METHOD FOR GENERING A SUITABLE TREATMENT PLAN FOR DEVELOPING A CULTIVATED AREA WITHOUT CHEMICAL PESTICIDES, LEARNING MODEL AND ASSOCIATED SYSTEM

Abstract

The invention relates to a method for generating a treatment plan adapted to the development conditions of a cultivated area, comprising a first natural formulation adapted to promote growth and stimulate the defense system of a plant variety, the method comprising the following steps: loading (100) parameter values ​​of the cultivated area, loading (200) at least one value of an environmental parameter, for a predetermined past time period, of the cultivated area, generating (500) a recommendation (R1, R2) for treating the cultivated area as a function of the value of the environmental parameter and the values ​​of the parameters of the cultivated area, said recommendation (R1, R2) comprising an application of at least one other natural formulation selected from: a second natural formulation adapted to stimulate the defense system of the plant variety,and / or a third, natural formulation adapted to prevent the development of a pest. Figure to be published with the abbreviation: Figure 1.

Description

Title of the invention: METHOD FOR GENERATING A TREATMENT PLAN ADAPTED FOR THE DEVELOPMENT OF A CULTIVATED SURFACE WITHOUT PESTICIDES OF CHEMICAL ORIGIN, ASSOCIATED LEARNING MODEL AND SYSTEM Technical field

[0001] The invention relates to the field of agriculture, more particularly to the use of data for optimizing the development of a cultivated area. In particular, the invention relates to a method for generating a treatment plan suitable for the development of a cultivated area.

[0002] The invention further relates to a system for monitoring a cultivated area and generating a treatment plan suitable for the development of the cultivated area.

[0003] The invention also relates to a method for supervised training of a deep learning model for generating a treatment plan suitable for the development of a cultivated area. Prior art

[0004] Below we describe the known prior art from which the invention was developed.

[0005] In the field of agriculture, to enable effective control of pests and diseases, it is essential to treat crops to ensure optimal development and yield. The development of agriculture in the 20th century was made possible by the advent of agrochemistry. Indeed, the production and massive use of chemical fertilizers and pesticides and the development of intensive agriculture made it possible to meet one of the major challenges of the 20th century: feeding an ever-increasing world population.The marketing of fertilizers intended to cover the needs of plants associated with wide ranges of very powerful chemical pesticides and having durations of action of several weeks to several months have made it possible to significantly increase the production yields of cultivated areas by boosting the growth of cultivated areas, by allowing the eradication of insects or plants considered harmful or by allowing the treatment of cultivated areas contaminated by pathogenic agents, such as fungi or others. However, at the end of the 20th century, scientists highlighted the deleterious impacts of a large number of synthetic molecules on the one hand on the . human health and on the other hand on the environment. The harmful effects of chemical pesticides used, for example Glyphosate or N-(phosphonomethyl)glycine, exposure to which considerably increases the risk of non-Hodgkin's lymphoma in humans, or DTT or DichlorodiphenylTrichloroethane, which has a half-life that can exceed 10 years, or more generally organochlorine compounds, synthetic molecules classified as endocrine disruptors, have polluted ecosystems for decades. Citizen and political awareness have led to the massive withdrawal of the most problematic substances. For example, 38 substances of concern were withdrawn at the European level between 2018 and 2019; France unilaterally decided to withdraw metam-sodium and epoxiconazole.

[0006] Although stopping the use of these substances has a positive environmental impact, the agricultural sector risks finding itself without effective tools to optimize the production of cultivated areas and to regulate pest populations.

[0007] Indeed, currently, conventional agriculture is based on a "1 problem = 1 solution" approach. When a pest is identified, a treatment, most often a chemical pesticide, is triggered and this is generally a product containing a specific molecule, targeting the identified pest, which is applied by the farmer to the cultivated area.

[0008] Solutions aimed at optimizing crop yields and controlling the population of pests have been developed.

[0009] A first solution described in patent application No. US 10768156 proposes a system and a method for providing a predictive yield analysis indicating the impact of various parameters on the yield of a cultivated area. For this, for a given type of crop, it is proposed to determine, from an aggregated crop database, an impact on the yield by one or more yield factors including in particular a variety of seeds used in the cultivated area.

[0010] A second solution described in patent application No. WO2020132092 proposes a device and a method for predicting the yield of a cultivated area using a multi-scale predictive framework, machine learning and driven by satellite data and by data from phenological observations on the ground (type of crop, climatic / weather data, ground measurements, satellite data for cultivation conditions and other additional environmental data such as, for example, the soil properties of the cultivated area, management data of the cultivated area, etc.

[0011] From the solutions mentioned above, farmers can thus use this information to make planting choices (for example, seed variety, row spacing, etc.) or post-planting choices (for example, irrigation, application of fertilizers / pesticides, etc.).

[0012] However, although these solutions allow for optimization of crop yield, they are neither suitable for proposing a treatment plan based on natural formulations, nor suitable for allowing the application of natural formulations, depending on environmental parameters, throughout the development cycle of the plant in the cultivated area.

[0013] Furthermore, the proposed solutions generally include the treatment of cultivated surfaces with chemical pesticides which have the advantage of combining power and persistence.

[0014] Nevertheless, the long-term adverse effects on biotopes have justified the gradual cessation of the use of the most worrying substances.

[0015] Finally, the systemic nature of some of these molecules still in use results in the presence of residues of these pesticides in plants, particularly fruits and vegetables, consumed as human or animal food. Technical Problem

[0016] The invention aims to overcome the drawbacks of the prior art. In particular, the invention aims to propose a method and a system for generating a treatment plan adapted for the development of a cultivated area. Such a method and system make it possible to provide a user with a treatment plan, based on natural formulations, for a cultivated area depending on environmental parameters. The application of these natural formulations, depending on pedoclimatic conditions or even upstream of the appearance or advanced development of bioaggressors, makes it possible to strengthen the plant's defense system and promote its growth while ensuring the substitution of chemical products commonly used for the treatment of cultivated areas.The method and system according to the invention thus make it possible to generate a treatment plan adapted to the development of the cultivated surface and its environmental conditions, to prevent the appearance or development of bioaggressors while eliminating the use of chemical pesticides.

[0017] The invention further aims to propose a supervised training method for a deep learning model for generating a treatment plan suitable for the development of a cultivated area. Summary of the invention

[0018] The invention aims to overcome these drawbacks.

[0019] The invention relates in particular to a method for generating a treatment plan adapted to the development conditions of a cultivated surface comprising a first natural formulation adapted to promote growth and stimulate the defense system of a plant variety, said method being implemented by a computer device, the generation method comprising the following steps: • loading values ​​of parameters of the cultivated area, said parameters comprising at least one plant variety, at least one varietal resistance index associated with the plant variety and a development index of the plant variety, • loading at least one value of an environmental parameter, for a predetermined past time period, of the cultivated area, • generation of a treatment recommendation for the cultivated area based on the value of the environmental parameter and the values ​​of the parameters of the cultivated area, said recommendation comprising an application of at least one other natural formulation selected from: • a second natural formulation adapted to stimulate the defense system of the plant variety, and / or • a third natural formulation adapted to prevent the development of a bioaggressor.

[0020] The invention makes it possible to optimize the development of a cultivated area while eliminating the use of chemical pesticides. The method according to the invention integrates the difference in action of natural formulations and proposes the use of natural formulations having synergies in their biological mode of action allowing a strengthening of the plant species and at least equivalent effectiveness against a predetermined bioaggressor. The complementarity of the biological actions between the different potentially usable natural formulations being complex, the invention makes it possible to provide a decision-making tool by generating recommendations adapted to the cultivated area and its environmental conditions.

