Cultivation Management System

Abstract

In a cultivation management system 1 for managing open-field crop cultivation, the future growth trajectory of crops can be estimated. [Solution] The cultivation management system 1 of the present invention comprises a state data acquisition unit 111 that acquires state data A1, a growth distribution calculation unit 112 that calculates crop growth distribution data A2, a disease and pest distribution calculation unit 113 that calculates crop disease and pest distribution data A3, a first growth data calculation unit 114 that calculates first growth data A5, and a growth state evaluation unit 115 that evaluates the quality of the first growth data A5. The state data A1 includes at least time-series imaging data Aa of the crop captured by an imaging device 31 mounted on an unmanned mobile body 30, fertilization data Ab, pesticide application data Ac, crop variety data Ad, and crop sowing time data Ae.

Description

Technical Field

[0006] ,

[0001] The present invention relates to a cultivation management system for managing open-field cultivation of crops.

Background Art

[0002] In open-field cultivation of crops, it is common to create a cultivation work plan from sowing to harvesting approximately. For example, in Patent Document 1, color values are extracted from image data of plants photographed with a digital camera, a color index value that is an index of the growth state of the plants is calculated from the color values, and a technique for objectively determining the growth status of the plants from the color index value is disclosed.

[0003] Also, in Patent Document 2, in a farming management system for managing a farm field divided into a plurality of small plots, the field environment information that is superimposed and displayed on a monitor together with the results of past events (work processes) includes weather information obtained in a section unit larger than the small plot unit (for example, the farm field unit).

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Patent Document 2

Disclosure of the Invention

Problems to be Solved by the Invention

[0005] However, since the technique of Patent Document 1 is applied to greenhouse cultivation where environmental management is easy, there is a problem that it is difficult to apply it to open-field cultivation where environmental changes are more drastic than in greenhouse cultivation.

[0006] Furthermore, the technology described in Patent Document 2 manages cultivation work on an event-by-event basis (work process) basis, that is, on a long-term basis such as a process unit, and does not manage on a finer time scale such as a day or hour. Therefore, there was a problem in that it was extremely difficult to allocate agricultural machinery, personnel or materials, or to schedule the automatic operation of agricultural machinery, based on the period management data from Patent Document 2.

[0007] Furthermore, in farm management, tasks are often carried out by multiple personnel, and farm machinery can be operated both manned and unmanned. Therefore, the combination and allocation of farm machinery, personnel, and materials are of great importance. However, Patent Document 2 does not disclose or suggest anything about the relationship between each event and the combination and allocation of farm machinery, personnel, and materials. [Means for solving the problem]

[0008] In view of the above problems, the present invention aims to provide a cultivation management system for managing open-field cultivation of crops that can estimate the future growth trajectory of crops and ultimately create a cultivation work plan for the crops.

[0009] The present invention relates to a cultivation management system for managing open-field cultivation of crops, comprising: a state data acquisition unit that acquires state data related to the growth state of the crop; a growth distribution calculation unit that calculates growth distribution data of the crop based on the state data; a pest and disease distribution calculation unit that calculates pest and disease distribution data of the crop based on the state data; a first growth data calculation unit that calculates first growth data representing the growth state of the crop based on the growth distribution data and the pest and disease distribution data; and a growth state evaluation unit that evaluates the quality of the first growth data by comparing it with a preset first reference data, wherein the state data includes at least time-series imaging data of the crop captured by an imaging device mounted on an unmanned mobile body, fertilization data for each unit field where the crop is cultivated in the open field, pesticide application data for each unit field, crop variety data, and crop sowing time data.

[0010] The cultivation management system of the present invention includes an environmental data acquisition unit that acquires environmental data related to the growth environment of the crop, an accumulated temperature calculation unit that calculates the accumulated temperature during the growth period of the crop based on the environmental data, an accumulated solar radiation calculation unit that calculates the accumulated solar radiation during the growth period of the crop based on the environmental data, a second growth data calculation unit that calculates second growth data representing the growth environment of the crop based on the accumulated temperature and the accumulated solar radiation, and a growth environment evaluation unit that evaluates the quality of the second growth data by comparing it with a preset second reference data, wherein the environmental data may include at least time-series weather data for a predetermined area where the unit field is located, time-series local weather data measured at the unit field, and field characteristic data related to the characteristics of the unit field.

[0011] The cultivation management system of the present invention may include a growth estimation unit that estimates the future growth trajectory of the crop from the growth data using a growth learning model.

[0012] The cultivation management system of the present invention may include a growth estimation unit that estimates the future growth trajectory of the crop from the first and second growth data using a growth learning model.

[0013] The cultivation management system of the present invention comprises a cultivation work data acquisition unit that acquires cultivation work data related to the cultivation management of the crop, and a cultivation work plan creation unit that creates a cultivation work plan for the crop from the cultivation work data using a planned learning model, wherein the cultivation work data includes at least schedule data, farm machinery data, personnel data and material data, and the cultivation work plan creation unit may create the cultivation work plan by allocating schedules, farm machinery, personnel and materials to each unit field based on a time axis. [Effects of the Invention]

[0014] According to the present invention, using state data A1 related to growth status and environmental data B1 related to growth environment, future growth transition data C1 of crops can be predicted in time series (for example, every hour, every day, or every week, etc.), and finally, a cultivation work plan D2 for crops can be created based on the time axis, enabling efficient operation of the entire cultivation work.

Brief Description of the Drawings

[0015] [Figure 1] It is an explanatory diagram showing the outline of the cultivation management system of the first embodiment. [Figure 2] It is a conceptual diagram showing the functional configuration of the cultivation management system. [Figure 3] It is a conceptual diagram showing the hardware configuration of the cultivation management system. [Figure 4] It is a conceptual diagram of cultivation work plan generation. [Figure 5] It is an explanatory diagram showing the outline of the cultivation management system of the second embodiment. [Figure 6] It is an explanatory diagram of an unmanned vehicle with a vinyl house specification, (a) is a side view, and (b) is a rear view. [Figure 7] It is an explanatory diagram of an unmanned vehicle with a plant factory specification, (a) is a side view, and (b) is a rear view. [Figure 8] It is an explanatory diagram of a furrow.

Modes for Carrying Out the Invention

[0016] Hereinafter, embodiments embodying the present invention will be described based on the drawings. Although the drawings show preferred embodiments, it can be implemented in many different forms and is not limited to the embodiments described in this specification.

