FR3075565A1 - METHOD FOR DIAGNOSING AND MONITORING A TARGET ELEMENT OF AN AGRICULTURAL FARM - Google Patents

Abstract

System (1) for monitoring and alerting a target natural element (20) of a farm, comprising a diagnostic platform (2) of the target element (20) comprising a plurality of sensors (3) ) and a module (4) for recording and analyzing the parameters, a pointer module (6) adapted to receive an input indicating that a target state of the target element (20) is in progress or on the not occur, a monitoring platform (10) of said target element comprising a plurality of monitoring sensors (13) and a monitoring module (11) designed to monitor the data in relation to the physical conditions related to the evolution. said target element (20) and / or its environment (21).

Description

Method for diagnosing and monitoring a target element of a farm

TECHNICAL FIELD OF THE INVENTION

The present invention relates to a monitoring and alert system for a target natural element of a farm, comprising a diagnostic platform of the target element and a monitoring platform.

STATE OF THE PRIOR ART

Despite the ever-increasing level of complexity of agricultural machinery, operating a farm or agricultural business remains a complex and demanding task. The operator is particularly stressed by a large number of parameters and events to monitor and control so that he can make the right decisions in relation to the daily tasks to be performed. Surveillance obliges him to continually move from plot to plot in order to fully understand the precise situation at each key point of the operation. Farms are becoming larger and often grouping several plots often far apart, the farmer spends much of his time traveling for the purpose of informing himself. These information phases, while essential for effective management and organization, are as such unproductive and time consuming.

[0003] In addition, the operator must regularly make decisions which are often fraught with consequences, both on the quality of the resulting products and on the management of time and tasks at the operational level. The farmer or breeder must be able to prioritize their decisions. However, very few effective tools still exist today for this purpose.

[0004] For example, diagnostic methods are known which allow the farmer to more easily detect the state of health of plants or certain infections which can affect crops.

For example, document FR2663192 describes a method for the early diagnosis of infection of cereal seedlings by a fungus, in particular Pseudocercosporella herpotrichoïdes, and a device specifically designed for the implementation of this method. The method consists in forcing the young cereal plants in a controlled atmosphere under conditions which ensure simultaneous growth of the fungus and the parasitized plant. The process described applies to the field of plant protection and more particularly to cereal farming. This method seems effective but is entirely manual. On the other hand, it requires prior knowledge of the phenomena sought.

Document FR2970621 describes a method which consists of analyzing branches in a vineyard sample by determining the concentration of macro-elements and trace elements. An equilibrium relationship is determined between the concentration of the macro-elements and the values of the trace-elements. The data obtained for juices extracted from conductive tissues and for branches at reference values for a given type of vine are compared. A statement of the corrections to be provided to maintain the nutritional status of the vines in the vineyard is produced. This method seems effective, but is limited to establishing a diagnosis.

Also known are soil moisture detection systems, intended to help the farmer who must decide when and how to water his crops.

Document US2017038749 describes a precision agriculture irrigation modeling framework using a combination of meteorological data, harvest data and other agricultural inputs to create personalized agronomic models to diagnose and predict the state of humidity of a field, as well as the need for irrigation. Specific combinations of various agricultural inputs can be applied, as well as meteorological information to identify or adjust the water characteristics of crops and soils, to model optimal irrigation activities and to provide advice, recommendations and guidelines for application targeted artificial precipitation. This system is particularly oriented towards the diagnostic and recommendation phases, only in the field of agricultural irrigation.

Also known are devices or methods for simplifying the use of certain agricultural machines.

For example, the document FR2764401 describes a method for automating recurring tasks to be implemented with an agricultural machine such as a tractor. The method is based on the memorization of the actions activated by the operator, and the subsequent re-reading of the control signals which activate these operations. The memorization can be carried out in advance by the vehicle manufacturer. The actions are classified into primary and secondary actions. The relative quantities of the actions are memorized with the state of the vehicle associated with the actions. Non-representative actions can be filtered. Secondary actions can be synchronized. Stored actions can be viewed and edited, and actions can be interrupted during execution and repeated. A computer is used to memorize the actions and to control the operation and the display. This process is directly related to the machine to which it is attached, makes it possible to record and repeat trivial actions, but without making it possible to establish a diagnosis or analyze the tasks performed.