[0021] According to other optional features of the method, the latter may optionally include one or more of the following features, alone or in combination: • the step of loading at least one value of an environmental parameter is followed by the steps of: • determination of at least one threshold, for the environmental parameter considered, based on the values ​​of the parameters of the cultivated area, • comparison of the value of the environmental parameter with at least one determined threshold and selection of the threshold corresponding to the value of the environmental parameter, and in which the step of generating the treatment recommendation for the cultivated area is a function of the selected threshold. • the parameters of the cultivated area also include soil parameters, nutrient parameters, prior treatment with a natural formulation and / or a type of irrigation. • the first natural formulation adapted to promote growth and stimulate the defense system of a plant variety is applied via the roots, and the recommendation of the other natural formulation includes foliar application. • It further comprises a step of loading at least one value of a forecast environmental parameter, for a predetermined future time period, of the cultivated surface from geographical coordinates of said cultivated surface. • It includes a step of modifying the generated recommendation based on the prior treatment with a natural formulation and / or the type of irrigation. • the step of modifying the generated recommendation also includes taking into account the value of the forecast environmental parameter. • the environmental parameter includes a temperature, a humidity level, a frost risk indicator, a precipitation level, a sunshine level, a wind speed, a detection of a pathogenic organism or a detection of a pathological symptom on the plant variety. • a step of calculating a quantity of the other natural formulation to be applied depending on the prior treatment with the natural formulation, the parameters of the cultivated surface and the value of the environmental parameter. • the first natural formulation comprises microorganisms capable of forming a mutualistic interaction with the plant variety, the second natural formulation comprises at least one elicitor originating from microorganisms, an insect, a nematode or plant fractions or extracts, and / or the third natural formulation comprises at least one active compound having at least one of the following properties: antibacterial, antiviral, insecticide, insect repellent, antigerminative, nematicide or antifungal, and where the recommendation includes the application of the second and third natural formulations, the second natural formulation is applied before the third natural formulation. • the parameters of the cultivated surface comprise soil and / or nutrient parameters, said method further comprising a step of comparing at least one value of a soil and / or nutrient parameter to a predetermined threshold and generating a recommendation for adding an input when the value of the soil and / or nutrient parameter is not equal to the predetermined threshold.

[0022] According to a second object, the invention relates to a system for monitoring a cultivated surface and generating a treatment plan adapted for the development of the cultivated surface comprising a first natural formulation adapted to promote growth and stimulate the defense system of a plant variety, the system comprising a computer device and at least one measuring device, in which: • the measuring device is suitable for measuring at least one environmental parameter of the cultivated area, • the computing device is configured to: • loading values ​​of parameters of the cultivated area, said parameters comprising at least one plant variety, at least one varietal resistance index associated with the plant variety and a development index of the plant variety, • load at least one value of an environmental parameter, for a predetermined past time period, from the cultivated area, • generate a treatment recommendation for the cultivated area based on the value of the environmental parameter and the values ​​of the parameters of the cultivated area, said recommendation comprising an application of at least one other natural formulation selected from: • a second natural formulation adapted to stimulate the defense system of the plant variety, and / or • a third natural formulation adapted to prevent the development of a bioaggressor.

[0023] According to other optional features of the system, the computing device may be configured to: • determine at least one threshold, for the environmental parameter considered, based on the values ​​of the parameters of the cultivated area, • compare the value of the environmental parameter to at least one determined threshold and select the threshold corresponding to the value of the environmental parameter, and to generate the treatment recommendation for the cultivated area based on the selected threshold.

[0024] According to a third object, the invention relates to a machine learning model trained to predict a treatment plan adapted to the development conditions of a cultivated surface comprising a first natural formulation adapted to promote growth and stimulate the defense system of a plant variety, the trained machine learning model being obtained according to the following steps: • acquisition, at each of the acquisition times of a plurality of successive acquisition times, of reference parameter values ​​of a plurality of reference cultivated areas, the reference parameters comprising, for each reference cultivated area, at least: • values ​​of parameters of the cultivated area, the parameters of the cultivated area comprising a plant variety, a varietal resistance index associated with the plant variety, a development index of the plant variety, • a value of an environmental parameter, • treatment data comprising at least one other natural formulation selected from a second natural formulation suitable for stimulating the defense system of the plant variety, and / or a third natural formulation suitable for preventing the development of a bioaggressor, and • data on the effectiveness of treatment using at least one other natural formulation, • calculation, at each acquisition time, for each of the plurality of reference cultivated areas, of mathematical quantities, from the values ​​of the parameters of the cultivated area, of the value of the associated environmental parameter, acquired at acquisition times included in a predetermined time period preceding the current acquisition time, • creation, for each of the plurality of reference cultivated areas, of a reference mathematical function from the parameter values ​​of the cultivated area, the value of the environmental parameter and the mathematical quantities, • association, at each acquisition time, for each of the plurality of reference cultivated areas, of the reference mathematical function with the treatment efficiency data of the reference cultivated area, so as to obtain machine learning model variables, and • training a machine learning model to predict a treatment plan comprising at least one alternative natural formulation, from a machine learning algorithm and the machine learning model variables. Brief description of the drawings

[0025] Other characteristics and advantages of the invention will be better understood on reading the description which follows and with reference to the appended drawings, given for illustrative purposes and in no way limiting.

[0026] [Fig-1] [Fig. 1] represents an embodiment of a method for generating a treatment plan.

[0027] [Fig.2] [Fig.2] represents a graph describing the level of effectiveness of natural formulations as a function of time in the treatment against potato blight (Solanum tuberosum).

[0028] [Fig.3] [Fig.3] represents an embodiment of a system for monitoring a cultivated area and generating a treatment plan.

[0029] The figures do not necessarily respect the scales, in particular in thickness, and this is for illustration purposes.

[0030] Aspects of the present invention are described with reference to flowcharts and / or functional diagrams of methods, systems according to embodiments of the invention.

[0031] In the figures, flowcharts and block diagrams illustrate the architecture, functionality, and operation of possible implementations of systems and methods according to various embodiments of the present invention. In this regard, each block in the flowcharts or block diagrams may represent a system, device, module, or code, which includes one or more executable instructions for implementing the specified logical function(s). In some implementations, the functions associated with the blocks may appear in a different order than shown in the figures. For example, two blocks shown in succession may, in fact, be executed substantially simultaneously, or the blocks may sometimes be executed in reverse order, depending on the functionality involved.Each block of the block diagrams and / or flowchart, and combinations of blocks in the block diagrams and / or flowchart, can be implemented by special hardware systems that . perform specified functions or acts or carry out combinations of special hardware and computer instructions. Description of the embodiments

[0032] Below, we describe a summary of the invention and the associated vocabulary, before presenting the disadvantages of the prior art, and finally showing in more detail how the invention overcomes them.

[0033] In the remainder of the description, the expression “computer device” corresponds to any device comprising a processing unit or a processor, for example in the form of a microcontroller cooperating with a data memory, possibly a program memory, said memories being able to be dissociated. The processing unit cooperates with said memories via internal communication buses.

[0034] By "Natural formulation" within the meaning of the invention is meant a formulation comprising one or more chemical compounds or substances which are found in nature, preferably produced by a living organism. A natural formulation within the meaning of the invention includes chemical compounds of natural origin which can also be obtained by chemical synthesis.

