Monitoring apparatus, monitoring system, monitoring method, and program

Abstract

To provide a monitoring device capable of estimating abnormality occurring in a crop.SOLUTION: A calculation unit 11 configured to estimate an abnormality occurring in a crop in an area where the crop exists, wherein the calculation unit acquires a first detection value for detecting a state change occurring in the crop at a first position of a first block among two or more blocks included in the area, and a second detection value for detecting the state change at a second position of a second block included in the area, the monitoring device 10 compares the first detection value and the second detection value in time series, determines whether or not a predetermined state is included in the first detection value and the second detection value, and estimates that the block is an abnormal area where the abnormality occurs when the predetermined state is included.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a monitoring device, a monitoring system, a monitoring method, and a program for monitoring abnormalities occurring in crops. [Background technology]

[0002] Patent Document 1 describes a technology that can estimate the ripeness of crops based on the detected values ​​of odorous components. According to this technology, a first detected value of the odorous components and a second detected value of the crop hardness are measured in advance, a regression model that estimates the second detected value from the first detected value is calculated, and the regression model is used to estimate the crop hardness (ripeness) from the detected values ​​of the odorous components. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 2018-132325 Summary of the Invention [Problem to be solved by the invention]

[0004] Crops can suffer from various abnormalities, such as damage from pests, viruses, weather, and the proliferation of undesired plants. Abnormalities in crops are discovered by visual inspection by workers, who then treat the crops. In recent years, the number of agricultural workers has been declining, creating a demand for technology to efficiently manage crops. The technology described in Patent Document 1 does not estimate abnormalities in crops caused by pests or weather.

[0005] Furthermore, when using sensors to remotely monitor the condition of crops in outdoor environments, the sensor detection values ​​are affected by the type of crop, its growth rate, the sensor installation environment, and external environmental factors (rain, temperature, solar radiation, humidity, and air pressure). As a result, even for crops of the same type and growth rate, there will be variations in the sensor detection values. With such remote monitoring methods, it may be difficult to simply set upper and lower limits on the individual sensor detection values ​​and infer abnormalities.

[0006] An object of the present invention is to provide a monitoring device, a monitoring system, a monitoring method, and a program that are capable of estimating abnormalities occurring in crops. [Means for solving the problem]

[0007] One aspect of the present invention is a monitoring device that includes a calculation unit that estimates an abnormality occurring in a crop in an area where the crop is present, and the calculation unit acquires a first detection value that detects a state change occurring in the crop at a first position of a first block among two or more blocks included in the area, and a second detection value that detects the state change at a second position of a second block included in the area, compares the first detection value and the second detection value in a time series, determines whether a predetermined state is included in the first detection value and the second detection value, and if the predetermined state is included, estimates the block to be an abnormal area where the abnormality is occurring. [Effects of the Invention]

[0008] According to the present invention, abnormalities occurring in crops can be estimated. [Brief explanation of the drawings]

[0009] [Figure 1] 1 is a block diagram showing a configuration of a monitoring device according to an embodiment; [Figure 2] FIG. 1 is a diagram showing a schematic configuration of an area where crops are grown. [Figure 3] FIG. 1 is a plan view showing the configuration of a sensor for detecting volatile substances produced from crops. [Figure 4]FIG. 2 is a cross-sectional view showing the configuration of a sensor. [Figure 5] FIG. 10 is a diagram showing an example of a predetermined waveform of a detection value. [Figure 6] FIG. 10 is a diagram illustrating a state in which an abnormality occurring in an area is detected. [Figure 7] FIG. 10 is a diagram illustrating the propagation state of an anomaly occurring in a region. [Figure 8] FIG. 10 is a diagram showing an information display image including information about an anomalous propagation region. [Figure 9] FIG. 10 is a diagram showing an instruction display image showing a countermeasure instruction for suppressing an abnormality. [Figure 10] 4 is a flowchart showing a flow of processing of a monitoring method executed in the monitoring device. [Figure 11] 4 is a flowchart showing a flow of processing of a monitoring method executed in the monitoring device. [Figure 12] 4 is a flowchart showing a flow of processing of a monitoring method executed in the monitoring device. DETAILED DESCRIPTION OF THE INVENTION