In all these examples, no document discloses a solution making it possible to establish a precise and exhaustive diagnosis of an element based on the values of the physical parameters making it possible to characterize this element during its evolution and more particularly at when this system goes to a particular state. Furthermore, no system allows simple and reliable monitoring of certain elements of a farm in order to detect the occurrence of one or more specific conditions.

To overcome these different drawbacks, the invention provides different technical means.

STATEMENT OF THE INVENTION

A first objective of the invention consists in making it possible to establish a precise and exhaustive diagnosis comprising all of the parametric values relating to an element to be monitored during its evolution and more particularly when this element changes to a particular state, this particular state being useful for carrying out a task of the agricultural exploitation.

Another object of the invention is to provide a method for monitoring the state of an element in order to identify and / or anticipate the occurrence of this particular state using parameters that are easy to measure and monitor. .

Another objective of the invention is to provide a method for supplying a farmer, automatically and over time, with state data for one or more elements of his farm.

MONITORING AND ALERT SYSTEM

To do this, the invention provides a monitoring and alert system of a target element of a farm, comprising: i) a diagnostic platform of the target element comprising a plurality of sensors diagnostic and a parameters recording and analysis module designed to accumulate, for a limited period corresponding to a diagnostic phase of said target element, data in relation to the physical conditions linked to the evolution of said target element; ii) a pointer module (or “flag”) designed to receive an input indicating that a target state of the target element is in progress or about to occur and a data identification module corresponding to said target state of the target element; ii) a platform for monitoring said target element comprising a reduced number of monitoring detectors and a monitoring module designed to monitor for a period corresponding to a monitoring phase of said target element the data related to the physical conditions linked to the evolution of said target element and / or its environment, the monitoring platform further comprising an alert module designed to generate an alert condition when the parametric data obtained by the monitoring platform converge to the data of said state target identified by the pointer module.

Such a system makes it possible, during a diagnostic phase, to acquire a wide range of physical state data of a system. The system provides a complete, efficient diagnostic platform which makes it possible to make a very high number of data captures, to make it possible to ensure that the exact values corresponding to the phenomenon to be monitored are well captured and taken into account. The diagnostic platform is designed with comprehensiveness and accuracy criteria prevailing over the energy saving, autonomy or sustainability criteria, which are priority at the level of the monitoring platform.

These data acquired by the diagnostic platform make it possible to acquire a knowledge of the physical comprising of the target element, and more particularly a detailed and exhaustive knowledge of the physical state data as a function of the evolution of the target item.

The pointer module makes it possible to establish a precise and reliable link between the occurrence of the phenomenon sought or of the event to be monitored and the physical state values of the target element at this precise instant, and possibly a little before and / or after.

The data in relation to the pointer makes it possible to determine a surveillance approach different from that conventionally used (generally manual and approximate) for carrying out surveillance of the target element. The user is thus able to set up a reliable, precise and durable automatic surveillance, using a simple surveillance platform and configured to operate over time in hostile environments, under variable conditions and often difficult. The monitoring phase is thus based on a reduced number of parameters, corresponding to the most significant parameters of the target state sought.

Advantageously, the number of detectors of the monitoring platform is at least 50% lower and more preferably at least 70% less than the number of sensors on the diagnostic platform.

This large difference between the number of sensors for diagnosis and the number of detectors for monitoring allows on the one hand to improve the reliability of the diagnostic phase with the taking into account of a large number of parameters, without seeking to eliminate at this stage, not knowing which will be the most significant, and on the other hand to greatly simplify and lighten the implementation of monitoring, with a simple, inexpensive installation, suitable for carrying out monitoring over time.