[0035] The term "bioaggressor" refers to any type of pest known in the field of agriculture, in particular insects or worms, any type of phytopathogenic agents, such as fungi or weeds.

[0036] For the purposes of the invention, the term "coupled" means connected, directly or indirectly, with one or more intermediate elements. Two elements can be coupled mechanically, electrically or linked by a communication channel.

[0037] The expression "plant species" defines a group of plants having similar characteristics and capable of reproducing among themselves, but sterile with any plant of another species. As a non-limiting example, the plant species considered is Solarium tuberosum.

[0038] In the remainder of the description, the various features presented and / or claimed may be advantageously combined. Their presence in the description or in different dependent claims does not exclude this possibility.

[0039] As mentioned, the current situation is that the use of chemical fertilizers and pesticides in agriculture is widespread worldwide. Chemical fertilizers are used to maximize crop yields. Chemical pesticides are used to eradicate the presence of pests and diseases in crops. Each chemical pesticide used is specifically effective against a or several bioaggressors, the development of resistance of cultivated plants against these bioaggressors is not a priority.

[0040] However, the withdrawal and the desire to limit the use of chemical pesticides leaves agricultural professionals without an alternative solution to chemical pesticides.

[0041] The substitution of chemical pesticides by natural formulations is possible but the modes of action of the latter are different. Programs composed of natural formulations combine actions that can be powerful / effective but of limited duration while the persistence of natural formulations with direct action, that is to say targeting a bioaggressor, is limited but indirect persistence is possible via elicitation mechanisms. In addition, the use of natural substances compared to chemical pesticides allows the absence of non-biodegradable residues on the external parts of the plant, the persistence of the natural protection linked to the stimulation of natural defenses is linked to the persistence of compounds endogenous to the plant cells and not to the compounds having allowed this external stimulation.

[0042] Although the substitution of chemical pesticides with natural formulations remains possible, managing the application of natural formulations is much more complex and requires choosing the right natural formulation and applying it at the right time in order to obtain protection equivalent to that obtained with chemical pesticides.

[0043] Active ingredients of natural origin most often have a contact action which allows to benefit from an instantaneous effect but do not offer the persistence to protect against a health risk in the long term. As a result, farmers engaged in organic farming most often use disease-resistant varieties which do not have the same yield potential as conventional varieties and thus resort, often abusively, to copper for its anti-germination effect on fungal spores. The use of copper in combination with resistant varieties constitutes the basis of organic farming but does not allow to achieve production yields of conventional varieties and, therefore, appears limiting to replace the current agricultural model based on the use of chemical pesticides.

[0044] In view of these findings, the invention proposes the generation of a treatment plan, based on natural formulations, comprising one or more recommendations for treating a cultivated area, depending on parameters linked to the cultivated area and its environment. Indeed, the different stages of development of a cultivated plant imply different needs, whether in the context of their growth, or resistance to climatic conditions and bioaggressors. Although the use of chemical fertilizers and pesticides makes it possible to easily meet the needs associated with the different stages of development of a cultivated plant, the use of alternatives based on natural formulations requires a more detailed analysis of environmental parameters in order to continually adapt the application of natural formulations according to these parameters and thus optimize the resistance of the cultivated plant according to the type of attacks to which it could be subjected.

[0045] Thus, according to a first aspect, the invention relates to a method for generating a treatment plan suitable for the development of a cultivated surface comprising a first natural formulation suitable for promoting growth and stimulating the defense system of a plant species, the method comprising a step of loading parameter values ​​of the cultivated surface, a step of loading at least one value of an environmental parameter and a step of generating a recommendation for treatment of the cultivated surface.

[0046] [Fig.l] illustrates an example of a method for generating a treatment plan 1 according to the invention implemented by a computer device.

[0047] As mentioned previously, the invention comprises a step of loading the parameter values ​​of the cultivated surface 100.

[0048] The parameters of the cultivated area include at least one plant variety, at least one varietal resistance index associated with the plant variety and at least one development index of the plant variety.

[0049] A plant variety corresponds to a set of plants identified by morphological, physiological and genetic characteristics which distinguish it from other plants of the same species. As a non-limiting example, the plant variety considered is the Allians variety.

[0050] In addition, the parameters of the cultivated area include one or more varietal resistance indices associated with the plant variety. Each varietal resistance index may, for example, describe the overall resistance of the variety of the plant species in question or describe the level of resistance of the plant variety with respect to a given environmental parameter, examples of which will be presented later. As illustrative examples, the varietal resistance index may correspond to a value approximated to unity ranging from 1 to 10, reflecting a sensitivity to diseases, pests and / or physiological accidents, a value ranging from 7 to 10 describing a plant variety called "resistant", a value ranging from 5 to 6 describing a plant variety called "semi-resistant" or a value ranging from 1 to 4 describing a plant variety called "sensitive" for a given environmental parameter.Thus, the closer the varietal resistance index is to 1, the higher the sensitivity of the corresponding variety to the parameter. environmental considered. High sensitivity generally indicates that the variety will likely not be able to survive, in the event of the development of a pest, without the application of a chemical pesticide.

[0051] In addition, the parameters of the cultivated area include a development index of the plant variety. The development index of the plant variety may correspond to a value describing the phenological stage of the plant variety of the cultivated area. By way of non-limiting examples, the development index of the cultivated area may take the value “1” to describe the germination of the plant variety, the value “2” to describe the growth phase of the plant variety once it has emerged from the ground until the start of senescence, the value “3” to describe the phase from the start of senescence until the time of harvesting of the plant variety and / or its fruits.

[0052] In order to facilitate the association of the development index of the plant variety with a stage of development of the plant variety, the development index of the plant variety may take the value "1" during a time interval of up to four weeks after the seeds have been sown, the value "2" during a time interval of from four weeks to eight weeks after the seeds have been sown, the value "3" during a time interval of from eight weeks to twelve weeks after the seeds have been sown. On the other hand, depending on the plant variety considered, said time intervals will be likely to vary so that they correspond to the phenological stages described above.

[0053] In another embodiment, the development index of the cultivated area can take a value describing a stage of the BBCH code (for Biologische Bundesanstalt, Bundessortenamt und CHemische Industrie according to German terminology).Thus, the development index can take the value "0" to describe the germination of the plant variety, the value "1" to describe the development of the leaves, the value "2" to describe the formation of secondary shoots / tillering, the value "3" to describe the elongation of the stem / the formation of the rosette / the development of the shoots, the value "4" to describe the formation of the vegetative parts of harvest or of the organs of vegetative propagation / development of the organs of sexual reproduction, swelling of the ear or panicle, the value "5" to describe the appearance of the inflorescence, the value "6" to describe flowering, the value "7" to describe the development of the fruits, the value "8" to describe the ripening of the fruits, the value "9" to describe the phase from the beginning of senescence until the time of harvesting of the plant variety and / or its fruits.

[0054] As resistance to certain environmental conditions or to certain bioaggressors can vary during the different phenological stages of the plant variety, the varietal resistance index, in relation to an environmental parameter given, can also be associated with the development index of the plant variety. Indeed, the plant variety may be sensitive to certain types of bio aggressors during the germination phase and not during the senescence phase and vice versa. In addition, during the development of the plant variety, certain agricultural practices, such as the closure of mounds, may complicate the application of natural formulations and may justify the application of these formulations during an earlier phenological stage to allow the development of the plant variety's defense system. Similarly, when the senescence phase of the plant variety begins, the application of certain natural formulations, in particular those containing an elicitor, is no longer of interest in the context of prevention against certain types of bio aggressors.In fact, the stimulation of the plants' defense systems becomes ineffective as soon as the plant enters the senescence phase.