[0017] Referring to FIGS. 1 to 4, the cultivation management system 1 of the first embodiment will be described. FIG. 1 is an explanatory diagram showing an overview of the cultivation management system 1 (hereinafter sometimes simply referred to as the system) of the first embodiment. The system 1 collects and manages data related to the open-field cultivation of crops in a predetermined area and provides various support services to users. There are various cultivation operations, such as irrigation, fertilization (top dressing), guiding, leaf pruning, flower pruning, fruit pruning, and pesticide application. Since the cultivation operations to be carried out may vary depending on the type of crop, the cultivation operations to be targeted can be appropriately designed according to the scope of application.

[0018] As shown in FIG. 1, the system 1 includes a management device 10, an external terminal 20, and an unmanned aerial vehicle 30 such as a drone. The management device 10 is a server or the like for managing and supporting the open-field cultivation of crops. For example, it can evaluate the quality of the growth state of crops, evaluate the quality of the growth environment of crops, estimate the future growth trend of crops, and create a cultivation work plan D2 for crops. The unmanned aerial vehicle 30 corresponds to the unmanned moving body described in the claims.

[0019] The external terminal 20 is an information processing device such as a smartphone, a tablet terminal, or a personal computer that users such as workers who perform cultivation operations or managers who comprehensively manage cultivation operations use an application program or a browser program (that is, use the system 1).

[0020] The unmanned aerial vehicle 30 is equipped with an imaging device 31, a positioning device 32, and the like. The imaging device 31 flies over the unit field F and images the unit field F and / or the crops planted in the unit field F from directly above. The imaging device 31 may include an infrared camera, or may include a visible light camera and / or an ultraviolet camera. The positioning device 32 uses GNSS signals transmitted from satellites to calculate the current position (latitude, longitude, altitude, etc.) and the current orientation, etc. of the unmanned aerial vehicle 30. The current position detected by the positioning device 32 is associated with the image captured by the imaging device 31 to form captured data Aa.

[0021] The management device 10, the external terminal 20, and the unmanned aerial vehicle 30 can communicate with each other either via a communication network N such as the Internet, or directly. Although Figure 1 shows one management device 10, one external terminal 20, and one unmanned aerial vehicle 30 for convenience, the system is not limited to this example, and System 1 may include at least one management device 10, one external terminal 20, and / or one unmanned aerial vehicle 30.

[0022] As shown in Figure 1, the management device 10 includes a growth status calculation unit 11, a growth environment calculation unit 12, a growth estimation unit 13, a cultivation work plan creation unit 14, and an information storage unit 15, etc. The growth status calculation unit 11 determines the growth status of the crop planted in unit field F. The growth environment calculation unit 12 determines the growth environment of the crop planted in unit field F. The growth estimation unit 13 estimates the future growth trajectory of the crop. The cultivation work plan creation unit 14 creates a cultivation work plan D2 for the crop.

[0023] The information storage unit 15 stores various programs and data. As will be described in detail later, the data stored in the information storage unit 15 includes status data A1, growth distribution data A2, pest and disease distribution data A3, first reference data A4, first growth data A5, environmental data B1, accumulated temperature during the crop growth period B2, accumulated solar radiation during the crop growth period B3, second reference data B4, second growth data B5, growth progression data C1, cultivation work data D1, and cultivation work plan D2. The information storage unit 15 also stores the growth distribution learning model MA2, the pest and disease distribution learning model MA3, the accumulated temperature learning model MB2, the accumulated solar radiation learning model MB3, the growth learning model MAB, and the planning learning model MCD.

[0024] Figure 2 is a conceptual diagram showing the functional configuration of System 1. The growth status calculation unit 11 of System 1 includes a status data acquisition unit 111, a growth distribution calculation unit 112, a disease and pest distribution calculation unit 113, a first growth data calculation unit 114, and a growth status evaluation unit 115.

[0025] The status data acquisition unit 111 acquires status data A1 related to the growth status of the crop. The growth distribution calculation unit 112 calculates the crop's growth distribution data A2 based on the status data A1. The pest and disease distribution calculation unit 113 calculates the crop's pest and disease distribution data A3 based on the status data A1. The first growth data calculation unit 114 calculates the first growth data A5 representing the crop's growth status based on the growth distribution data A2 and the pest and disease distribution data A3. The growth status evaluation unit 115 evaluates the quality of the first growth data A5 by comparing it with a preset first standard data A4.

[0026] Status data A1 includes at least time-series image data Aa of crops captured by the imaging device 31 mounted on the unmanned aerial vehicle 30, fertilization data Ab for each unit field F where crops are grown in open fields, pesticide application data Ac for each unit field F, crop variety data Ad, and crop sowing time data Ae. Status data A1 primarily uses image data (map data). Variety data Ad and sowing time data Ae are string data.

[0027] As described above, the imaging data Aa is constructed by associating the image captured by the imaging device 31 on the unmanned aerial vehicle 30 with the current position detected by the positioning device 32. The management device 10 acquires the imaging data Aa from the unmanned aerial vehicle and stores it in the information storage unit 15. The imaging data Aa is stored in chronological order.

[0028] The imaging data Aa may be, for example, images taken hourly, daily, or weekly. However, it is preferable that the imaging period for imaging data Aa be short, as detailed changes can be determined by capturing a group of imaging data Aa over a short period. The imaging data Aa may also include data obtained by analyzing vegetation indices such as DVI, RVI, NDVI, GNDVI, SAVI, TSAVI, CAI, MTCI, REP, PRI, or RSI.

[0029] Fertilization data Ab is a map (image) of the fertilizer distribution in unit field F, including the type and amount of fertilizer applied to each micro-area within unit field F, and also includes the positional data of each micro-area within unit field F. Fertilization data Ab is obtained, for example, by sensing unit field F with an unmanned aerial vehicle 30 and analyzing the spectral data obtained at the time of sensing. The management device 10 acquires fertilization data Ab from the unmanned aerial vehicle 30 and stores it in the information storage unit 15.

[0030] The pesticide application data Ac is a map (image) of the pesticide distribution in the unit field F, including the type and amount of pesticide applied in each micro-area within the unit field F, and also includes the positional data of each micro-area within the unit field F. Like the fertilization data Ab, the pesticide application data Ac is obtained, for example, by sensing the unit field F with an unmanned aerial vehicle 30 and analyzing the spectral data obtained at the time of sensing. The management device 10 acquires the pesticide application data Ac from the unmanned aerial vehicle 30 and stores it in the information storage unit 15.

