parameters [QGP] is reduced within an reasonably acceptable limited value.

The method for breeding one or more aquatic animals by means from a computer- implemented simulation model (10SM), as described above, can be execute in full scale farming plants or in smaller test farming plants, anyhow these steps can be done repeatedly as described in step g1). The intention to execute these steps repeatedly is to gather useful information that can bring the simulation model (10SM) forward, such that the simulation model (10SM) can contribute to better end products in the real aquatic animal production and to design aquatic farming plants to reasonably achieve better production processes of the product.

Figure 7 illustrates step f1 ) to g1 ).

Figure 1 , 9 and 10 shows an illustrates of the process of method as described in claim 1 , through the steps from a1 ) to g1 ). In an embodiment of the invention wherein said computer-implemented simulation model (10SM) is based on Liebig's Law of the Minimum.

This means that the relevant parameters, as factors and variables, have to be chosen and their interaction has to be built into the structure of the simulation model (10SM), based on at least the Liebig's Law of the Minimum theoretical consideration and common knowledge.

Liebig's law states that growth is dictated not by total resources available, but by the scarcest resource, limiting factor. This law is not directly sufficient to understand growth in aquaculture as it is more complex relationship between conditions and effects which also have interactions directly and in higher orders. E.g. growth for a fish is depending of available feed, however the intake and conversion of feed to fish meat is depending of water quality, e.g. temperature, oxygen, CO2, genes, competition, speed of water/ fish swimming activity, light conditions and a lot more factors. See especially Figure 11 .

In an embodiment of the invention wherein said computer-implemented simulation model (10SM) is based on Shelford's law of tolerance.

This means that the relevant parameters, as factors and variables, have to be chosen and their interaction has to be built into the structure of the simulation model (10SM), based on at least the Shelford's law of tolerance theoretical consideration and common knowledge.

Shelford's law of tolerance states that an organism's success is based on a complex set of conditions and that each organism has a certain minimum, maximum, and optimum environmental factor or combination of factors that determine success.

For production of aquatic product, the relationship between conditions and effects are both direct and coupled in many dimensions. E.g. the optimal water temperature is not in one dimension but a function of fish size, gas content, salinity, feed rate, density of fish and other factors as well as interaction of these factors. See especially on Figures 12 - 13.

In an further embodiment of the invention wherein said computer-implemented simulation model (10) are based on a combination of Liebig's Law of the Minimum and Shelford's law of tolerance. This means that the relevant parameters, as factors and variables, have to be chosen and their interaction has to be built into the structure of the simulation model (10SM), based on at least the Liebig's Law of the Minimum and the Shelford's law of tolerance theoretical consideration and common knowledge.

In an embodiment of the invention wherein said empirical analysis and calibration are based on machine learning [ML] from said logged data and said set of quality gap parameters [QGP].

In this subject we can define machine learning (ML) as a type of artificial intelligence (Al) that allows software applications to become more accurate at predicting outcomes without being explicitly programmed to do so. Machine learning algorithms use historical data as input to predict new output values.

In an embodiment of the invention wherein said selectable model quality parameters [QSMP], is a multidimensional matrix QSMP[ ] of variables and factors.

In an embodiment of the invention, wherein said target quality parameters [QTP] is a multidimensional matrix QTP[ ] of variables and factors.

In an embodiment of the invention, wherein said real quality obtained parameters [QOP] is a multidimensional matrix QOP[ ] of variables and factors.

In an embodiment of the invention, wherein said target gap parameters [QGP] is a multidimensional matrix QGP[ ] of variables and factors.

In an embodiment of the invention, wherein selectable model aquatic population parameters (210SMP), is a multidimensional matrix 21 0SMP[ ] of variables and factors.

In an embodiment of the invention, wherein model aquatic population parameters (21 0MP), is a multidimensional matrix 21 0MP[ ] of variables and factors. In an embodiment of the invention, wherein selectable model aquatic farming plant parameters (310SMP), is a multidimensional matrix 310SMP[ ] of variables and factors.

In an embodiment of the invention, wherein model aquatic farming plant parameters (310MP), is a multidimensional matrix 310MP[ ] of variables and factors.

In an embodiment of the invention, wherein real aquatic population parameters (21 RP), is a multidimensional matrix 21 RP[ ] of variables and factors.

In an embodiment of the invention, wherein real aquatic farming plant parameters (31 RP), is a multidimensional matrix 31 RP[ ] of variables and factors.

Figure 14 illustrates an embodiment of the invention wherein inputs in a fish farming plant are typically fish, water, 02, Energy, Heat, Feed, light and more, and that the internal system for data catch are Fish, Water, Energy, Heat, Feed, Process Parameters, Velocity, Pressure, Resistance, O2, CO2, N and more, and that the outputs are Fish, Organic matter, Phosphorus, Nitrogem solid, Protein, Fat, Water, O2, N, CO2, NH3, NH4, Energy, Heat Feed and more.