In an advantageous embodiment, the parameters are selected from: pressure, temperature, humidity, wind speed, brightness, UV index, mass, force, torque, current , a voltage, a magnetic flux, a flow, an image (with image recognition), a sound level, a color, a reflectivity, a concentration of a gas or other element, an impedance (in particular on the ground).

According to an advantageous embodiment, the pointer module is independent of the diagnostic platform and adapted to communicate remotely with the latter.

This embodiment makes it possible to set up a pointer module (or “flag”) at any location, preferably to facilitate the occurrence of the target event, while the diagnostic platform is advantageously placed at the 'shelter, in a place facilitating the work of an operator.

Advantageously, the diagnostic platform and the pointer module each include a clock adapted for the time stamping of the data, and a clock synchronization module is provided (either at the level of the diagnostic platform and / or at the pointer module).

This implementation ensures that the time stamping data of the pointer module and the sensors of the diagnostic platform are perfectly synchronized, in order to ensure that the right data is used during the pointing step.

In a version comprising an artificial intelligence module, the diagnostic platform includes a self-learning module (for example based on artificial intelligence), designed to automatically identify the data of specific physical parameters in which the system to be monitored is likely to enter the specific state mode to be monitored.

DIAGNOSTIC PROCESS AND MONITORING

The invention also provides a method for diagnosing and monitoring a target element of a farm using a system and monitoring as previously described, comprising the following steps: PHASE 1 i) setting in place of a diagnostic platform, said platform comprising a plurality of diagnostic sensors cooperating with said target element; ii) acquisition and recording of diagnostic data of said target element using the plurality of sensors; PHASE 2 iii) in the event of the occurrence of a target state of said target element, entry of identification data for said target state; iv) identification of the parameters and threshold values corresponding to said target state; PHASE 3 v) configuration of a monitoring platform based on the parameters and threshold values corresponding to said target state; vi) setting up the configured monitoring platform, said platform comprising a reduced number of monitoring detectors cooperating with said target element; vii) monitoring of said target element by the monitoring platform; viii) in the event of convergence between the parametric data obtained by monitoring and the threshold values corresponding to the target state, issuance of an alert.

The three-phase process first allows a complete and exhaustive acquisition of a large amount of data, able to identify the parameters of a specific state of the target element of a farm. It then makes it possible to correlate, using a pointer (or “flag”), a specific state sought or monitored with the values corresponding to this state for each of the parameters. Finally, it makes it possible to set up a monitoring system, simple, reliable, and efficient, strictly based on the specific state parameters identified.

Once the specific state parameters have been identified, the method makes it possible to monitor the target element and anticipate a situation in which the specific state is highly likely to occur again. The farmer thus has technology enabling him to be alerted in order to anticipate certain specific states for which urgent actions are required. The monitoring phase is based on a reduced number of parameters, corresponding to the most significant parameters of the target state sought.

These parametric data are certain and precise, which makes it possible to have reliable monitoring.

In an alternative embodiment, the step of identifying the data of the parameters corresponding to said target state allows the identification of pre-alert data.

The pre-alert allows the operator to be warned early enough to allow intervention prior to the occurrence of the specific condition.

In an advantageous embodiment, the diagnostic and monitoring method also comprises a step of synchronizing the clocks.

This step makes it possible for the timestamps of the data of the diagnostic platform and of the pointer data to be perfectly synchronized, thus allowing the pointing step to give rise to the identification of the state data. closest to the target state.

According to another advantageous embodiment, the diagnostic method also provides at least one post-alert monitoring step. For example, post-alert surveillance includes monitoring the application of curative treatment. This step ensures that the treatment provided in the event of an alert is applied appropriately. In the case of a spreading of curative product for example, detectors previously configured during the configuration phase make it possible to monitor the quantity of product which reaches the area to be treated, to indicate whether the level of product is sufficient or not, or even applied in excess.

According to yet another advantageous embodiment, post-alert monitoring includes monitoring a target state corresponding to healthy cultures. Thus, when the situation has returned to a normal state, that is to say with healthy cultures, the operator is informed and can for example stop curative treatments.