[0055] In one embodiment of the generation method according to the invention, the parameters of the cultivated surface may comprise geographic coordinates which make it possible to locate the cultivated surface.

[0056] In another embodiment of the generation method according to the invention, the latter may comprise a step 150 of comparing at least one value of a pedological and / or nutrient parameter to a predetermined threshold and generating a recommendation for adding an input when the value of the pedological and / or nutrient parameter is not equal to the predetermined threshold.

[0057] Advantageously, the input may correspond to any type of material of known type intended to improve / modify the pedological parameters of the cultivated surface. The recommended input may comprise a composition adapted to modify the pedological and / or nutrient parameters of the cultivated surface so that the value of the pedological and / or nutrient parameter is equal to the predetermined threshold.

[0058] For example, the input may comprise Vinasse, a hydrolyzate of fish proteins and marine algae extracts, trace elements and magnesium sulfate, organic nitrogen in the form of Magnesium Oxide and Sulphur.

[0059] Generally speaking, the soil parameters describe the condition of the soil in the cultivated area.

[0060] Soil parameters may include, but are not limited to: • A soil texture. The soil texture may correspond to a ratio between different elements likely to make up the soil of the cultivated surface such as clay, silt, sand. Within the framework of the invention, clay / silt / sand ratios may be defined in order to describe a suitable support for the cultivation of a predetermined plant variety. A suitable support may, for example, correspond to a clay / silt / sand ratio making it possible to avoid soil anoxia, i.e. a reduction in dissolved or present and bioavailable oxygen in the soil, and low inertia in the face of abiotic stresses. • A hydrogen potential (pH) of the soil. Indeed, a pH adapted to the development of certain microorganisms can prove advantageous in particular for the activity of bacteria responsible for the transformation of organic matter, or for earthworms. Depending on the texture of the soil, the pH-water value can range from 6.5 to 8, preferably 6.8 to 7.5 and even more preferably 7 to 7.2, this generally allows optimal development of these microorganisms. • Organic matter content. Indeed, organic matter is a support but also a source of nutrients for the development of microorganism communities and the value of the organic matter content can range from 1% to 2.5%, preferably from 1.2% to 2.1%, depending on the texture of the soil. • A carbon to nitrogen (C / N) mass ratio. The C / N ratio is an indicator that allows us to judge the degree of evolution of organic matter, that is to say its ability to decompose more or less quickly in the soil and ensure optimal development of microorganisms. For example, the C / N ratio can range from 10 to 25.

[0061] Alternatively or in addition, the parameters of the cultivated area may include nutrient parameters. The nutrient parameters describe the accessibility, by the plant variety, to the elements necessary for optimal elicitation. Without this being limiting, the nutrient parameters may correspond to a concentration, in the soil of the cultivated area, of nitrogen, phosphorus and potassium.

[0062] Still within the scope of the invention, the method for generating a treatment plan 1 comprises a step 200 of loading at least one value of an environmental parameter, for a predetermined past time period, of the cultivated surface.

[0063] By way of non-limiting examples, the environmental parameter may comprise a temperature, a humidity level, for example of the air and / or the foliage of the plant variety, a frost risk indicator, a precipitation level, a sunshine level, a wind speed, a detection of a pathogenic organism and / or a detection of a pathological symptom on the plant variety, for example leaf blight or tuber blight.

[0064] The past time period may correspond to a duration, preferably expressed in days, prior to the time when the generation method according to the invention is implemented. For example, the past time period may correspond to a duration of one hour or several hours, one day or several days. The value of the environmental parameter considered may be loaded at regular intervals over the entire time period, for example once per hour. The value of the environmental parameter may also correspond to an averaged value from all the values ​​loaded at regular intervals for the environmental parameter considered, such as in particular an average precipitation level over a time period of three days. The value of the environmental parameter may also correspond to a value indicating or not the detection of an environmental parameter, in particular that of a pathogenic organism or pathological symptoms, due for example to a bioaggressor, on the plant variety.

[0065] In one embodiment according to the invention, the method may comprise a step of loading at least one value of a forecast environmental parameter 250, for a predetermined future time period, of the cultivated surface from geographical coordinates of said cultivated surface.

[0066] The term “predetermined future time period” means a duration, like the past time period, preferably expressed in days, subsequent to the time when the generation method according to the invention is implemented.

[0067] Like the environmental parameter, the forecast environmental parameter may include a temperature, a humidity level, for example of the air and / or of the foliage of the plant variety, a frost risk indicator, a precipitation level, a sunshine level, a wind speed, a detection of a pathogenic organism and / or a detection of a pathological symptom on the plant variety, for example leaf blight or tuber blight.

[0068] In a particular embodiment, the method according to the invention may further comprise a step 300 of determining a threshold, for the environmental parameter considered, as a function of the parameter values ​​of the cultivated surface.

[0069] In the context of the invention, threshold means at least one threshold associated with a given environmental parameter. The threshold may be represented by a numerical value or by a range of values ​​forming an interval. Indeed, in the invention it may be advantageous to determine several thresholds for a given environmental parameter in order to be able to generate a recommendation for the application of a natural formulation adapted to the environmental conditions of the cultivated area.

[0070] In the context of the invention, it is provided that the parameters of the cultivated area as well as environmental parameters and one or more predetermined thresholds are included in a data repository. The values ​​of the environmental parameters considered are for example communicated or loaded by the computer device and then encoded in the data repository. Depending on the time period considered, the value of the environmental parameter encoded in the data repository may correspond to an average of the values ​​of the environmental parameter loaded at regular intervals, for example every hour.

[0071] According to an embodiment of the method for generating a treatment plan 1 according to the invention, the latter may comprise a step 400 of comparing the value of the environmental parameter to the at least one determined threshold and the selection of the threshold corresponding to the value of the environmental parameter.

[0072] For example, depending on the development index of the plant variety and the environmental parameter considered, one or more thresholds can be predetermined. Each threshold can be associated with one or more natural formulations. The value of the environmental parameter considered is compared to the threshold(s) and the threshold comprising the value of the environmental parameter is selected.

[0073] Alternatively or in addition, a threshold may take a single value, for example the value “0” or “1”. In this case, the environmental parameter considered may correspond to the detection or not of a pathogenic organism or a pathological symptom on the plant variety, and the value of the environmental parameter may take the value “0” if no pathogenic organism or pathological symptom is detected or the value “1” if a pathogenic organism or a pathological symptom is detected. The value of the environmental parameter considered is compared to the threshold(s) and the threshold corresponding to the value of the environmental parameter is selected.

[0074] In a particular embodiment, in order to optimize the application of natural formulations and maximize their effects, the parameters of the cultivated surface may include a prior treatment with a natural formulation according to the invention, for example during a given time period.

[0075] As already mentioned, different thresholds may be applied depending on the development of the plant variety, but taking into account certain environmental parameters may remain relevant during the development cycle of the plant variety. However, any application of a natural formulation that would not be necessary should be avoided, so taking into account previous treatments already applied to the cultivated SI surface may prove necessary.

[0076] Alternatively or in addition, the parameters of the cultivated area may also include a type of irrigation. Indeed, depending on the water requirements, certain cultivated areas may include sprinkler irrigation systems, micro-irrigation systems or even drip irrigation or underground drip irrigation systems of known types. Thus, natural formulations may be selected according to the type of irrigation, in particular to facilitate their application to cultivated areas.