[0031] Fertilization data Ab may be obtained by associating the type and amount of fertilizer applied by a fertilizer application device such as a rice transplanter or transplanter with the position data of the rice transplanter or transplanter. Similarly, pesticide application data Ac may be obtained by associating the type and amount of pesticide applied by a pesticide application device such as a rice transplanter or transplanter with the position data of the rice transplanter or transplanter. In this case, the management device 10 acquires the fertilization data Ab and pesticide application data Ac from the rice transplanter or transplanter and stores them in the information storage unit 15. Data on the types of fertilizers and pesticides may also be provided to the management device 10 from an external terminal 20, which is then input by the user.

[0032] The crop variety data Ad and sowing time data Ae are entered by the user into an external terminal 20 and provided from the external terminal 20 to the management device 10. Alternatively, the sowing time data Ae may be provided from the sowing machine to the management device 10 based on the operating date and time of the sowing machine, which is an agricultural machine.

[0033] In the growth distribution calculation unit 112, crop growth distribution data A2 is calculated based on state data A1 using the growth distribution learning model MA2. The growth distribution data A2 is a map (image) of the size and growth rate of crops in each micro-area within a unit field F, i.e., the growth distribution of crops in the unit field F, and also includes positional data for each micro-area within the unit field F. The growth distribution data A2 is generated in a time series, similar to the imaging data Aa.

[0034] The growth distribution learning model MA2 is a model used for machine learning related to crop growth distribution. Although not shown in the diagram, it has an input layer, an intermediate layer, and an output layer. The training data for the growth distribution learning model MA2 includes historical state data A1, which includes historical imaging data Aa, fertilization data Ab, pesticide application data Ac, variety data Ad, and sowing time data Ae.

[0035] The growth distribution calculation unit 112 trains the growth distribution learning model MA2 using the past state data A1, which serves as training data. Any known machine learning method may be used. For example, the growth distribution learning model MA2 can be trained using deep learning algorithms such as convolutional neural networks (CNNs) or recurrent neural networks (RNNs). The past state data A1 may be added or updated at appropriate intervals. The growth distribution learning model MA2 may also be added to or retrained using the added or updated past state data A1.

[0036] In the pest and disease distribution calculation unit 113, the pest and disease distribution learning model MA3 is used to calculate crop pest and disease distribution data A3 based on the state data A1. The pest and disease distribution data A3 is a map (image) of the occurrence of pests and diseases in each micro-area within a unit field F, i.e., the distribution of crop pests and diseases in the unit field F, and also includes the positional data of each micro-area within the unit field F. The pest and disease distribution data A3 is also generated in a time series, similar to the imaging data Aa and the growth distribution data A2.

[0037] The disease and pest distribution learning model MA3 is a machine learning model used for the distribution of diseases and pests in crops, and it has an input layer, an intermediate layer, and an output layer. The training data for the disease and pest distribution learning model MA3 includes historical state data A1, which is historical imaging data Aa, fertilization data Ab, pesticide application data Ac, variety data Ad, and sowing time data Ae, similar to the growth distribution learning model MA2.

[0038] The pest and disease distribution calculation unit 113 trains the pest and disease distribution learning model MA3 using the past state data A1, which serves as training data. Similar to the growth distribution learning model MA2, known methods can be used for machine learning. In this case as well, the past state data A1 may be added or updated at appropriate times. The pest and disease distribution learning model MA3 may also be added to or retrained using the added or updated past state data A1.

[0039] In the growth state calculation unit 11, when the state data acquisition unit 111 acquires state data A1, the growth distribution calculation unit 112 accesses the growth distribution learning model MA2, inputs the state data A1 into the growth distribution learning model MA2 for calculation, and outputs the crop growth distribution data A2 from the growth distribution learning model MA2. Then, the pest and disease distribution calculation unit 113 accesses the pest and disease distribution learning model MA3, inputs the state data A1 into the pest and disease distribution learning model MA3 for calculation, and outputs the crop pest and disease distribution data A3 from the pest and disease distribution learning model MA3 (see Figures 2 and 4).

[0040] The first growth data calculation unit 114 calculates the first growth data A5, which represents the growth state of the crop, from the growth distribution data A2 and the pest and disease distribution data A3. The first growth data A5 is a time-series map (image) of the growth distribution data A2 and the pest and disease distribution data A3, that is, the growth distribution and pest and disease distribution of the crop in a unit field F.

[0041] The growth status evaluation unit 115 evaluates the quality of the first growth data A5 by comparing it with the pre-set first standard data A4. The first standard data A4 is image data (map data) that shows the standard growth status of crops for each unit field F. In this case, it is possible to extract standard data from past first growth data A5 and set it as the first standard data A4. For example, a user may create and set the first standard data A4 from past growth (cultivation) history of the same type of crop.

[0042] The configuration of the growth status calculation unit 11 allows for accurate diagnosis of the crop growth status in a unit field F and proper determination of its quality by comparing the mapped (imaged) first growth data A5 with the first reference data A4. Regardless of the size of the field (even large fields), the crop growth status can be diagnosed comprehensively and quickly. In particular, since the mapped (imaged) first growth data A5 and first reference data A4 are obtained from a combination of crop growth distribution and pest / disease distribution in a unit field F (a combination of growth distribution data A2 and pest / disease distribution data A3), the crop growth status can be evaluated with higher accuracy and precision (see Figures 2 and 4).

[0043] The growth environment calculation unit 12 of System 1 includes an environmental data acquisition unit 121, an accumulated temperature calculation unit 122, an accumulated solar radiation calculation unit 123, a second growth data calculation unit 124, and a growth environment evaluation unit 125. The environmental data acquisition unit 121 acquires environmental data B1 related to the crop's growth environment. The accumulated temperature calculation unit 122 calculates the accumulated temperature B2 during the crop's growth period based on the environmental data B1. The accumulated solar radiation calculation unit 123 calculates the accumulated solar radiation B3 during the crop's growth period based on the environmental data. The second growth data calculation unit 124 calculates second growth data B5 representing the crop's growth environment based on the accumulated temperature B2 and accumulated solar radiation B3. The growth environment evaluation unit 125 evaluates the quality of the second growth data B5 by comparing it with a preset second standard data B4.

[0044] Environmental data B1 includes at least time-series weather data Ba for a predetermined area containing a unit field F, time-series local weather data Bb measured at unit field F, and field feature data Bc related to the characteristics of unit field F. Of the environmental data B1, the weather data Ba and local weather data Bb are string data. The field feature data Bc may be string data or image data (map data).