Figure 15 illustrates a simplified overview of the concept named Oceanfront Aquaculture System (OAS), with integration between some of main factors, that shall be considered and processed in the computer-based simulation model for breeding aquatic animals.

Figure 16 illustrates the OAS using the phrase «From genes to fork» by designing aquatic animals according to customers expectations.

Figure 17 illustrates an embodiment of invention with a simplified method with fewer steps and with examples on how processes are connected and different tools for modelling, update/calibrate the simulator, collect data from the process.

Figure 18 illustrates an example of a parameter tree-structure for a simulation model (10SM). The invention provides a fish farming method for farming by using a fish population (20R) with a set of fish population parameters (21 RP) comprising;

- weight and length, and a fish farming plant (30R) with a set of farming plant parameters (31 RP) comprising;

- salinity, water temperature, flow velocity, feeding and energy consumption, and fish farming plant parameter sensors (32RPS) comprising;

- salinity meter, temperature meter, flow velocity meter, feed meter and energy meter, weight meter and length meter, and by utilizing a computer-implemented algorithm with a fish farm simulation model (10SM), comprising the steps of; a1 ) establishing a model quality (QM) for fish farming with a set of selectable model quality parameters (QSMP) for fish farming comprising;

- model energy consumption, model feeding, model weight and model length, a2) establishing a model fish population (200M) with a set of selectable model fish population parameters (210SMP) comprising;

- model weight and model length, a3) establishing a model farming plant (300M) with a set of selectable model farming plant parameters (310SMP) comprising;

- model salinity, model water temperature, model flow velocity, model feeding and model energy consumption, b1) determining a target quality (QT) by specifying target quality parameters (QTP) comprising;

- target energy consumption, target feeding, target weight and target length, c) running said fish farm simulation model (10SM); c1 ) with said target quality parameters (QTP) on said model quality (QM) to specify a set of quality model parameters (QMP) comprising;

- model energy consumption, model fodder consumption, model weight and model length, c2) use said set of specified model quality parameters [QMP] on said model fish population (200M) to specify a set of model fish population parameters (210MP) comprising;

- model weight and model length, in order to approach said target quality [QT], c3) use said set of specified model quality parameters [QMP] and said set of model fish population parameters (210MP) on said model fish farming plant (300M) to specify a set of model fish farming plant parameters (310MP) comprising; - model salinity, model water temperature, model flow velocity and model energy consumption, in order to approach said target quality [QT], d1) establish said fish population (20R) with said set of fish population parameters (21 RP) based on said set of specified model fish population parameters (210MP) from c2), d2) establish said fish farming plant (30R) with said set of fish farming plant parameters (31 RP) based on said set of specified model fish farming plant parameters (310MP) from c3), d3) establish said fish population (20R) in said fish farming plant (30R) for farming said fish population (20R), d4) grow said fish population (20R) in said fish farming plant (30R) subject to measuring and monitoring and logging said set of fish farming plant parameters (31 RP) and said set of fish population parameters (21 RP), by means of said fish farming plant parameter sensors (32RPS), e1) slaughtering all or part of said fish population (20R) and measuring obtained quality (QO) with a set of obtained quality parameters (QOP) comprising;

- obtained energy consumption, obtained fodder consumption, obtained weight and obtained length, e2) determining a quality gap (QG) with a set of quality gap parameters (QGP) comprising;

- gap energy consumption, gap fodder consumption, gap weight and gap length, between said obtained quality (QO) and said target quality (QT), e3) empirically analyze said logged fish farming plant parameters (31 RP) and said logged fish population parameters (21 RP) and said set of fish quality gap parameters (QGP), f) empirically adjust said fish farming simulation model (10SM), said model fish population (200M) and said model farming plant (300M) in order to adjust said set of farming plant parameters (31 RP) and said set of fish population parameters (21 RP) to reduce said quality gap (QG) in a subsequent farming of an accordingly adjusted population in an adjusted farming plant.

In an embodiment of the invention, the above steps are repeated until said obtained quality (QG) is sufficiently near said quality target (QT) or said quality gap (QG) is reduced to an reasonably acceptable limit. In an embodiment of the invention, said computer-implemented simulation model (10), as used in step c), is based on Liebig's Law of the Minimum in order to dictate not by total resources available, but by a limiting scarcest resource.

In an embodiment of the invention, said computer-implemented simulation model (10), as used in step c), is based on Shelford's law of tolerance in order to state that an organism's success based on a complex set of conditions and that each organism has a certain minimum, maximum, and optimum environmental factor or combination of factors that determine success or not.