[0077] Data relating to the type of irrigation can be encoded in the data repository like data relating to the development index of the plant variety or the natural formulation applied.

[0078] Alternatively, the comparison step 400 can be implemented for a second environmental parameter. In this case, the steps of the method according to the invention are also implemented for the second environmental parameter considered. Thus, the value of the second environmental parameter is compared 400 to a second predetermined threshold. When the values ​​of the first and second environmental parameters correspond respectively to the first and second predetermined thresholds, the following step of the method can be implemented. Those skilled in the art will understand that the application of a natural formulation, in particular with a view to preventing the appearance or development of bioaggressors, may require taking into account different environmental parameters.Indeed, certain bioaggressors require, in order to develop, that very specific environmental parameters be met, such as in particular a specific humidity level in the air and a specific ambient temperature. The invention cannot be limited to these environmental parameters; other environmental parameters characterizing the optimal conditions for the development of a bioaggressor may be used alone or in combination.

[0079] In a particular embodiment of the method according to the invention, when the step 400 of comparing the value of the environmental parameter to the at least one determined threshold does not make it possible to identify a threshold corresponding to the value of the environmental parameter, the method stops.

[0080] Alternatively, when the step 400 of comparing the value of the environmental parameter to the determined threshold does not make it possible to identify a threshold corresponding to the value of the environmental parameter, a new step of loading at least one value of an environmental parameter 200 is implemented.

[0081] When no threshold is identified during the comparison step 400, the new step of loading at least one value of an environmental parameter 200 can be implemented for another environmental parameter. Furthermore, it is provided that the new step of loading at least one value of an environmental parameter 200 is implemented for the same environmental parameter when a new value of the environmental parameter is communicated to the computing device 10.

[0082] Finally, a method according to the invention comprises a step 500 of generating a recommendation RI, R2 for treating the cultivated surface as a function of the value of the environmental parameter and the values ​​of the parameters of the cultivated surface.

[0083] In a particular embodiment of the generation method 1 according to the invention, when the step 200 of loading at least one value of an environmental parameter is followed by the steps of: • determination 300 of at least one threshold, for the environmental parameter considered, based on the values ​​of the parameters of the cultivated area, • comparison 400 of the value of the environmental parameter with the determined threshold and selection of the threshold corresponding to the value of the environmental parameter, the generation step 500 of the RI, R2 recommendation for treatment of the cultivated area depends on the selected threshold.

[0084] Thus, when the threshold is selected, the corresponding natural formulation(s) are encoded in a recommendation RI, R2, for example in the form of a message displayable via a suitable human-machine interface.

[0085] On the other hand, it is provided that the method according to the invention can take into consideration other parameters for the generation of a recommendation. The other parameters can for example be another environmental parameter and / or another parameter of the cultivated surface.

[0086] In a particular embodiment, the method according to the invention may comprise a step of modifying the generated recommendation as a function of the prior treatment with a natural formulation during a given time period and / or the type of irrigation.

[0087] In a particular embodiment of the recommendation modification step, the generated recommendation further comprises taking into account the value of the forecast environmental parameter.

[0088] Indeed, as described above, the value of the environmental parameter may correspond to an average value of the environmental parameter for a predetermined past time period. On the other hand, it may be advantageous to take into consideration the evolution of the environmental parameter considered for a future time period in order to recommend or not the application of one or more natural formulations.

[0089] Generally speaking, the application of certain natural formulations may be recommended during a more or less intense rainy episode, in particular when the average precipitation level measured exceeds a predetermined threshold, in order to prevent the development of bioaggressors. However, the generation of this recommendation may be subject to a value of the forecast average precipitation level and / or to other forecast environmental parameters such as temperature or humidity level for example. Indeed, it will be understood that depending on the development conditions of certain bioaggressors, it may be advantageous to take into account takes into account the value of an environmental parameter for a past period and for a future period in order to generate or not a recommendation for the application of one or more natural formulations.

[0090] Still within the framework of the invention, the recommendation RI, R2 comprises the application of at least one other natural formulation selected from: • a second natural formulation adapted to stimulate the defense system of the plant variety, and / or • a third natural formulation adapted to prevent the development of a bioaggressor.

[0091] In a preferred embodiment of the method according to the invention, the first natural formulation suitable for promoting growth and stimulating the defense system of a plant variety is applied via the roots, and the recommendation of the other natural formulation comprises an application via the leaves.

[0092] By way of non-limiting examples, the application by root route can be carried out by coating or pralining the seeds or via a soil irrigation technique of known types, in particular by micro-irrigation. In addition, the recommended foliar application can be an application by spraying, that is to say the projection of a composition in the form of drops, droplets or even an aerosol by a spraying device of known type, a mounted, semi-mounted or self-propelled sprayer towed or pulled by a motorized machine such as a tractor for example, but also by aerial spreading, such as by plane or helicopter or even via the use of drones, autonomous or not.

[0093] In the context of the invention, the plant species-soil interaction is of particular importance since it makes it possible to improve the effectiveness of the immune response of the plant variety following the application of one or more natural formulations. The soil of the cultivated surface is a biotope in its own right within which exchanges take place permanently with the roots of the plant variety, the soil provides nutrients and in return the microorganisms feed on the root exudates. A high level of biological life in the soil allows it to act as a suppressor of biotic and abiotic stress.

[0094] Thus, by way of examples, the natural formulation adapted to promote growth and stimulate the defense system of the plant variety may comprise microorganisms beneficial to the rhizosphere of known type such as symbiotic microorganisms like Glomus intraradices, microorganisms making it possible to solubilize nutrients and increase their bioavailability for the roots, microorganisms antagonistic to soil-borne pathogens, elicitor microorganisms (capable of stimulating the defenses of the plant by stimulating the root cells), in particular fungi of the genus Trichoderma or bacteria of the genus Bacillus or Pseudomonas, microorganisms producing hormones stimulating plant growth, prebiotics for example rich in nitrogen, carbon and are selected to be metabolized only by target microorganisms (for example by targeting enzymatic activities specific to the microbial communities of interest), such as simple sugars, amino acids, algae extracts, or humic acids, biostimulant substances having positive effects on the growth and general vigor of the plant, such as for example plant extracts, amino acids of animal origin, extracts of degraded organic matter and having the capacity to stimulate germination and root and / or leaf growth (directly or indirectly).

[0095] Alternatively or in addition, the natural formulation adapted to stimulate the defense system of the plant variety may comprise Natural Defense Stimulators adapted to activate the defense system of plants, such as the elicitor compounds mentioned above, but also extracts of plant, animal or microbial origin mimicking fungal aggression by activating the membrane receptors of the plant, purified plant extracts stimulating the plant by acting directly in the cascades of metabolic reactions mimicking the elicitation reaction.

[0096] Finally, the natural formulation suitable for preventing the development of a bioaggressor may generally comprise any substance having an adverse impact on the development of the bioaggressor in question. By way of non-limiting examples, said natural formulation may comprise an anti-germination substance such as copper or a substance suitable for creating a barrier effect such as a plant substance or a mineral substance such as talc or kaolin.

[0097] Thus, in a preferred embodiment of the method according to the invention: • the first natural formulation includes microorganisms capable of forming a mutualistic interaction with the plant variety, and preferably an elicitor microorganism, • the second natural formulation comprises at least one elicitor originating from a microorganism, for example a bacterium, a virus, or even an insect, a nematode, plant fractions or extracts, and / or • the third natural formulation comprises at least one active compound having at least one of the following properties: antibacterial, antiviral, insecticide, insect repellent, nematicide, antigerminative, or antifungal.