DESCRIPTION OF THE FIGURES

All the details of embodiment are given in the description which follows, supplemented by FIGS. 1 to 4, presented only for the purposes of nonlimiting examples, and in which: FIG. 1 is a schematic representation of a example of a surveillance and alert system for one or more target elements of a farm; FIG. 2 is a functional flow diagram with the main steps of an example of monitoring and alerting process for a target element of a farm; FIG. 3a is an example of a diagnostic and monitoring table for an example of a target element corresponding to a measurement of the level of humidity of crop grains; FIG. 3b is a variant of the table in FIG. 3a in which only the relevant data have been kept in order to simplify or lighten the monitoring phase; FIG. 4 is an example of a diagnostic and monitoring table for an example of a target element corresponding to the lift of the ground.

DETAILED DESCRIPTION OF THE INVENTION

DEFINITIONS

By "target element" or "target natural element" means an element related to a crop or a breeding carried out on a farm. It can be for example a cereal, a fruit, a vegetable, farm animals or used for the production of milk or eggs. It can also be a part of one of these elements, such as for example a grain of a cereal, a leaf of a tree, etc. The target natural element also includes environmental elements in which the object of exploitation evolves, such as soil, water, air, light, wind, etc. These examples are given by way of illustration and not by way of limitation. These can be essential factors in farming. They represent in certain cases the very object of exploitation (for example the production of cereal grains), or, in other cases, harmful elements such as parasites or diseases which one wishes to be able to detect at most early. The monitoring of these elements is conventionally carried out by the operators, generally manually. The elements that are the subject of the operation are monitored in order to determine, for example, a level of maturity for harvesting or harvesting. The elements of the environment are monitored in order to adjust the parameters of the operation, such as for example watering, the addition of fertilizers, etc., or in order to monitor the quality of a vector of the operation, such as for example the hardness of the soil.

By "sensor" means a device for transforming a physical quantity into a quantity that can be associated with a unit of measurement.

By "detector" means a device for monitoring that a physical quantity approaches or reaches a threshold corresponding to a determined value of a parameter to be monitored.

DIAGNOSTIC AND MONITORING SYSTEM

Figure 1 is a schematic representation of an example of a system 1 for monitoring and alerting a target element 20 of a farm. In this example, the monitoring and alerting system 1 comprises a diagnostic platform 2 comprising a plurality of sensors 3 connected to a recording and analysis module 4. Finally, a database 5 of the usage parameters of the target element makes it possible to keep the data received during the diagnostic phase.

A pointer module 6 allows a user (or a monitoring device for example) to enter a pointer (or "flag") when a target event occurs. Several pointers 6 may possibly be entered if the user (or other monitoring module) is able to determine the moments preceding the occurrence of the target event. As a variant, the pointer 6 is provided outside the diagnostic platform, at a distance from the latter. This module 6 makes it possible to precisely identify the values supplied by the various sensors 3 when the target state occurs and / or is in progress. Each parameter is pointed and / or marked in order to recognize the moment when the target state takes place. Optionally or alternatively, several types of pointers (or “flags”) are provided, for example a type for a target event, a type for normal operation, a type for an unforeseen event, etc.

A module 7 for identifying the data corresponding to the target state (that pointed to by the pointer module) is also provided, in relation to a database 8 of parameters of the target state. Alternatively, databases 5 and 8 can be arranged in a single database.

Optionally, the diagnostic platform 2 may include a self-learning module 9, intended either to assist an operator during a diagnostic phase, for example by proposing ranges of evolution where an event target could potentially arise, or to allow an automatic diagnosis to be made.

The diagnostic platform 2 is designed so that a user can easily set up a wide range of sensors, while promoting recordings of a very large amount of data. Since before the diagnostic phase, the operator does not know the parameters that are most relevant, nor the values of these parameters, the architecture of the diagnostic platform allows the use of different types of sensors and record many values for all of these sensors.