In an embodiment of the invention, said computer-implemented simulation model (10), as used in step c), is based on a combination of Liebig's Law of the Minimum and Shelford's law of tolerance, in order to combine selections based on a limiting scarcest resource, and a complex set of conditions and that each organism has a certain minimum, maximum, and optimum environmental factor or combination of factors that determine success or not.

In an embodiment of the invention, said empirical analysis, in e3), and empirical adjusting, in f), is based on machine learning (ML) from said monitored and logged data and said set of fish quality gap parameters (QGP).

In an embodiment of the invention, said fish selectable model quality parameters (QSMP), is a multidimensional matrix QSMP[ ] further comprising model taste, model color, model smell, model texture, model art/species, model boilable, model sushi, model aquatic, model energy effective to produce.

In an embodiment of the invention, said quality target parameters (QTP) is a multidimensional matrix QTP[ ] further comprising target taste, target color, target smell, target texture, target art/species, target boilable, target sushi, target aquatic, target energy effective to produce.

In an embodiment of the invention, said quality obtained parameters (QOP) is a multidimensional matrix QOP[ ] further comprising obtained taste, obtained color, obtained smell, obtained texture, obtained art/species, obtained boilable, obtained sushi, obtained aquatic, obtained energy effective to produce. In an embodiment of the invention, said target gap parameters (QGP) is a multidimensional matrix QGP[ ] further comprising gap taste, gap color, gap smell, gap texture, gap art/species, gap boilable, gap sushi, gap aquatic, gap energy effective to produce.

In an embodiment of the invention, said selectable fish model population parameters (210SMP), is a multidimensional matrix 210SMP[ ] further comprising model taste, model color, model smell, model texture, model art/species, model boilable, model sushi, model aquatic, model energy effective to produce.

In an embodiment of the invention, said fish model population parameters (210MP), is a multidimensional matrix 210MP[ ] comprising model taste, model color, model smell, model texture, model art/species, model boilable, model sushi, model aquatic, model energy effective to produce.

In an embodiment of the invention, said wherein selectable model farming plant parameters (310SMP), is a multidimensional matrix 310SMP[ ] comprising model gas content, model rain, model illumination, model turbidity, model location, model temperature, model fresh or salt water, model oxygen, model particles, model wave, model pump power.

In an embodiment of the invention, said wherein model farming plant parameters (310MP), is a multidimensional matrix 310MP[ ] comprising model gas content, model rain, model illumination, model turbidity, model location, model temperature, model fresh or salt water, model oxygen, model particles, model wave, model pump power.

In an embodiment of the invention, said fish population parameters (21 RP), is a multidimensional matrix 21 RP[ ] further comprising taste, color, smell, texture, art/species, boilable, sushi, aquatic, energy effective to produce.

In an embodiment of the invention, said farming plant parameters (31 RP), is a multidimensional matrix 31 RP[ ] comprising gas content, rain, illumination, turbidity, location, temperature, fresh or salt water, oxygen, particles, wave, pump power.

The invention provides a fish farming plant (30R), wherein - said fish farming plant (30R) comprises a set of farming plant parameters (31 RP) and further comprises fish farming plant parameter sensors (32RPS),

- a fish population (20R) having a set of fish population parameters (21 RP), and

- a computer with a computer-implemented algorithm with a fish farm simulation model (10SM), wherein said fish farm simulation model (10SM) comprises;

- a model farming plant (300M) with a set of selectable model farming plant parameters (310SMP) for simulating said fish farming plant (30R) and said set of farming plant parameters (31 RP), and

- a model fish population (200M) with a set of selectable model fish population parameters (210SMP) for simulating said fish population (20R) and said set of fish population parameters (21 RP).

In an embodiment of the invention, said fish farm simulation model (10SM) is arranged for, based on a model quality [QM] for fish farming with a set of selectable model quality parameters [QSMP] established from a target quality [QT] with a set of target quality parameters [QTP] specified as input,

- selecting said set of selectable model farming plant parameters (310SMP) for said model farming plant (300M) for providing a set of specified model fish farming plant parameters (310MP), and

- selecting said set of selectable model fish population parameters (210SMP) for said model fish population (200M) for providing a set of specified model fish population parameters (210MP).

In an embodiment of the invention, said fish population (20R) is established with

- said set of fish population parameters (21 RP) based on said set of specified model fish population parameters (210MP).

In an embodiment of the invention, said fish farming plant (30R) is established with

- said set of fish farming plant parameters (31 RP) and said fish farming plant parameter sensors (32RPS) based on said set of specified model fish farming plant parameters (310MP).