[0045] Weather data Ba includes data such as temperature, humidity, sunshine amount, and precipitation for a predetermined area where a unit field F is located, as well as data for the date and time of these observations. Weather data Ba is stored in a time series. The management device 10 acquires weather data Ba from a weather server 40, for example, located at the Japan Meteorological Agency or a weather provider, via a communication network N, and stores it in the information storage unit 15.

[0046] Local weather data Bb includes data such as temperature, humidity, sunshine amount, and precipitation for a unit field F or its surroundings, as well as data for the date and time of these observations. Like weather data Ba, local weather data Bb is also stored in a time series. The management device 10 acquires local weather data Bb from a group of field sensors 50 located in or around the unit field F via a communication network N and stores it in the information storage unit 15. The group of field sensors 50 includes, for example, a temperature sensor, a humidity sensor, an illuminometer, and a precipitation sensor. The temperature sensor detects the temperature of the unit field F or its surroundings. The humidity sensor detects the humidity of the unit field F or its surroundings. The illuminometer detects the amount of sunshine for the unit field F or its surroundings. The precipitation sensor detects the amount of precipitation for the unit field F or its surroundings.

[0047] Field characteristic data Bc is a map (image) of the past growth (cultivation) history of each micro-area within a unit field F, and also includes the positional data of each micro-area within the unit field F. Past growth history includes distributions of poor growth and lodging. Field characteristic data Bc is input by the user into an external terminal 20 and provided from the external terminal 20 to the management device 10. Alternatively, field characteristic data Bc may be obtained by, for example, imaging the unit field F with the imaging device 31 of an unmanned aerial vehicle 30 and obtaining the resulting imaging data Aa.

[0048] In the cumulative temperature calculation unit 122, the cumulative temperature B2 during the crop's growth period is calculated based on the environmental data B1 using the cumulative temperature learning model MB2. The cumulative temperature B2 is an index that represents the cumulative temperature necessary for the growth and maturation of the crop, and is the sum of the average temperatures over a predetermined period. As the cumulative temperature B2, the effective cumulative temperature, which is the value obtained by accumulating temperatures above the zero growth temperature on an hourly basis in order to understand the growth, flowering, and fruiting of the crop, may also be used.

[0049] The cumulative temperature learning model MB2 is a model used for machine learning regarding the cumulative temperature in a single field F, and has an input layer, an intermediate layer, and an output layer. The training data for the cumulative temperature learning model MB2 includes historical environmental data B1, namely historical weather data Ba, local weather data Bb, and field feature data Bc.

[0050] The integrated temperature calculation unit 122 trains the integrated temperature learning model MB2 using the past environmental data B1, which serves as training data. Similar to the growth distribution learning model MA2, known methods can be used for machine learning. In this case as well, the past environmental data B1 may be added or updated at appropriate intervals. The integrated temperature learning model MB2 may also be added to or retrained using the added or updated past environmental data B1.

[0051] The integrated solar radiation calculation unit 123 uses the integrated solar radiation learning model MB3 to calculate the integrated solar radiation B3 during the crop growing season based on the environmental data B1. The integrated solar radiation B3 is the value obtained by integrating the amount of solar energy per unit time over a predetermined period on an hourly basis.

[0052] The cumulative solar radiation learning model MB3 is a model used for machine learning regarding cumulative solar radiation in a single field F, and has an input layer, an intermediate layer, and an output layer. The training data for the cumulative solar radiation learning model MB3 includes historical environmental data B1, i.e., historical weather data Ba, local weather data Bb, and field characteristic data Bc, similar to the cumulative temperature learning model MB2.

[0053] The integrated solar radiation calculation unit 123 trains the integrated solar radiation learning model MB3 using the past environmental data B1, which serves as training data. Similar to the integrated temperature learning model MB2, known methods can be used for machine learning. In this case as well, the past environmental data B1 may be added or updated at appropriate intervals. The integrated solar radiation learning model MB3 may also be added to or retrained using the added or updated past environmental data B1.

[0054] In the growth environment calculation unit 12, when the environmental data acquisition unit 121 acquires environmental data B1, the accumulated temperature calculation unit 122 accesses the accumulated temperature learning model MB2, inputs the environmental data B1 into the accumulated temperature learning model MB2 for calculation, and outputs the accumulated temperature B2 for the crop over a predetermined period from the accumulated temperature learning model MB2. Then, the accumulated solar radiation calculation unit 123 accesses the accumulated solar radiation learning model MB3, inputs the environmental data B1 into the accumulated solar radiation learning model MB3 for calculation, and outputs the accumulated solar radiation B3 for the crop over a predetermined period from the accumulated solar radiation learning model MB3 (see Figures 2 and 4).

[0055] Furthermore, the calculation of the accumulated temperature B2 in the accumulated temperature calculation unit 122 and the calculation of the accumulated solar radiation B3 in the accumulated solar radiation calculation unit 123 are not limited to the above-described forms, but may also be performed by calculating the degree of variation in numerical values, such as by simple deviation matching, least squares method, or component analysis method. For example, the difference in these calculation results can be used in the estimation calculation of future growth progression data C1, which will be described later.

[0056] The second growth data calculation unit 124 calculates second growth data B5, which represents the crop's growth environment, from the accumulated temperature B2 and accumulated solar radiation B3. The second growth data B5 is a time-series combination of the accumulated temperature B2 and accumulated solar radiation B3 for each unit field F, and is an indicator that shows what kind of environment is necessary for the crop's growth and maturation.

[0057] The growth environment evaluation unit 125 evaluates the quality of the second growth data B5 by comparing it with the pre-set second standard data B4. The second standard data B4 is data that shows the standard growth environment for crops in each unit field F. This data may be in the form of text data or image data (map data). Standard data from past second growth data B5 may be extracted and set as the second standard data B4. For example, the user may create and set the second standard data B4 from past accumulated temperature B2 and accumulated solar radiation B3 for unit field F.

[0058] According to the configuration of the growth environment calculation unit 12, the quality of the crop growth environment in unit field F can be appropriately determined by comparing the second growth data B5 with the second standard data B4, and the second growth data B5 can be used as supplementary data to the first growth data A5, which indicates the growth status of the crop. Therefore, the diagnostic accuracy of crop growth status based on the first growth data A5 can be further improved (see Figures 2 and 4).

[0059] In the growth estimation unit 13 of System 1, future crop growth trend data C1 is estimated from the first and second growth data A5 and B5 using the growth learning model MAB, which uses past first and second growth data A5 and B5 as training data. The growth estimation unit 13 includes a growth data acquisition unit 131. The growth data acquisition unit 131 acquires the first growth data A5 calculated by the first growth data calculation unit 114 and the second growth data B5 calculated by the second growth data calculation unit 124.