[0098] Furthermore, when the recommendation includes the application of the second and third natural formulations, the second natural formulation is applied before the third natural formulation. Indeed, as mentioned previously, one of the objectives of a method according to the invention is to provide a treatment plan to prevent the appearance or development of bioaggressors. Thus, the application of a natural formulation directly targeting the bioaggressor will preferably be used as a last resort. Therefore, the development of immunity of the plant variety of the cultivated SI surface will be preferred before the application of the third natural formulation, and if necessary, a third natural anti-germination formulation may preferably be used.

[0099] In one embodiment of the method according to the invention, the latter may further comprise a step 550 of calculating a quantity of the other natural formulation to be applied as a function of the prior treatment with the formulation, the values ​​of the parameters of the cultivated surface and the value of the environmental parameter. The quantity of the other natural formulation may further be recorded in the data repository and be encoded in the generated treatment recommendation.

[0100] In another embodiment of the invention, the generation method 1 comprises a step 600 of displaying at least one treatment recommendation for the cultivated SI surface. For example, the treatment recommendation may take the form of a message displayable by a human-machine interface of a known type of the computing device.

[0101] Alternatively or in addition, the generation 1 method may comprise a step of application (not shown in the figures) of the recommended natural formulation(s), by a user. The application step may be a foliar or soil application such as those described previously.

[0102] Example of representation of a data repository

[0103] As described above, the parameters of the cultivated area, the environmental parameters and possibly one or more predetermined thresholds may be included in a data repository. The table below presents an example of representation of a data repository according to the invention. [Tables 1] Paramètres de la wrfac® cuiwée jAjA -À-,* Jw A- U4JI JAM A. WAL A 'WÂ-jt-K-'lt Recommended formufo^on naturel Espèce végétale : Scmwm toswro Index de dévefeppemeh I de la va?lésé végétale (stade BBCH) index de résistance wièfoîe Formulait n naturelle appliquée Niveau de précipitai® n mm en mesmé efou prévistwi ei (en mnsj Periode temporelle 8e<<s) 7 Eliciteur foliaire 4 ÎÎW 3 iWS > S mm 1 - Oismern feteBe 2 7 na 12 SW) 4 hours >20 mm; i - s-.,PsUm\e « MK oarwe ne B mm 3 purs ]0 mm-S mm] 1 - Eiidtem Wiaire p mm« 10 mm] 1 - Eimiiem foUaBC 2-antjgefminatif 3 - sukmv.m-■eiM bat se'** )18 mm 20 mm] i -amigernWi? 2 - Eiidtem fcliaim ■> am^ennmadt' 3 -antgerminatif na § mm § pars >20 mm l--m3pmmir^

[0104] Table 1 presents an example of a data repository accessible by, or stored in memory in, the computing device. In this example, the parameters of the cultivated area include a pre-treatment with a natural foliar elicitor-type formulation, carried out during the leaf development phase (BBCH index = 1).

[0105] As mentioned previously in connection with the comparison step 400 and as presented in Table 1, depending on the development index of the plant variety and the environmental parameter considered, one or more thresholds may be predetermined. In these examples, the thresholds are characterized by ranges of values, for example ]0 mm-5 mm], or limit values, for example >20 mm. Each threshold is associated with one or more natural formulations. The value of the environmental parameter considered is compared to the threshold(s) and the threshold comprising the value of the environmental parameter is selected.

[0106] As an illustrative example, in Table 1, the plant species Solanum tuberosum, variety Allians, is in the secondary shoot formation / tillering phase (see value “2”), a plurality of values ​​of an environmental parameter have been loaded into the data repository. These different values ​​correspond to averaged measurements of the precipitation level over predetermined time periods. It can thus be seen that for each predetermined time period, i.e. 4 hours, 3 days and 5 days, the average precipitation level has been calculated. For each time period, the average value of the precipitation level is then compared to the associated threshold(s). In the example presented, the average precipitation level measured at 8 mm for a time period of 3 days is the only value that corresponds to one of the determined thresholds, in this case the threshold 5 mm-10 mm.Thus, the threshold ]5 mm-10 mm] is selected and the next step of the process can be implemented.

[0107] Indeed, as presented in Table 1, each threshold can be associated with at least one predetermined natural formulation. In particular, when the forecast environmental parameter corresponds to the forecast average precipitation level, over a time period of 4 hours, and the latter is greater than a threshold of 20 mm, a natural formulation of the barrier substance type is selected.

[0108] In the example presented in connection with the measured average precipitation level of 8 mm, the threshold ]5 mm-10 mm] with which three distinct natural formulation recommendations are associated is selected. By default, the recommendation will include the natural formulation noted “1”, in this case a foliar elicitor. The natural formulation noted “1” is normally selected when the cultivated area has not been subject to an application of a natural formulation during the germination phase. On the other hand, when the cultivated area has been subject to an application of a natural formulation, in this case a foliar elicitor, the recommendation can be modified so that it includes the natural formulation noted “2”, in this case an anti-germination agent.

[0109] Furthermore, depending on the different phenological stages of the plant variety (BBCH stage), a recommendation may include the application of an input. Indeed, as detailed previously, the parameters of the cultivated SI surface may include soil parameters or even nutrient parameters. These parameters can be monitored during the development of the cultivated SI surface and, like the environmental parameter described in relation to the average precipitation level measured, other thresholds may be associated with soil or nutrient parameters.

[0110] Furthermore, when the cultivated area parameter includes a type of irrigation (not shown in Table 1). For example, for spray type irrigation, at a predetermined flow rate, the average quantity of water delivered during the predetermined time period (e.g. 4 hours, 3 days and / or 5 days) can be calculated and added to the measured and / or forecasted precipitation level to generate an average of the measured and / or forecasted precipitation level and the quantity of water delivered. When it is the forecasted or measured and forecasted precipitation level that is determined, the recommendation can be modified so that it includes the natural formulation denoted "3", in this case a barrier substance.Although the level of precipitation and / or the average quantity of water delivered by spraying can promote the development of certain bioaggressors, other environmental parameters can be taken into account in addition, such as in particular the humidity level and / or the temperature.

[0111] As already mentioned previously, one of the objectives of the invention is to prevent the appearance or development of pests and diseases while eliminating the use of chemical pesticides. For this, the different natural formulations applied preferably have a complementary mode of action in order to provide protection equivalent to a chemical pesticide commonly used in the treatment of a predetermined pest. [Fig. 2] shows a graph describing the level of effectiveness of three natural formulations as a function of time. The level of effectiveness of each natural formulation is expressed as a percentage of protection compared to a reference chemical pesticide noted PI, in this case the fungicide Infinito® comprising the active chemical compounds fluopicolide and propamocarb hydrochloride for the treatment against potato blight.Here, a first natural formulation 51 to promote growth and stimulate the defense system was applied when planting seedlings of the plant species Solanum tuberosum, variety Allians. The first natural formulation is a root elicitor, in this case the Gaia sol® formulation, which has an efficiency of approximately 30% for a period of approximately 42 days. The second natural formulation 52 adapted to stimulate the defense system of the plant variety is a foliar elicitor. in this case the Milvax® formulation, and is selected based on its complementary biological mode of action to the root elicitor, here the Gaia sol® formulation. This improves the efficacy against potato blight by up to approximately 50% compared to the Infinito® fungicide. The second natural formulation can be applied at regular intervals to maintain protection for a given period. The third natural formulation 53 adapted to prevent the development of a pest is an antigermination or antifungal substance, in this case Euclean® and is selected based on its complementary biological mode of action to the first and second natural formulations. This improves the efficacy against the pest by up to 100% compared to the Infinito® fungicide.The selection of natural formulations recommended within the framework of a method according to the invention allows for coordination of the actions of the natural formulations. Thus, it is the combination of natural formulations applied at predetermined times and according to environmental conditions which makes it possible to effectively combat a bioaggressor and eliminate the use of chemical pesticides.