In an embodiment of the invention, said fish farming plant (30R) is arranged to receive and grow said fish population (20R) while subject to measuring and monitoring said set of farming plant parameters (31 RP) and said set of of fish population parameters (21 RP). In an embodiment of the invention, said fish farming plant parameter sensors (32RPS) are further arranged for logging of said farming plant parameters (31 RP) and also arranged for logging said fish population parameter data (23RPD).

In an embodiment of the invention, said fish farming plant (30R) is arranged for slaughtering all or part of said fish population (20R) and measuring an obtained quality (QO) with a set of obtained quality parameters (QOP), wherein said computer-implemented algorithm in said computer is arranged for determining a quality gap (QG) with a set of quality gap parameters (QGP) between said obtained quality (QO) and said target quality (QT).

In an embodiment of the invention, said algorithm is further arranged to empirically analyze fish farming plant parameter data (33RPD,) and said fish population parameter data (23RPD) and said set of fish quality gap parameters (QGP) to adjust said fish farm simulation model (10SM) and said model fish population (200M) and said model farming plant (300M) in order to adjust said set of farming plant parameters (31 RP) and said set of fish population parameters (21 RP) to reduce said quality gap [QG] in a subsequent farming of adjusted fish population in an adjusted farming plant.

In an embodiment of the invention, said fish farming plant (30R) is arranged for farming said fish population (20R) having said set of fish population parameters (21 RP) comprising;

- weight and length, said fish farming plant (30R) having said set of farming plant parameters (31 RP) comprising;

- salinity, water temperature, flow velocity, feeding rate and energy consumption, and said fish farming plant parameter sensors (32RPS) comprising;

- salinity meter, temperature meter, flow velocity meter, feed meter, and energy consumption meter, said fish farming plant (30R) further comprising;

- said computer with said computer-implemented fish farm simulation model (10SM) comprising; said model quality [QM] with said set of selectable model quality parameters [QSMP] for said model fish population (200M) and model farming plant (300M), comprising; - model energy consumption, model fodder consumption, model weight and model length, said model fish population (200M) with said set of selectable model fish population parameters (210SMP) comprising;

- model weight and model length, said model farming plant (300M) with said set of selectable model farming plant parameters (310SMP) comprising;

- model salinity, model water temperature, model flow velocity, model feeding, and model energy consumption, said computer-implemented fish farm simulation model (10SM) arranged for receiving said target quality [QT] with said specified target quality parameters [QTP] comprising;

- target energy consumption, target feeding, target weight and target length, and further arranged for;

- specifying, based on said target quality parameters [QTP] as input to said model quality [QM], said set of specified quality model parameters [QMP] comprising;

- model energy consumption, model feeding, model weight and model length,

- and further arranged to use said set of specified model quality parameters [QMP] on said model fish population (200M) to specify said set of model fish population parameters (210MP) comprising;

- model weight, and model length, in order to approach said target quality [QT],

- specifying, based on said set of specified model quality parameters [QMP] and said set of model fish population parameters (210MP) on said model fish farming plant (300M), said set of model fish farming plant parameters (310MP) comprising;

- said model salinity, said model water temperature, said model flow velocity and said model energy consumption, in order to approach said target quality [QT],

- said fish population (20R) established with said set of fish population parameters (21 RP) based on said set of specified model fish population parameters (210MP),

- said fish farming plant (30R) established with said set of fish farming plant parameters (31 RP) based on said set of specified model fish farming plant parameters (310MP),

- said fish farming plant (30R) arranged to receive said fish population (20R),

- said fish farming plant (30R) arranged for growing said fish population (20R) while subject to measuring and monitoring said set of farming plant parameters (31 RP) and said set of fish population parameters (21 RP) while said fish farming plant parameter sensors (32RPS) are arranged for logging of said farming plant parameters (31 RP) comprising;

- salinity, water temperature, flow velocity and energy consumption, and also arranged for logging fish population parameters (21 RP) comprising;

- weight and length, and

- said fish farming plant (30R) further arranged for slaughtering all or part of said fish population (20R) and measuring an obtained quality [QO] with a set of obtained quality parameters [QOP] comprising;

- obtained energy consumption, obtained feeding, obtained weight and obtained length, and

- said algorithm in said computer is further arranged for determining a quality gap [QG] with a set of quality gap parameters [QGP] comprising;

- gap energy consumption, gap fodder consumption, gap weight and length, between said obtained quality [QO] and said target quality [QT], and

- said algorithm is also arranged to empirically analyze logged fish farming plant parameters (31 RP,) and said logged fish population parameters (21 RP) and said set of fish quality gap parameters [QGP], and

- said algorithm arranged to empirically adjust said fish farm simulation model (10SM), said model fish population (200M) and said model farming plant (300M) in order to adjust said set of farming plant parameters (31 RP) and said set of fish population parameters (21 RP) to reduce said quality gap [QG] in a subsequent farming of an accordingly adjusted population in an adjusted farming plant.