[0060] Crop growth progression data C1 includes, for example, the size and color of the crop's stems, leaves, and fruits (which can also be described as the size and growth rate of the crop), the degree of crop deterioration due to pests and diseases, and the degree of lodging of the crop. In other words, growth progression data C1 is a map (image) of the overall growth distribution of crops in each micro-zone within a unit field F, and also includes the location data of each micro-zone within the unit field F. Growth progression data C1 is generated over time.

[0061] The Growth Learning Model (MAB) is a machine learning model used for the overall growth distribution of crops, and it has an input layer, an intermediate layer, and an output layer. The training data for the Growth Learning Model (MAB) includes historical first and second growth data A5 and B5.

[0062] The growth estimation unit 13 trains the growth learning model MAB using the past first and second growth data A5 and B5, which serve as training data. Known methods may be used for machine learning. The past first and second growth data A5 and B5 may be added or updated at appropriate intervals. The growth learning model MAB may also be added to or retrained using the added or updated past first and second growth data A5 and B5.

[0063] When the growth estimation unit 13 acquires the first and second growth data A5 and B5, it accesses the growth learning model MAB, inputs the first and second growth data A5 and B5 into the growth learning model MAB for calculation, and outputs crop growth progression data C1 from the growth learning model MAB (see Figures 2 and 4). Note that since the second growth data B5 functions as complementary data to the first growth data A5 which indicates the growth state of the crop, the growth learning model MAB may learn only past first growth data A5 as training data, and the growth estimation unit 13 may use the growth learning model MAB to estimate the future growth progression data C1 of the crop from the first growth data A5.

[0064] The configuration of the growth estimation unit 13 allows for the prediction of future crop growth trends in a time series for each unit field F. This enables highly accurate planning when creating the cultivation work plan D2, and allows for the acquisition of a cultivation work plan D2 and various data that closely reflect actual conditions. Furthermore, because future crop growth trends can be predicted for each unit field F, the rate of crop growth failure can also be predicted with high accuracy (see Figures 2 and 4).

[0065] The cultivation work plan creation unit 14 of System 1 includes a growth progression data acquisition unit 141 and a cultivation work data acquisition unit 142. The growth progression data acquisition unit 141 acquires growth progression data C1 estimated by the growth estimation unit 13. The cultivation work data acquisition unit 142 acquires cultivation work data D1 related to crop cultivation management. The cultivation work plan creation unit 14 uses a planned learning model MCD, which uses past growth progression data C1 and past cultivation work data D1 as training data, to create a cultivation work plan D2 for the crop from the growth progression data C1 and cultivation work data D1 (see Figures 2 and 4).

[0066] In other words, the cultivation work plan creation unit 14 creates a cultivation work plan D2 to select the optimal timing for cultivation work in accordance with the future growth trajectory estimated by the growth estimation unit 13. The work content includes, for example, soil preparation, ridge making, levee construction, tilling, sowing, transplanting, rice planting, puddling, furrowing, fertilization, pesticide application, top dressing (pest control), irrigation, or harvesting. In this case, the cultivation work plan D2 may be planned in hourly units, daily units, or weekly units. However, it is understood that a shorter period unit for the cultivation work plan D2 is preferable, as it enables the efficient operation of the entire cultivation work.

[0067] The cultivation work data D1 includes at least the following: schedule data Da, which serves as a guideline for the dates of cultivation work; agricultural machinery data Db, which pertains to various agricultural machines such as tractors, transplanters, and combine harvesters; personnel data Dc, which pertains to workers performing cultivation work; and material data Dd, which pertains to the types and quantities of fertilizers and pesticides (usage and inventory). The cultivation work data D1 (schedule data Da, agricultural machinery data Db, personnel data Dc, and material data Dd, etc.) is entered by the user into an external terminal 20 and provided from the external terminal 20 to the management device 10.

[0068] Alternatively, past cultivation work data D1 stored in the information storage unit 15 may be modified using an external terminal 20 and provided to the management device 10 as the added or updated cultivation work data D1. Alternatively, the cultivation work data D1 may be pre-stored in the information storage unit 15 of the management device 10, with only the modifications being input from the management device 10 or the external terminal 20.

[0069] The cultivation work plan creation unit 14, in creating a cultivation work plan D2 using the planning learning model MCD, allocates agricultural machinery, personnel, and materials to each unit field F based on a time axis (e.g., hourly, daily, or weekly). The planning learning model MCD is a machine learning model used for creating cultivation work plans for crops, and has an input layer, an intermediate layer, and an output layer. The training data for the planning learning model MCD includes past growth progression data C1 and past cultivation work data D1, etc.

[0070] The cultivation work plan creation unit 14 trains the planning learning model MCD using the training data, which includes past growth progression data C1 and past cultivation work data D1. Known methods can be used for machine learning. The past growth progression data C1 and past cultivation work data D1 may be added or updated as appropriate. The planning learning model MCD may also be added to or retrained using the added or updated past growth progression data C1 and past cultivation work data D1.

[0071] In the cultivation work plan creation unit 14, upon acquiring growth progression data C1 and cultivation work data D1, it accesses the planning learning model MCD, inputs the growth progression data C1 and cultivation work data D1 into the planning learning model MCD for calculation, and outputs the crop cultivation work plan D2 from the planning learning model MCD.

[0072] When a user, such as a worker or manager, performs a predetermined operation on the external terminal 20, the external terminal 20 retrieves the cultivation work plan D2 from the management device 10 and displays the cultivation work plan D2 on the terminal display unit 206 (see Figure 3). The user then performs the cultivation work according to the cultivation work plan D2 displayed on the terminal display unit 206.

[0073] According to the configuration of the cultivation work plan creation unit 14, a cultivation work plan D2 can be created based on a time axis (e.g., hourly, daily, or weekly), taking into account adjustments to work hours according to field size, postponement of cultivation work according to weather, selection of agricultural machinery and maintenance requirements, personnel allocation, and the amount and inventory of fertilizers and pesticides used. Therefore, even inexperienced or less experienced individuals can schedule cultivation work, and the management of agricultural machinery, personnel, and materials can be easily carried out.

[0074] Furthermore, cultivation tasks in each unit field F can be carried out in an order suitable for such tasks as harvesting, eliminating wasted allocation and use of personnel and materials. For example, the schedule, allocation of agricultural machinery and personnel can be calculated considering the number and usage of personnel and agricultural machinery, enabling efficient management of the entire cultivation operation (see Figures 2 and 4).