[0010] As shown in Figures 1 and 2, the monitoring system 1 includes a detection unit S provided in an area R to be managed, and a monitoring device 10 communicatively connected to the detection unit S. The monitoring system 1 includes a countermeasure unit M that operates based on commands output by the monitoring device 10. The area R is a predetermined area where a crop H is cultivated. The area R may be an outdoor facility or an indoor facility. The crop H is, for example, a plant of a predetermined cultivable variety. The crop H may be one from which a desired harvest product is harvestable. The crop H is, for example, a plant whose state changes when an abnormal condition occurs.

[0011] Abnormal conditions include factors that affect the crop H, such as damage caused by pests, disease stress due to virus infection, damage due to the environment such as temperature changes, excess or shortage of water, excess or shortage of fertilizer, etc. State changes include changes in conditions that affect the growth of the crop H, such as a state in which the crop H emits a specified substance, a state in which the appearance of the crop H changes, or a state in which the electric potential of the crop H changes. The crop H is grown in a specified area R.

[0012] The area R is divided into n blocks Bn (n is a natural number). One block Bn is an area to be managed. Each block Bn may be set to have approximately the same shape and area, or may have a non-uniform shape and area. A predetermined number of crops H are cultivated in one block Bn. The area R is provided with a detection unit S that detects changes in the state of the crops H. The detection unit S is composed of n sensors Sn.

[0013] One block Bn is provided with, for example, one sensor Sn. The sensor Sn in this embodiment is composed of a substance sensor Tn capable of detecting a predetermined substance. The substance sensor Tn is configured to detect, for example, a volatile substance emitted into the atmosphere from crops H. The volatile substance is, for example, leaf alcohol. Leaf alcohol is a volatile substance emitted from crops H when abnormalities such as insect damage occur in the crops H. The volatile substance emitted from crops H has a signaling function to neighboring crops H. The volatile substance has the effect of activating the immunity of crops H. The immune system of neighboring crops H exposed to the volatile substance is activated.

[0014] The sensor Sn may be provided with a supply unit Kn that supplies a medium to stabilize the measurement of the substance sensor Tn. The supply unit Kn generates, for example, a gas mixture for supplying the volatile substance to be detected to the substance sensor Tn.

[0015] The supply unit Kn supplies an inert carrier gas such as nitrogen or air. The supply unit Kn supplies the carrier gas at a constant amount for a fixed time. The supply unit Kn generates a mixed gas by mixing the carrier gas with the detection target and supplies it to the substance sensor Tn. The supply unit Kn supplies only the carrier gas to the substance sensor Tn at a predetermined timing, removes the detection target sorbed on the substance sensor Tn, and resets the output of the substance sensor Tn. The supply unit Kn may be configured to supply a medium other than the carrier gas depending on the configuration of the substance sensor Tn.

[0016] The sensor Sn is provided with a position sensor Gn that detects its own position. The position sensor Gn is configured, for example, by a GPS (Global Positioning System) sensor. The position sensor Gn may be configured not only by a GPS sensor but also by other sensors that can detect its own position. The position sensor Gn may be configured to detect its own position indoors. The position sensor Gn may not be provided, for example, when the sensor Sn is managed by the monitoring device 10 based on the installation position of the sensor Sn and identification information that individually identifies the sensor.

[0017] The sensor Sn includes a communication unit Cn that communicates with the monitoring device 10 via the network W. The communication unit Cn is a communication interface that can be connected to the network W via a wired or wireless connection. The sensor Sn includes a control unit Pn for controlling its own operation and a memory unit Dn. The control unit Pn is configured with at least one hardware processor such as a CPU (Central Processing Unit). The memory unit Dn is configured with a non-transitory storage medium such as a hard disk drive (HDD) or a solid state disk (SSD). The control unit Pn and the memory unit Dn may be provided for each sensor Sn, or at least one may be provided for the detection unit S.