Once the diagnostic phase has been carried out, the operator will know the most relevant parameters and will therefore be able to recommend a specific, more targeted architecture for the monitoring phase. The monitoring platform, which often has fewer different types of detectors, is therefore less expensive. This platform is designed to promote low energy consumption, less frequent data transmission, and the implementation of optimized technological elements for an efficient and reliable monitoring installation over time.

Thus, as illustrated in the example of Figure 1, a monitoring platform 10 includes a plurality of detectors 13 connected to a monitoring module 11, in relation to the data 12 of the parameters of the target state . The number of detectors 13 of the monitoring platform is at least 50% lower and more preferably at least 70% less than the number of sensors 3 on the diagnostic platform. An alert module 14 makes it possible to generate a signal or alert when the actual observed data converge or are in correspondence with the data of the target state.

The alert is intended to notify the farmer, and / or another person or entity, of the next or current occurrence of the target state according to the settings previously made.

DIAGNOSTIC AND SURVEILLANCE PROCESS

FIG. 2 illustrates, using a functional flowchart, an example of a monitoring and alerting method advantageously suitable for implementation with a diagnostic and monitoring system as previously described in relation to the figure 1.

In the example illustrated, the process is subdivided into three phases.

First, in step 100, a diagnostic platform 2 is set up, with installation of the various sensors, for example of humidity, temperature, wind speed, etc. . The chosen and installed sensors are related to the target element 20 and / or its environment, depending on the parameters to be measured. In this way, the data collected is relevant to the target element. However, as previously indicated, since the most relevant parameters are unknown or uncertain before carrying out the diagnosis, the operator advantageously provides for a greater number of sensors, to avoid neglecting a possible parameter which could prove to be relevant for perform effective monitoring.

In step 101, data acquisition from the sensors is carried out. The data captured and recorded correspond to the evolution of the target element and its environment over time. The state data is preferably saved before the target state sought, during this state and, if necessary, after this state.

The second phase of the process corresponds to the marking of a target state. The first step (step 103) is the referencing of the target state, when it occurs, using the pointer module 5. This step corresponds to the entry of an identification datum (or “flag”) of the target state sought. At the precise moment of the occurrence of this state and / or when it is in progress, each of the data of the various sensors is recorded and marked in order to distinguish the pointed data from the other data recorded before the occurrence of the target state and after that -this. This marking makes it possible to obtain (in step 104) all of the precise values of the data relating to the occurrence of the target state. The operator, or a selection module, can then identify the significant parameters and the threshold values corresponding to the target state sought. It is advantageous to select the parameter or parameters which make it possible to discriminate the target state with respect to all of the available parameters and the target values of these parameters indicating the target state.

The third phase of the process is that of monitoring the target element 20 by means of the monitoring platform 10.

In step 105, a monitoring platform 10 is configured, based on the parameters and threshold values previously identified in step 104. In step 106, the monitoring platform 10 configured is set up in relation to the element to be monitored 20 and / or its environment, according to the parameters identified in step 104.

The monitoring can then be set up, in step 107. The monitoring platform 10 makes it possible to monitor the evolution of the parameters measured by the detectors 13 set up to determine when the target state risks occur or actually occurs.

If there is convergence between the data collected during monitoring and the threshold values, an alert intended to warn the operator is issued in step 108. This alert can be issued sooner or later depending on the setting. performed upstream. In certain cases where this may prove useful, and as far as possible, the selected threshold values allow the issue of an alert sufficiently in advance to allow the operator to anticipate the occurrence as well as possible. of the target state sought.

SYNCHRONIZATION OF CLOCKS

The method includes an optional step indicated in step 102. This step is useful and implemented for example in the case where one or more pointer modules 6 are independent of the diagnostic platform 2 and are likely to have a or several clocks with values different from those of the diagnostic platform. A clock synchronization step is then carried out using the clock synchronization module 15. One of the clocks is identified as a reference, for example the platform clock or an external clock whose value is reliable and precise. The other clock (s) are then synchronized with the reference clock.