[0075] Furthermore, it is user-friendly because external terminals 20 can access the management device 10, allowing users to create, view, and modify cultivation work plans D2 on the management device 10. Since modifications can be made to cultivation work plans D2, it is possible to revise the cultivation work plans D2 to accommodate sudden schedule changes, such as personnel changes due to illness or leave, or the reassignment of farm equipment due to breakdowns or maintenance.

[0076] Furthermore, the cultivation work plan creation unit 14 can set not only the items of various cultivation tasks but also the procedures for those tasks. The cultivation work plan D2 in the first embodiment includes, for example, soil preparation, levee construction, tilling, sowing, rice planting, puddling, furrowing, fertilization, pesticide application, top dressing (pest control), irrigation, or harvesting. Therefore, the management device 10 can calculate the materials necessary for each task (for example, the types and amounts of seeds, fertilizers, and pesticides, the amount of irrigation, etc.) and execute various cultivation tasks based on the amount of these materials.

[0077] It is also possible to set the travel route of agricultural machinery within the unit field F, or set the travel time (cultivation work time), by utilizing the positional data of each minute area of ​​the unit field F. The width of the agricultural machinery (work width), the overlap width in the width direction where agricultural machinery overlaps in adjacent work areas, and the speed of the agricultural machinery may be set and configured to be modifiable via an external terminal 20 or the like.

[0078] Furthermore, since the management device 10 collects status data A1 and generates first growth data A5, it can, for example, consider the past fertility of the unit field F to determine the necessary materials (e.g., the type and amount of seeds, fertilizers, and pesticides), manage inventory, and suggest the timing for ordering materials. In this case, the suggestion can be made, for example, by displaying it on the terminal display unit 206 of the external terminal 20, or by issuing a warning from the external terminal 20. The management device 10 can also be set to automatically order materials according to the calculation results. Moreover, if soil data representing the hardness or softness of each minute area of ​​the unit field F is acquired as status data A1, the management device 10 can also set the transplanting accuracy (spacing between plants, planting depth, etc.) of agricultural machinery such as transplanters and rice transplanters.

[0079] Now, while the agricultural machinery data database (Db) for various types of agricultural machinery can include model data and maintenance data for the machinery in question, it may also include disposal and recycling data. For example, when disposing of agricultural machinery, a user such as an operator can enter the address (city, town, etc.) from an external terminal 20, which will allow access to the city, town, and responsible sales company (including waste disposal companies), and enable them to understand the disposal procedure for the agricultural machinery based on the disposal and recycling data.

[0080] The disposal procedures that can be confirmed from the waste recycling data include general disassembly and dismantling methods for each agricultural machine. The explanations of disassembly and dismantling methods may include not only text data, image data, and still image data, but also video data. This allows users to easily understand the disposal and sorting methods. By providing access to the waste recycling data for the agricultural machine via a QR code attached to the machine or its instruction manual, the cumbersome process of searching for disposal procedures can be avoided.

[0081] In explaining the disassembly and dismantling methods, emphasis should be placed on recyclable parts. For example, a system could be adopted where customers receive a cashback when they bring their agricultural machinery to the sales company for disposal. This would promote recycling and contribute to reducing industrial waste. It is also possible to establish a system where the sales company sells the recyclable parts of the agricultural machinery brought in to recycling companies or other related parties.

[0082] Figure 3 is a conceptual diagram illustrating the hardware configuration of System 1 (management device 10 and external terminal 20). The management device 10 includes a processor 101, memory 102, storage 103, communication IF 104, input device 105, display unit 106, and input / output IF 107. The processor 101 is a CPU, GPU, or MPU, and performs various calculations according to the program to realize the functions of the management device 10. The number of processors 101 may be one or multiple. When multiple processors 101 are used, these processors 101 may execute processing simultaneously or sequentially.

[0083] Memory 102 functions as a work area when the processor 101 performs processing. Memory 102 includes, for example, RAM and ROM. Storage 103 stores various programs and data, and includes, for example, an SSD and / or HDD. Storage 103 stores programs that realize the functions of the management device 10.

[0084] The communication interface 104 communicates with other devices according to a predetermined communication standard (e.g., Ethernet). The communication interface 104 includes, for example, a network interface card (NIC). The input device 105 inputs information in response to user operations. The input device 105 includes, for example, at least one of a touch sensor, keyboard, keypad, mouse, and microphone.

[0085] The display unit 106 displays various types of information. The display unit 106 includes, for example, an LCD. The input / output IF 107 is an interface for connecting the display unit 106 and the input device 105. The processor 101 is an example of a growth status calculation unit 11, a growth environment calculation unit 12, a growth estimation unit 13, and a cultivation work plan creation unit 14. At least one of the memory 102 and storage 103 is an example of an information storage unit 15.

[0086] The external terminal 20 includes a processor 201, memory 202, storage 203, communication IF 204, input device 205, terminal display unit 206, and input / output IF 207. The processor 201, memory 202, storage 203, communication IF 204, input device 205, terminal display unit 206, and input / output IF 207 perform the same functions as the processor 101, memory 102, storage 103, communication IF 104, input device 105, display unit 106, and input / output IF 107 described above. However, the storage 203 stores a program that implements the functions of the external terminal 20.

[0087] As is clear from the above explanation, according to the first embodiment, the growth status of crops in unit field F can be accurately diagnosed and their quality can be appropriately determined by comparing the mapped (imaged) first growth data A5 with the first reference data A4. Regardless of the size of the field (even large fields), the growth status of crops can be diagnosed comprehensively and in a short time. In particular, since the mapped (imaged) first growth data A5 and first reference data A4 are obtained from a combination of crop growth distribution and pest and disease distribution in unit field F (a combination of growth distribution data A2 and pest and disease distribution data A3), the growth status of crops can be evaluated with higher accuracy and precision.

[0088] Furthermore, by comparing the second growth data 55 with the second standard data B4, it is possible to appropriately determine the quality of the crop growth environment in unit field F, and the second growth data B5 can be used as supplementary data to the first growth data A5, which indicates the crop's growth status. Therefore, the diagnostic system for crop growth status based on the first growth data A5 can be further improved.