[0112] The values ​​relating to the effectiveness of natural formulations are presented for illustrative purposes and may vary depending on the combination of natural formulations used, the values ​​of the parameters of the cultivated surface, in particular the conditions of the cultivated surface in relation to the soil and / or nutritional parameters.

[0113] According to a second aspect, and as described in connection with [Fig. 3], the invention relates to a system 2 for monitoring a cultivated SI surface and generating a treatment plan adapted for the development of the cultivated SI surface, the cultivated SI surface comprising a first natural formulation 51 adapted to promote growth and stimulate the defense system of a plant variety, the system 2 comprising a computer device 10 and at least one measuring device 20.

[0114] Preferably, the first natural formulation 51 comprises at least one root elicitor. Without this being limiting, the first natural formulation 51 comprises a marine algae extract, for example Ascophyllum nodosum, a fish meal hydrolyzate, a bacterium of the Bacillus amyloliquefaciens type, a fungus of the Trichoderma harzianum type and humic and fulvic acids.

[0115] As illustrative examples, the computing device 10 may be a smartphone or a tablet connected and configured to communicate via a wireless communication network with the measuring device 20.

[0116] In a particular embodiment of the system according to the invention, the measuring device 20 may be a weather station of known type. The weather station, or more generally the measuring device 20, may comprise one or more sensors of known type suitable for measuring one or more environmental parameters as mentioned above.

[0117] Alternatively or additionally, the measuring device 20 may further be adapted to measure a detection of a pathogenic organism or a detection of a pathological symptom on the plant variety. For example, the measuring device 20 may comprise one or more spore sensors of known type placed in the cultivated SI surface.

[0118] Thus, depending on the environmental parameters of interest, a system according to the invention may comprise a plurality of sensors positioned in or near the cultivated SI surface.

[0119] In the invention, the measuring device 20 is adapted to measure at least one environmental parameter P2S1 of the cultivated SI surface.

[0120] Further, the measuring device 20 may be configured to perform measurements at regular intervals, for example every hour, and to communicate the measurements to the computing device 10.

[0121] Still in the invention, the computing device 10 is configured to: • load P1S1 parameters of the cultivated SI area, said parameters comprising at least one plant variety, a varietal resistance index associated with the plant variety and a development index of the plant variety, • load at least one value of an environmental parameter P2S1, for a predetermined past time period, of the cultivated SI surface, • determine at least one threshold, for the environmental parameter considered, depending on the plant variety, the varietal resistance index and the development index of the plant variety, • compare the value of the environmental parameter P2S1 with at least one determined threshold and select the threshold corresponding to the value of the environmental parameter.

[0122] In addition, the P1S1 parameters of the cultivated SI surface may include a pre-treatment with a natural formulation during a given time period, geographical coordinates of the cultivated SI surface and / or a type of irrigation.

[0123] The value of the environmental parameter P2S1 of the cultivated SI surface can be communicated directly to the computing device 10 or to the remote server 40, in particular when said value comes from a measurement carried out by a measuring device 20 positioned in or near the cultivated SI surface. However, it is provided that certain environmental parameter values ​​can come from a third-party remote server 30.

[0124] Furthermore, the P2S1 parameters of the cultivated SI surface can be recorded on a remote server 40 or directly in the computing device 10.

[0125] In addition, the remote 40 and third-party 30 servers can communicate with each other, in particular by sending a message comprising the value of the environmental parameter P2S1. In all cases, the remote 40 and third-party 30 servers are configured to communicate with the computing device 10 to which the value of the environmental parameter P2S1 and the parameters P1S1 of the cultivated surface SI are sent.

[0126] In the invention, the computing device 10 is further configured to generate a recommendation RI, R2 for treatment of the cultivated surface SI as a function of the selected threshold, said recommendation RI, R2 comprising the application of at least one other natural formulation 50 selected from: • a second natural formulation 52 adapted to stimulate the defense system of the plant variety, and / or • a third natural formulation 53 adapted to prevent the development of a bioaggressor.

[0127] For example, the second natural formulation 52 may comprise a foliar elicitor. More particularly, the second natural formulation 52 may comprise a decoction of field horsetail, clove (Syzygium aromaticum), nettle (Urtica dioica), tea plant extract (CameIlia sinensis), soy lecithin, sunflower oil and guar gum (Cyamopsis tetragonolobus).

[0128] In addition, the third natural formulation 53 may comprise a so-called “intervention” substance such as anti-germination, curative, barrier, and / or natural biocide substances.

[0129] For example, the third natural formulation 53 may comprise a decoction of field horsetail, Chinese cinnamon (Cinnamomum aromaticum), clove (Syzygium aromaticum), common eucalyptus (Eucalyptus globulus), soy lecithin, sunflower oil and guar gum (Cyamopsis tetragonolobus), or a solution based on copper sulfate and plant extracts.

[0130] According to a third aspect, the invention relates to a machine learning model trained to predict a treatment plan adapted to the development conditions of a cultivated SI surface comprising a first natural formulation adapted to promote growth and stimulate the defense system of a plant variety.

[0131] The general principle of the machine learning model trained according to the invention is based on the observation of correlations between the development of a bioaggressor as a function of environmental parameters. Thus, by measuring the effectiveness of the treatment of a plant variety by one or more natural formulations applied, compared to a known chemical pesticide, against a predetermined pest, it is possible to determine a treatment plan. Subsequently, this treatment plan can be used to eliminate the use of the chemical pesticide commonly used against the pest.

[0132] In the invention, the trained machine learning model is a so-called predictive model in which significant correlations are discovered in a set of past observations and in which an attempt is made to generalize these correlations to cases not yet observed. As such, the trained machine learning model according to the invention is distinguished from so-called explanatory models in which an attempt is made to understand the causal mechanism underlying the phenomenon to be predicted. In addition, the trained machine learning model according to the invention is obtained using a so-called supervised learning approach in which past observations are “labeled”. In practice, observations are said to be “labeled” when each of them is accompanied by a label that identifies the phenomenon to be predicted.Training the machine learning model according to the invention requires a plurality of reference cultivated areas, for which environmental parameters are measured, via suitable measuring devices, such as those described previously.

[0133] The system according to the invention being intended to be used by numerous farmers / users and therefore numerous cultivated areas, the quantity of data analyzed represents a relevant source of information for carrying out the continuous improvement of the learning model.

[0134] Generally speaking, the presence or absence of a bioaggressor, following the application of predetermined natural formulations, is crucial information for validating the relevance or for correcting the recommendations generated.

[0135] It is thus possible to propose numerous variants of recommendations for the same plant variety and for similar environmental conditions, with for example earlier or later use of this or that natural formulation.

[0136] In particular, if several less secure treatment data, i.e. which include a limited number of applications of natural formulations, result in the absence of triggering of a disease (absence of bioaggressor) on a given plant variety, it may be considered that these more economical treatment data are sufficient and will be used to generate a recommendation.