[0018] The control unit Pn acquires detection values ​​detected by the substance sensor Tn and the position sensor Gn. The control unit Pn stores the detection values ​​in the memory unit Dn. If the substance sensor Tn is provided with a supply unit Kn, the control unit Pn controls the supply unit Kn to adjust the sensitivity of the substance sensor Tn. The control unit Pn transmits the detection values ​​to the monitoring device 10 via the communication unit Cn. The memory unit Dn stores data and computer programs necessary for control. The memory unit Dn stores detection value data at a predetermined timing. The control unit Pn erases the detection value data stored in the memory unit Dn in chronological order at a predetermined timing, and maintains the detection value data capacity at a predetermined value. The control unit Pn controls the communication unit Cn at a predetermined timing to output the detection value data to the monitoring device 10 via the network W.

[0019] The monitoring device 10 acquires detection value data via a network W. The monitoring device 10 includes a calculation unit 11 that executes calculation processing and control for monitoring the area R based on the detection values. The monitoring device 10 includes a memory unit 12 that stores data and programs necessary for the calculation processing. The calculation unit 11 is configured with at least one hardware processor such as a CPU. The memory unit 12 is configured with a non-transitory storage medium such as an HDD or SSD. The calculation unit 11 estimates abnormalities occurring in the crops H in the area R based on the detection values.

[0020] The monitoring device 10 includes a communication unit 13 that communicates with the detection unit S via a network W. The communication unit 13 is a communication interface that can be connected to the network W via a wired or wireless connection. The communication unit 13 is controlled by the calculation unit 11. The monitoring device 10 includes a display unit 14 that outputs the calculation results calculated by the calculation unit 11 based on images. The display unit 14 is configured with a display device such as a liquid crystal display. The display unit 14 is controlled by the calculation unit 11. When the calculation unit 11 determines that an abnormality has occurred in the crop H, it outputs a countermeasure command to suppress the abnormality. The calculation unit 11 outputs information related to the countermeasure command to the display unit 14. The calculation unit 11 outputs the countermeasure command to the countermeasure unit M via the network W.

[0021] The countermeasure unit M is configured to execute measures to reduce abnormalities occurring in the crops H based on the countermeasure command output from the monitoring device 10. The countermeasure unit M may be equipment installed in the area R. The countermeasure unit M may be a mobile body that moves within the area R. The countermeasure unit M may be an autonomous vehicle or an autonomously flying drone device. The countermeasure unit M may not be required if a worker executes predetermined measures based on the countermeasure command. The countermeasure unit M includes, for example, a countermeasure device M1 that executes predetermined measures on the crops H in which an abnormality has occurred. The countermeasure device M1 executes one or more measures, including, for example, spraying pesticides, cutting the crops H, supplying water or fertilizer, opening and closing windows, opening and closing eaves, adjusting the air conditioning unit, etc.

[0022] The countermeasure unit M includes a control unit M2 that controls the countermeasure device M1. The control unit M2 acquires a countermeasure command and causes the countermeasure device M1 to execute a predetermined measure according to the content of the countermeasure command. The control unit M2 is configured with at least one hardware processor such as a CPU. The countermeasure unit M includes a communication unit M3 that communicates with the monitoring device 10 via the network W. The communication unit M3 is a communication interface that can be connected to the network W via a wired or wireless connection.

[0023] The countermeasure unit M is equipped with a moving device M4 that moves itself. The moving device M4 is controlled by the control unit M2. The moving device M4 is configured, for example, by a running device that enables travel. The moving device M4 may also be configured by a flying device that enables flight. The moving device M4 may not be required if the countermeasure unit M is installed in the area R. The countermeasure unit M is equipped with a position sensor M5 that detects its current position. The position sensor M5 is configured, for example, by a sensor that can detect position, such as a GPS sensor. The position sensor M5 may be configured not only by a GPS sensor, but also by other sensors that can detect markers or the like provided in the area R, as long as they can detect the position of the countermeasure unit M.