[0089] Furthermore, because the future growth trajectory of crops can be predicted in time series (e.g., hourly, daily, or weekly) for each unit field F, it becomes possible to create cultivation work plans D2 with high accuracy, and to obtain cultivation work plans D2 and various data that are closer to the actual situation. In addition, because the future growth trajectory of crops can be predicted for each unit field F, the rate of crop growth failure can also be predicted with high accuracy.

[0090] Furthermore, it is possible to create a cultivation work plan D2 based on a time axis (e.g., hourly, daily, or weekly), taking into account adjustments to working hours according to field size, postponement of cultivation work according to weather, selection and maintenance status of agricultural machinery, personnel allocation, and the amount and inventory of fertilizers and pesticides used. As a result, even inexperienced or less experienced individuals can schedule cultivation work, and the management of agricultural machinery, personnel, and materials can be easily carried out.

[0091] Furthermore, cultivation tasks in each unit field F can be carried out in an order suitable for harvesting and other cultivation operations, eliminating wasted allocation and use of personnel and materials. For example, the system can calculate the schedule and allocation of agricultural machinery and personnel, taking into account the number and usage of personnel and agricultural machinery, thereby improving the efficiency of cultivation operations.

[0092] Furthermore, it is user-friendly because external terminals 20 can access the management device 10, allowing users to create, view, and modify cultivation work plans D2 on the management device 10. Since modifications can be made to cultivation work plans D2, it is possible to revise the cultivation work plans D2 to accommodate sudden schedule changes, such as personnel changes due to illness or leave, or the reassignment of farm machinery due to breakdowns or maintenance.

[0093] In summary, by using growth-related status data A1 and growth-related environmental data B1, future crop growth trend data C1 can be generated in time series (e.g., hourly, daily, or weekly), and finally, a cultivation work plan D2 for the crop can be created based on the time axis, enabling efficient management of the entire cultivation operation.

[0094] Next, the cultivation management system 1 of the second embodiment will be described with reference to Figures 5 to 8. Figure 5 is an explanatory diagram showing an overview of the cultivation management system of the second embodiment, Figure 6(a) is a side view of the unmanned vehicle for greenhouse specifications, Figure 6(b) is a rear view of the unmanned vehicle for greenhouse specifications, Figure 7(a) is a side view of the unmanned vehicle for plant factory specifications, Figure 7(b) is a rear view of the unmanned vehicle for plant factory specifications, and Figure 8 is an explanatory diagram of a furrow. In the second embodiment, components whose configuration and operation are common to the second embodiment are denoted by the same reference numerals as in the first embodiment, and their detailed explanation is omitted.

[0095] In the second embodiment of System 1, an unmanned mobile vehicle 60 is used as an unmanned mobile vehicle instead of an unmanned flying vehicle 30 such as a drone. The unmanned mobile vehicle 60 is a self-propelled vehicle used in cultivation facilities such as greenhouses and plant factories, and is divided into three parts: a driving unit 61, a main body 62, and a measurement and work unit 63. The main body 62 is mounted on the driving unit 61, and the measurement and work unit 63 is mounted on the main body 62.

[0096] The unmanned vehicle 60 is formed in a front-to-back symmetrical shape when viewed from the left and right sides intersecting the direction of travel, and the unmanned vehicle 60 itself has a structure that does not distinguish between front and back. For this reason, the unmanned vehicle 60 of the second embodiment can be configured in multiple patterns, making it easy to respond to various user requests. In addition, because the unmanned vehicle 60 has a structure that does not distinguish between front and back, it also has the advantage of being easy to perform automatic driving control.

[0097] The running section 61 has two types: a crawler-type running section 611 and a wheeled running section 612. The crawler-type running section 611 is used for soil cultivation in greenhouses, while the wheeled running section 612 is used in plant factories. In greenhouses, the unmanned vehicle 60 travels between adjacent rows 64. In plant factories, the unmanned vehicle 60 travels along the running rails 65.

[0098] The crawler-type running section 611 and the wheel-type running section 612 are configured to be interchangeable with the main body section 62. Safety bars 614 are provided on both the front and rear sides of the running section 61 so as to be extendable and retractable in the front-to-back direction. In other words, safety bars 614 are present on both the front and rear sides of the crawler-type running section 611, and safety bars 614 are also present on both the front and rear sides of the wheel-type running section 612.

[0099] The main body 62 consists of two types: a crawler body 621 and a wheel body 622. An operating section 627 is provided on one side of the main body 62. End determination plate detection sensors 623 are provided on both the front and rear sides of the main body 62 to detect end determination plates 66 that indicate the starting and ending positions of, for example, ridges 64 or running rails 65. With this configuration, the unmanned vehicle 60 can accurately recognize the starting and ending positions of ridges 64 or running rails 65, making it easier to perform subsequent turning and stopping operations smoothly and improving the accuracy of automatic driving.

[0100] On both the left and right sides of the crawler body 621, there are furrow position determination plate detection sensors 624 that detect furrow position determination plates 67 placed on the upper surface of adjacent furrows 64 during travel. With this configuration, the unmanned vehicle 60 can accurately recognize the position of adjacent furrows 64, enabling more accurate automated travel. Furthermore, by aligning the determination height positions of the end determination plate 66 and the furrow position determination plate 67, and aligning the mounting height positions of the end determination plate detection sensor 623 and the furrow position determination plate detection sensor 624 with these, it is easier to improve and stabilize detection accuracy.

[0101] Multiple bumper arms 625 are provided symmetrically on both the left and right sides of the crawler body 621. In the second embodiment, a total of four bumper arms 625 are provided, two on each side of the operating unit 627 and the furrow position determination plate detection sensor 624. The bumper arms 625 are configured to be extendable and retractable in their longitudinal direction, and spherical rollers 626 are attached to their tips, which contact the longitudinal side of adjacent furrows 64 when traveling. When the unmanned vehicle 60 travels between adjacent furrows 64, the vehicle 60 travels while the spherical rollers 626 of each bumper arm 625 are in contact with the longitudinal side of the furrows 64.

[0102] With this configuration, even if there are uneven surfaces between adjacent furrows 64, there is no risk of the unmanned vehicle 60 tipping over, enabling stable automatic driving. Furthermore, after the unmanned vehicle 60 has passed between adjacent furrows 64, each bumper arm 625 can also function as a safety bar 614.

[0103] The measurement and operation unit 63 includes three types of equipment: a growth diagnostic device 631, a pest control device (not shown), and a harvesting device (not shown). The growth diagnostic device 631 detects chlorophyll fluorescence to measure the stress level of the crop. By detecting chlorophyll fluorescence, it becomes possible to quantitatively evaluate the state of photosynthesis (stress level) of the crop. In other words, if the amount of chlorophyll fluorescence is low, it indicates that photosynthesis is active and the plant is under low stress, and if the amount of chlorophyll fluorescence is high, it indicates that photosynthesis is poor and the plant is under high stress.