SELF-LEARNING MODULE

Alternatively, to determine when the target element is likely to change to a target state, the diagnostic platform may include a self-learning module 9. This module is designed to automatically identify the parameter data specific physical when the target element approaches a target state. This module can be based on artificial intelligence. It preferably includes a link to a database of target elements and target states facilitating the analysis of convergence to one of these states. For example, a self-learning module can compare evolution ranges of data obtained for an element under diagnosis in order to compare these evolutionary ranges with similar ranges likely to be found in the database. Any strong similarities between longer and longer ranges can be used as indicators to automatically determine target state data. The module can also operate semi-automatically, generating proposals for a user who may or may not accept them, for example based on his own observations.

EXAMPLES OF APPLICATIONS EXAMPLE 1 FIG. 3a is a first example of a referencing table of data acquired in the world of cereal farming, with the aim of establishing a diagnosis making it possible to identify in a simple and reliable manner. the grain moisture level corresponding to an optimal time for cutting plants and harvesting grains. Conventionally, the verification of the humidity level is carried out manually by taking a plurality of grain samples for the purpose of taking measurements in the laboratory or in an industrial environment, with measurement equipment suitable for carrying out measurements. internal to the grain. This approach is very time-consuming and requires the operator to make numerous trips to various areas of the operation where harvests are about to be completed. The implementation of automated monitoring, based on reliable parameters, defined during a diagnostic phase, allows not only better productivity, but also more reliable and more precise detection of the best time to harvest. The operator can also organize the harvests of several plots optimally based on the data he automatically obtains for each of the plots.

The first part of the table illustrates an example of data corresponding to a diagnosis of the moisture level of a grain. Not knowing, before the diagnosis, which are the significant parameter (s), the user sets up a complete and exhaustive diagnostic platform, for taking into account a large number of parameters likely to prove significant , either alone or with matches. In this diagnostic example, sensors measure air temperature and humidity, wind speed, light and precipitation.

An example of data pointing is presented in the table presented as an example of principle. The analysis of the values of the parameters by comparing, for example, the values pointed to those in the vicinity of these, makes it possible to determine the most significant parameters, corresponding for example to those for which the differences in values are the largest.

The second part of the table illustrates an example of parameters retained for the implementation of automatic monitoring. In this example, all parameters are retained, and the relevant values are used to target the monitoring to be performed.

FIG. 3b is a variant of the table in FIG. 3a in which the step of identifying significant parameters has made it possible to retain a reduced number of parameters, thus making it possible to simplify and lighten the monitoring phase, which can be implemented with a reduced number of detectors and parameters to be monitored. In this example, the parameter corresponding to the brightness was judged to be insignificant and not retained in the monitoring phase. EXAMPLE 2 The table of FIG. 4 illustrates another example of application aiming to establish a diagnosis of the significant parameters to determine in a simple, reliable and precise manner the level of lift of the ground. This diagnosis is important because a minimum level of lift is required to allow the various equipment to be able to move on the soil of the farm without damaging the crops in place and / or risk sinking into the soil. Conventionally, this parameter is determined by an operator who must move in order to go walking on the ground to test the level of noticeable hardness. He can also use a penetrometer to take precise measurements, but the latter approach is particularly time-consuming and therefore rarely used in practice. In this application, the establishment of an automatic monitoring mode with reliable parameters, without the operator having to intervene, allows significant productivity gains, and makes it possible to avoid incidents in cases where an operator would circulate on a floor without first checking the level of hardness and / or lift.

The first part of the table illustrates an example of data corresponding to a diagnosis of the bearing capacity of the soil. Not knowing, before the diagnosis, which are the significant parameter (s), the user sets up a complete and exhaustive diagnostic platform, for taking into account a large number of parameters likely to prove significant , either alone or with matches. In this diagnostic example, sensors measure the humidity and temperature of the soil at various depths, air temperature and humidity, infrared radiation and wind speed.