[0137] Similarly, if initial processing data results in the presence of a bioaggressor, and second processing data has made it possible to get through the risk episodes without the development of bioaggressors, the learning model can be configured to reference second processing data as the new standard and thus include it when generating a recommendation.

[0138] Still in the invention, the trained machine learning model is obtained according to the following steps: • acquisition, at each of the acquisition times of a plurality of successive acquisition times, of reference parameter values ​​of a plurality of reference cultivated areas, the reference parameters comprising, for each reference cultivated area, at least: • values ​​of parameters of the cultivated area, the parameters of the cultivated area comprising a plant variety, a varietal resistance index associated with the plant variety, a development index of the plant variety, • a value of an environmental parameter, • treatment data comprising at least one other natural formulation selected from a second natural formulation suitable for stimulating the defense system of the plant variety, and / or a third natural formulation suitable for preventing the development of a bioaggressor, and • data on the effectiveness of treatment with at least one other natural formulation.

[0139] In the learning model according to the invention, the successive acquisition times are spaced apart from each other by a predetermined duration, for example 1 hour.

[0140] In the learning model according to the invention, the effectiveness data can indicate the absence of development of a predetermined bioaggressor.

[0141] Alternatively, when the value of the environmental parameter corresponds to the detection of a predetermined bioaggressor, the efficacy data may indicate a degree of efficacy of the treatment compared to a chemical pesticide commonly used against said bioaggressor.

[0142] The trained machine learning model is further obtained according to a step of: • calculation, at each acquisition time, for each of the plurality of reference cultivated areas, of mathematical quantities, from the values ​​of the parameters of the cultivated area, of the value of the associated environmental parameter, acquired at acquisition times included in a predetermined time period preceding the current acquisition time.

[0143] In an example of the calculation step, the mathematical quantities are obtained from a mathematical function chosen from: square root, power, logarithm, exponential, gradient, mean, or a combination of these. However, other mathematical functions may be considered.

[0144] In another example, the predetermined time period may be 4 hours, 1 or more days, or one or more weeks.

[0145] Then, the trained machine learning model is obtained according to the steps of: • creation, for each of the plurality of reference cultivated areas, of a reference mathematical function, for example describing a reference vector, from the parameter values ​​of the cultivated area, the value of the environmental parameter and the mathematical quantities, • association, at each acquisition time, for each of the plurality of reference cultivated areas, of the reference mathematical function, for example describing the reference vector, with the data on the effectiveness of the treatment of the reference cultivated area, so as to obtain machine learning model variables, and • training a machine learning model to predict a treatment plan comprising at least one alternative natural formulation, from a machine learning algorithm and the machine learning model variables.

[0146] In the training step, the training of the machine learning model is carried out by providing the machine learning model with a training data set in the form of pairs (X, Y), in which X corresponds to a set of predictor variables and Y corresponds to a predicted variable. In practice, a pair (X. Y) is defined such that the predictor variable X comprises the characteristic mathematical function, for example the characteristic vector, of the machine learning model variables and the predicted variable Y comprises the treatment data comprising the natural formulation(s) used to treat the reference cultivated area, the first machine learning model variables which are associated with the predetermined time period.

[0147] In one example of the learning step, the machine learning algorithm is based on the Random Forest method. However, other supervised regression machine learning algorithms may also be considered.

[0148] The invention may be the subject of numerous variants and applications other than those described above. In particular, unless otherwise indicated, the various structural and functional characteristics of each of the implementations described above should not be considered as combined and / or closely and / or inextricably linked to each other, but on the contrary as simple juxtapositions. Furthermore, the structural and / or functional characteristics of the different embodiments described above may be the subject in whole or in part of any different juxtaposition or any different combination.

Claims

1. Claims Method for generating (1) a treatment plan adapted to the development conditions of a cultivated surface (SI) comprising a first natural formulation (51) adapted to promote growth and stimulate the defense system of a plant variety, said method being implemented by a computer device (10), the generation method (1) comprising the following steps: loading (100) of parameter values ​​of the cultivated surface, said parameters comprising at least one plant variety, at least one varietal resistance index associated with the plant variety and a development index of the plant variety, loading (200) of at least one value of an environmental parameter, determined by a first measuring device, for a predetermined past time period, of the cultivated surface, a step of loading at least one value of a forecast environmental parameter (250) determined by a second measuring device, for a predetermined future time period, of the cultivated surface from geographical coordinates of said cultivated surface, the value of the forecast environmental parameter comprising a precipitation level, generation (500) of a recommendation (RI, R2) for treatment of the cultivated surface as a function of the value of the environmental parameter,of the value of the forecast environmental parameter and the values ​​of the parameters of the cultivated area, said recommendation (RI, R2) including an application of at least two other natural formulations selected from:, • an application, at regular intervals over a predetermined period, of a second natural formulation (52) adapted to stimulate the defense system of the plant variety, and • an application, at a first predetermined time, of a third natural formulation (53) comprising an antigerminative, and / or • an application, at a second predetermined time, of a third natural formulation (53) comprising an antifungal, • a step of application by foliar or soil route, by a user, of at least two other recommended natural formulations.

2. Generation method (1) according to claim 1, in which the step of loading (200) at least one value of an environmental parameter is followed by the steps of: • determining (300) at least one threshold, for the environmental parameter considered, as a function of the values ​​of the parameters of the cultivated surface, • comparing (400) the value of the environmental parameter with the at least one determined threshold and selecting the threshold corresponding to the value of the environmental parameter, and in which the step of generating (500) the recommendation (RI, R2) for treating the cultivated surface is a function of the selected threshold.

3. Generation method (1) according to one of claims 1 or 2, wherein the parameters of the cultivated surface (SI) further comprise soil parameters, nutrient parameters, prior treatment with a natural formulation and / or a type of irrigation.

4. Generation method (1) according to any one of claims 1 to 3, wherein: • the first natural formulation adapted to promote growth and stimulate the defense system of a plant variety is applied by root application, and • the recommendation of the other natural formulation comprises an application by foliar application.

5. A generation method (1) according to claim 3, the method comprising a step of modifying the generated recommendation depending on the prior treatment with a natural formulation and / or the type of irrigation.

6. A generation method (1) according to any one of claims 1 to 5, wherein the environmental parameter comprises a temperature, a humidity level, a frost risk indicator, a sunshine level, a wind speed, a detection of a pathogenic organism or a detection of a pathological symptom on the plant variety.

7. A generation method (1) according to claim 3, the method comprising a step of calculating (550) a quantity of the at least two other natural formulations to be applied as a function of the prior treatment with the natural formulation, the parameters of the cultivated area and the value of the environmental parameter.

8. A generation method (1) according to any one of claims 1 to 7, wherein: • the first natural formulation comprises microorganisms capable of forming a mutualistic interaction with the plant variety, • the second natural formulation comprises at least one elicitor originating from microorganisms, an insect, a nematode or plant fractions or extracts, and / or • the third natural formulation further comprises at least one active compound having at least one of the following properties: antibacterial, antiviral, insecticidal, insect repellent, or nematicide, and the second natural formulation is applied before the third natural formulation.

9. Generation method (1) according to any one of claims 3 to 8, wherein the parameters of the cultivated surface (SI) comprise pedological and / or nutrient parameters, said method further comprising a step of comparing (150) at least one value of a pedological and / or nutrient parameter to a predetermined threshold and generating a recommendation for adding an input when the value of the pedological and / or nutrient parameter is not equal to the predetermined threshold.

Images