[0104] The growth diagnostic device 631 includes an optical unit 632 consisting of a blue LED light source 633 and a CCD camera 634, and a CO2 sensor 635. The optical unit 632 and the CO2 sensor 635 are controlled at appropriate timings to diagnose crop growth. The P / S value data Af obtained from the optical unit 632, which compares chlorophyll fluorescence intensity, and the CO2 concentration data Ag detected by the CO2 sensor 635 are included in the system 1 state data A1 in the second embodiment. In particular, the P / S value data Af makes it possible to distinguish between crops of different varieties, to distinguish between different parts within a single crop (e.g., leaves and stems), and to identify damage caused by pests and diseases.

[0105] Although detailed illustrations are omitted, the pest control machine, for example, uses an electric pump to spray pesticides from a pest control tank onto the crops through a sprayer. The harvesting machine has an imaging device to recognize crops, a harvesting hand capable of gripping crops, and a storage bucket. The imaging device acquires positional data of the harvesting portion of the crop, and based on this positional data, the harvesting hand grips the harvesting portion, picks the crop, and stores it in the storage bucket.

[0106] The unmanned vehicle 60, like the unmanned aircraft 30 of the first embodiment, is equipped with an imaging device 31 and a positioning device 32. For example, the imaging device 31 may be included in the measurement and work unit 63, and the positioning device 32 may be included in the main unit 62. The unmanned vehicle 30 can collect status data A1, such as time-series imaging data Aa of crops, fertilization data Ab, pesticide application data Ac, P / S value data Af, and CO2 concentration data Ag.

[0107] The system 1 of the second embodiment, configured as described above, also produces the same effects as the first embodiment. Using state data A1 related to growth status and environmental data B1 related to the growth environment, it is possible to generate time-series data C1 on the future growth progression of the crop, ultimately creating a cultivation work plan D2 for the crop, and enabling efficient management of the entire cultivation operation.

[0108] The configuration of each part in the present invention is not limited to the illustrated embodiment, and various modifications are possible without departing from the spirit of the present invention. [Explanation of Symbols]

[0109] 1. Cultivation Management System 10 Management device 11. Growth Status Calculation Section 111 Status data acquisition unit 112 Growth distribution calculation section 113 Pest distribution calculation section 114 First-stage child development data calculation unit 115 Growth Status Evaluation Department 12 Growth Environment Calculation Department 121 Environmental Data Acquisition Unit 122 Integrated Temperature Calculation Unit 123 Integrated solar radiation calculation unit 124 Second Growth Data Calculation Unit 125 Growth Environment Evaluation Department 13 Growth Estimation Department 131 Growth Data Acquisition Unit 14. Cultivation Work Plan Creation Department 141 Growth Progression Data Acquisition Unit 142 Cultivation work data acquisition unit 15 Information storage section 20 External terminals 30 Unmanned aircraft 31 Imaging device 32 Positioning device 40 Weather Servers 50 field sensor groups A1 Status Data Aa Imaging data Ab Fertilization Data Ac application data Ad Variety Data Ae sowing time data Af P / S value data Ag CO2 concentration data A2 Growth distribution data A3 Pest and Disease Distribution Data A4 First Reference Data A5 First-year childcare data B1 Environmental Data Ba Weather Data Bb Local Weather Data Bc Field Characteristics Data B2 Accumulated temperature B3 Cumulative solar radiation B4 Second standard data B5 Second Growth Data C1 Growth Progression Data D1 Cultivation work data Da Schedule Data Db Agricultural Machinery Data DC Personnel Data Dd Material Data D2 Cultivation Work Plan F Unit Field MA2 Growth Distribution Learning Model MA3 Pest and Disease Distribution Learning Model MB2 Integrated Temperature Learning Model MB3 Integrated Solar Radiation Learning Model MAB Growth and Learning Model MCD (Multi-Career Designed Learning Model) N Communication Network

Claims

1. A cultivation management system for managing open-field cultivation of crops, The system comprises: a state data acquisition unit that acquires state data related to the growth state of the crop; a growth distribution calculation unit that calculates growth distribution data of the crop based on the state data; a pest and disease distribution calculation unit that calculates pest and disease distribution data of the crop based on the state data; a first growth data calculation unit that calculates first growth data representing the growth state of the crop based on the growth distribution data and the pest and disease distribution data; and a growth state evaluation unit that evaluates the quality of the first growth data by comparing it with a preset first standard data. The aforementioned status data includes at least time-series imaging data of the crop captured by an imaging device mounted on an unmanned mobile vehicle, fertilization data for each unit field where the crop is cultivated in the open field, pesticide application data for each unit field, crop variety data, and crop sowing time data. Cultivation management system.

2. The system comprises: an environmental data acquisition unit that acquires environmental data related to the growth environment of the crop; an accumulated temperature calculation unit that calculates the accumulated temperature during the growth period of the crop based on the environmental data; an accumulated solar radiation calculation unit that calculates the accumulated solar radiation during the growth period of the crop based on the environmental data; a second growth data calculation unit that calculates second growth data representing the growth environment of the crop based on the accumulated temperature and the accumulated solar radiation; and a growth environment evaluation unit that evaluates the quality of the second growth data by comparing it with a preset second reference data. The environmental data includes at least time-series weather data for a predetermined area where the unit field is located, time-series local weather data measured at the unit field, and field characteristic data related to the characteristics of the unit field. A cultivation management system as described in claim 1.

3. The system includes a growth estimation unit that uses a growth learning model to estimate the future growth trajectory of the crop from the aforementioned growth data. A cultivation management system as described in claim 1.

4. The system includes a growth estimation unit that uses a growth learning model to estimate the future growth trajectory of the crop from the first and second growth data. A cultivation management system as described in claim 2.

5. The system includes a cultivation work data acquisition unit that acquires cultivation work data related to the cultivation management of the crop, and a cultivation work plan creation unit that creates a cultivation work plan for the crop from the cultivation work data using a planning learning model. The aforementioned cultivation work data includes at least schedule data, farm equipment data, personnel data, and material data. The cultivation work plan creation unit creates the cultivation work plan by allocating the schedule, agricultural machinery, personnel, and materials to each of the unit fields based on the time axis. A cultivation management system according to claim 3 or 4.

Images