In the example illustrated in FIG. 4, six parameters are noted: the humidity of the soil and the subsoil, the temperature of the soil and the subsoil, the temperature of the air at a distance x from the ground, air humidity at a distance x from the ground, infrared radiation and wind speed.

Several sensors are installed to collect data related to the humidity and temperature of the soil and subsoil. The sensors are placed at a distance of 5cm below ground level, then at 10cm below ground level, then from 5cm in 5cm until reaching 30cm below ground level. The line marked with a flag corresponds to the values pointed during the pointing phase, when a target state is observed or detected.

The second part of the table illustrates an example of parameters retained for the implementation of automatic monitoring. In this example, the measurements made it possible, for example, to identify a level of soil depth giving more relevant results. The other levels of measures are therefore not used for the monitoring phase, in order to make the latter as simple as possible to set up and manage.

As illustrated, the monitoring phase is based on a reduced number of parameters, corresponding to the most significant parameters of the target state sought. In this example, the value of a single sensor, located at a certain distance from the ground, is used for the data on soil humidity and the data on soil temperature. In this example, the raised sensor is the one placed 10cm below ground level. EXAMPLE 3 The table of FIG. 5 illustrates another example of application aiming to establish a diagnosis of the significant parameters to determine in a simple, reliable and precise manner the appearance of a fungus such as yellow rust or brown rust on cereal crops such as wheat. This diagnosis, carried out at the earliest possible stage of the appearance of the fungus, makes it possible to quickly carry out targeted treatment to prevent all of the cultures from being invaded by a massive attack.

Conventionally, the risk associated with fungi is treated in two ways. Most often, fungicide treatments are applied to a large part of the crops. This approach, not very ecological and expensive, involves a great consumption of products, and monopolizes the operator for long periods to make the spreading. When this approach is not sufficient, or has not been carried out and fungi have appeared on crops, the operator then applies a curative treatment, also expensive, not very ecological, and time-consuming.

The first part of the table illustrates an example of data corresponding to the earliest possible diagnosis of the appearance of the first fungi. Not knowing, before the diagnosis, which are the significant parameter (s), the user sets up a complete and exhaustive diagnostic platform, for taking into account a large number of parameters likely to prove significant , either alone or with matches. In this diagnostic example, as shown in Figure 5, sensors measure the level of humidity and temperature of the ground, the temperature and humidity of the air, the infrared radiation and the luminosity, the evolution of the color of a given area and / or surveillance by imagery and image processing, reflectivity, impedance and current at ground level.

The line marked with a flag corresponds to the values pointed during the pointing phase, when a target state is observed or detected, in this case when the first signs of fungi or the first fungi appear.

The second part of the table illustrates an example of parameters retained for the implementation of automatic monitoring. As illustrated, the monitoring phase is based on a reduced number of parameters, corresponding to the most significant parameters of the target state sought.

As a variant, the alert phase is supplemented by a phase for monitoring the treatment applied. For example, the surveillance detectors also include detectors capable of measuring the implementation of the post-detection treatment, in this case a treatment for spreading fungicide and / or other product making it possible to return to a normal situation, with healthy cultures. . For example, capacitive sensors are provided to detect if the fungicide is applied sufficiently, insufficiently, or in excess.

Still in a variant, the monitoring sensors can comprise sensors capable of monitoring the effectiveness of the treatment applied and / or the return to a normal situation, with healthy cultures.

In this example, but also in general, the sensors and / or detectors preferably include geolocation data in order to allow the operator to identify the plot or plots concerned. Detectors and sensors on each plot, or each representative plot, are therefore expected, both during diagnostics and for monitoring, because the results, in particular the results of the monitoring phases, often vary from one plot to another.

Again as a variant, the input data, in particular for the diagnostic phases, comprise data from sources external to the agricultural holding, such as for example weather data from a third-party supplier, and / or data various models, such as data from meteorological, pluviometric, epidemiological or other models.