[0024] 3 and 4, the substance sensor Tn is formed by a fine nanomechanical sensor capable of detecting surface stress generated in response to a substance to be detected. The substance sensor Tn may be installed in any position during use. The substance sensor Tn is formed, for example, in a circuit formed on a substrate Te.

[0025] The substrate Te is formed of, for example, a silicon substrate. The substrate Te has, for example, a rectangular frame portion Tt. A rectangular opening Th is formed in the center of the frame portion Tt. A circular sensitive portion Tj is formed in the opening Th.

[0026] The sensitive part Tj includes a circular substrate part Ta. The substrate part Ta is formed, for example, in the shape of a disk with a diameter of approximately 300 μm. The substrate part Ta is formed of a silicon substrate, like the frame part. The substrate part Ta is formed with a thinner cross-sectional thickness than the frame part Tt. A sensitive layer Tb that is sensitive to the substance to be detected is formed on the upper surface of the substrate part Ta. The sensitive layer Tb is formed of a material that mechanically expands or contracts when the substance to be detected is sorbed. The expansion or contraction of the sensitive layer Tb is a reversible reaction; when a carrier gas is supplied, the substance is removed and the sensitive layer Tb returns to its original shape. Even when a carrier gas is not used, the substance sorbed on the sensitive layer Tb is released into the atmosphere over time. As a result, the expanded or contracted sensitive layer Tb returns to its original shape.

[0027] The sensitive layer Tb is formed of an organic material such as polymethyl methacrylate, polyvinylidene fluoride, polystyrene, or polycaprolactone. The sensitive layer Tb may be formed of not only an organic material but also an inorganic material, a metal material, or a biomaterial. The substrate portion Ta elastically deforms so as to bulge toward or opposite the sensitive layer Tb based on the expansion or contraction of the sensitive layer Tb. The substrate portion Ta returns to its original shape as the sensitive layer Tb returns to its original state from the expanded or contracted state.

[0028] The sensitive part Tj is supported on the frame part Tt by four support parts Tc. The support parts Tc are formed to have approximately the same cross-sectional thickness as the substrate part Ta. The support parts Tc elastically deform in response to deformation occurring in the substrate part Ta. A piezoresistor Td is provided inside the support part Tc. The piezoresistor Td may be provided on the upper surface side of the support part Tc. The piezoresistor Td elastically deforms in response to deformation occurring in the support part Tc. The electrical resistance value of the piezoresistor Td changes in response to deformation. An electrical path Tf is electrically connected to the piezoresistor Td.

[0029] The electrical path Tf is electrically connected to a plurality of piezoresistors Td so as to form a bridge circuit. A pair of input parts Ts for inputting a bias voltage is electrically connected to the electrical path Tf. A pair of output parts Tg for outputting a signal voltage is electrically connected to the electrical path Tf. The signal voltage input to the input parts Ts changes in voltage value in accordance with changes in the electrical resistance value of the piezoresistors Td, and is output from the output part Tg. In other words, the output voltage output from the output part Tg changes in accordance with the degree of deformation of the sensitive part Tj.

[0030] FIG. 5 shows an example of a change over time in the detection value detected by the sensor Sn. FIG. 5 shows a graph (time series waveform) of the change over time in the detection value in a normal state and a graph (time series waveform) of the change over time in the detection value in an abnormal state. The illustrated time series waveform is a plot of the detection values ​​detected by each sensor Sn, and visualizes the difference between the abnormal state and the normal state. The sensor Sn detects volatile substances produced by crops. Crops constantly emit volatile substances such as leaf alcohol. When an abnormal state occurs, such as damage by pests, the balance of volatile substance emission from crops changes over time.

[0031] In this embodiment, the sensor Sn detects a characteristic change over time in the detection value of the leaf alcohol produced by the crop when the crop is damaged by a pest. By monitoring the characteristic change over time in the detection value of the leaf alcohol, an abnormal state such as stress experienced by the crop H can be detected.