Reference numbers used in the figures

Claims (13)

1. System (1) for monitoring and alerting a target natural element (20) of a farm, comprising: i) a diagnostic platform (2) of the target element (20) comprising a plurality of diagnostic sensors (3) and a module (4) for recording and analyzing parameters designed to accumulate for a limited period corresponding to a diagnostic phase of said target element (20) of data related to physical conditions linked to the evolution of said target element (20); ii) a pointer module (6) designed to receive an input indicating that a target state of the target element (20) is in progress or about to occur, and a module (7) for identifying the data corresponding to said target state of the target element (20); iii) a monitoring platform (10) of said target element comprising a reduced number of monitoring detectors (13) and a monitoring module (11) designed to monitor for a period corresponding to a monitoring phase of said target element (20 ) the data in relation to the physical conditions linked to the evolution of said target element (20) and / or of its environment (21), the monitoring platform (10) further comprising an alert module (14) designed to generate an alert condition when the parametric data obtained by the monitoring platform (10) converge on the data of said target state dentified by the pointer module (6). 2. Monitoring and alert system according to claim 1, in which the number of detectors (13) of the monitoring platform (10) is less than at least 50% and more preferably at least 70% than the number of sensors (3) of the diagnostic platform (2). 3. Monitoring and alert system according to one of claims 1 or 2, in which the parameters are selected from: pressure, temperature, humidity, wind speed, brightness, UV index, a mass, a force, a couple, a current, a tension, a magnetic flux, a flow, an image, a sound level, a color, a reflectivity, a concentration of a gas or other element, an impedance, a level sound. 4. Monitoring and alert system according to any one of the preceding claims, in which the pointer module (6) is independent of the diagnostic platform (2) and adapted to communicate remotely with the latter. 5. The monitoring and alert system as claimed in claim 4, in which the diagnostic platform (2) and the pointer module (6) each comprise a clock adapted for time stamping of the data, and in which a module for synchronization (15) of the clocks is provided. 6. Monitoring and alert system according to any one of the preceding claims, in which the diagnostic platform (2) comprises a self-learning module (9) designed to automatically identify the data of specific physical parameters wherein said target element is capable of entering the target mode. 7. A method of diagnosing and monitoring a target natural element (20) of a farm using a system (1) and monitoring and alert according to one of claims 1 to 4, comprising the following steps: PHASE 1 i) setting up (100) of a diagnostic platform (2), said platform (2) comprising a reduced number of diagnostic sensors (3) cooperating with said target natural element (20) or its environment (21); ii) acquisition and recording (101) of diagnostic data of the system using the plurality of sensors (3); PHASE 2 iii) in the event of the occurrence of a target state of said target element (20), entry (103) of an identification datum of said target state of said target element (20); iv) identifying (104) the parameters and threshold values corresponding to said target state; PHASE 3 v) configuration (105) of a monitoring platform (10) based on the parameters and threshold values corresponding to said target state; vi) positioning (106) of the configured monitoring platform (10), said platform comprising a plurality of monitoring detectors (13) cooperating with said target element (20) or its environment (21); vii) monitoring (107) of said target element (20) and / or of its environment by the monitoring platform (10); viii) in the event of convergence between the parametric data obtained by monitoring and the threshold values of said target state, issuance (108) of an alert. 8. A diagnostic and monitoring method according to claim 7, in which the parametric data are selected from: pressure, temperature, humidity, wind speed, luminosity, UV index, mass, force , a torque, a current, a voltage, a magnetic flux, a flow, an image and a sound level 9. The diagnostic and monitoring method according to claim 7, in which the step of identifying the data of the parameters corresponding to said target state allows the identification of pre-alert data. 10. A diagnostic and monitoring method according to any one of claims 7 to 9, further comprising a step of synchronizing the clocks. 11. A diagnostic and monitoring method according to any one of claims 7 to 10, further comprising at least one step of post-alert monitoring. 12. The diagnostic and monitoring method as claimed in claim 11, in which the post-alert monitoring comprises monitoring the application of a curative treatment. 13. The diagnostic and monitoring method according to claim 11, wherein the post-alert monitoring comprises monitoring a target state corresponding to healthy cultures.