[0032] The detection values ​​include a predetermined waveform for a normal state (normal waveform) that shows a change over time when volatile substances are produced from crops in a normal state. The detection values ​​include a predetermined waveform for an abnormal state (abnormal waveform) that shows a change over time when volatile substances are produced from crops in an abnormal state. The predetermined waveform of the detection values ​​includes a rising portion (first characteristic portion 51) where the numerical value rises sharply. The predetermined waveform of the detection values ​​includes a decreasing portion (second characteristic portion 52) where the numerical value decreases from the rising portion. The predetermined waveform of the detection values ​​includes a falling portion (third characteristic portion 53) where the numerical value drops sharply from the decreasing portion.

[0033] In the predetermined waveform in an abnormal state, the rate of decrease in the second characteristic portion 52 is lower and the amount of volatile substance emitted is increased compared to the predetermined waveform in a normal state. The monitoring system 1 detects changes in the volatile substance over time with the detection unit S, and determines whether an abnormal state has occurred in the crops H grown in the area R to be managed based on the detected values ​​with the monitoring device 10. The calculation unit 11 acquires, for example, the detection values ​​detected for each different block in the area R. The calculation unit 11 monitors changes in the time-series waveform of the detection values ​​and determines whether a change in the state of the crops has occurred.

[0034] For example, if the time-series waveform of the detection value exceeds a preset threshold range, the calculation unit 11 determines that a change in the state of the crop has occurred compared to the normal state (see FIG. 5). When the calculation unit 11 acquires a detection value indicating a change in the state of the crop, it determines whether the detection value contains a feature of a predetermined state. The predetermined state is a measurable factor indicating an abnormal state. If the detection value contains a feature of the predetermined state, the calculation unit 11 estimates the block to be an abnormal region where an abnormality has occurred.

[0035] The monitoring device 10 communicates with multiple sensors Sn via a network W and acquires multiple detection values. The calculation unit 11 observes the multiple detection values ​​over time. The calculation unit 11 detects state changes occurring in the crop H based on the observed detection values ​​in the area R where the crop H is cultivated. The calculation unit 11 can also normalize the detection values. For example, the calculation unit 11 estimates abnormalities occurring in the crop H in the area R where the crop H is present.

[0036] The calculation unit 11 determines whether or not a predetermined state is included in detection values ​​detected from two of two or more blocks included in the region R. The calculation unit 11 acquires a first detection value that detects a state change occurring in a crop at a first position in a first of the two blocks, and a second detection value that detects a state change at a second position in a second block. The calculation unit 11 compares the first detection value and the second detection value in time series, and determines whether or not the predetermined state is included in the first detection value and the second detection value. For example, in determining a feature value that indicates a predetermined state, the calculation unit 11 extracts a time series waveform that indicates a change over time included in the detection value. The calculation unit 11 determines whether or not there is a correlation between the extracted multiple time series waveforms.

[0037] The calculation unit 11, for example, individually calculates a cross-correlation function indicating the correlation between two time-series waveforms included in the extracted multiple time-series waveforms. The calculation unit 11 normalizes the cross-correlation function and calculates a cross-correlation coefficient indicating the linear relationship between the two time-series waveforms. The calculation unit 11 normalizes the calculated multiple cross-correlation coefficients to smooth out variations in the detected values ​​of the time-series waveforms. The calculation unit 11 may calculate an approximation of the time derivative of the cross-correlation coefficient based on finite differences to numerically calculate changes over time in the cross-correlation coefficient.

[0038] The calculation unit 11 calculates a cross-similarity indicating the similarity between the time-series waveforms of different blocks based on the calculation results of the cross-correlation coefficient. The calculation unit 11 may normalize the calculated cross-correlation coefficient and calculate an approximation of the time derivative of the normalized cross-correlation coefficient based on finite differences, and calculate the cross-similarity based on the calculation results. The calculation unit 11 calculates an abnormal region where an abnormality has occurred in the block based on the result of comparing the calculated cross-similarity with a preset threshold.

[0039] The calculation unit 11 performs cluster analysis, for example, to group the calculated mutual similarity data set based on the positions of detection units with similar values. The calculation unit 11 calculates an abnormal area based on the clustering results of the calculated mutual similarity data set. The calculation unit 11 monitors the temporal progression of the abnormal area in a block where an abnormal crop is present. The calculation unit 11 predicts an abnormal area that will occur in the future and the time of occurrence based on the monitoring results of the temporal progression of the abnormal area in the block. When the calculation unit 11 calculates an abnormal area that will occur in the future, it outputs a countermeasure command to suppress the abnormality in the calculated abnormal area. The countermeasure command will be described later.

[0040] The calculation unit 11 may, for example, determine whether or not a predetermined waveform is included in the normalized detection value of each sensor Sn, and calculate an abnormal region where an abnormality occurs in the block. For example, the calculation unit 11 compares the time-series data of the detection value with a preset waveform using a determination method such as pattern matching.

[0041] The calculation unit 11 is configured to be able to extract a predetermined state from the detected value by, for example, executing machine learning such as deep learning based on a neural network using teacher data in advance. The predetermined state is analyzed by the calculation unit 11 based on the detected value.

[0042] The calculation unit 11 is configured to execute a process of extracting a normal waveform from a detected value by, for example, executing machine learning based on data of a normal waveform in advance. The calculation unit 11 is configured to execute a process of extracting an abnormal waveform from a detected value by, for example, executing machine learning based on data of an abnormal waveform in advance.

[0043] The calculation unit 11 determines, for example, whether the detected value includes a first characteristic portion 51, a second characteristic portion 52, and a third characteristic portion 53. If the detected value includes the first characteristic portion 51, the second characteristic portion 52, and the third characteristic portion 53, the calculation unit 11 determines that the detected value is a predetermined waveform. If the calculation unit 11 extracts a predetermined waveform from the detected value, it classifies the predetermined waveform into a normal waveform and an abnormal waveform. For example, the calculation unit 11 determines whether the gradient of the second characteristic portion 52 included in the waveform is within a first range that is a criterion for a normal waveform. If the gradient of the second characteristic portion 52 is within the first range that is a criterion for a normal waveform, the calculation unit 11 determines that the predetermined waveform is a normal waveform.

[0044] If the gradient of the second feature portion 52 is outside a first range that is the criterion for a normal waveform, the calculation unit 11 determines whether the gradient of the second feature portion 52 is within a second range that is the criterion for an abnormal waveform. If the gradient of the second feature portion 52 is within the second range that is the criterion for an abnormal waveform, the calculation unit 11 determines that the predetermined waveform is an abnormal waveform. If no detection value is obtained from the detection unit S within a predetermined period that is set in advance, the calculation unit 11 changes from a normal mode in which calculations can be performed to a sleep mode that reduces power consumption. If the calculation unit 11 receives a detection value, it changes from the sleep mode to the normal mode and performs calculations.

[0045] As shown in Fig. 6, when an abnormal waveform is detected in the detection value of sensor Sk (k is a natural number equal to or less than n) in a first block included in region R, calculation unit 11 determines that an abnormality has occurred in the crops H in block Bk that includes the position of sensor Sk. Calculation unit 11 determines whether or not an abnormal waveform is detected in the detection value of sensor Sm (m is a natural number equal to or less than n and different from k) in a second block included in region R that is installed in a different position from sensor Sk. When an abnormal waveform is detected in the detection value of sensor Sm, calculation unit 11 determines that an abnormality has occurred in the crops H in block Bm that includes the position of sensor Sm.

[0046] The calculation unit 11 estimates the propagation state of the abnormality in the region R, for example, based on the relationship between the positions and time of two blocks in which an abnormality has occurred in the crop H. The calculation unit 11 estimates the propagation state of the abnormality occurring in the crop H in the region R, for example, based on the appearance pattern of a predetermined waveform included in the detection value of the sensor Sk and the detection value of the sensor Sm. The appearance pattern is the state of the abnormality estimated based on parameters including detected values ​​and calculated values ​​such as time, position, number of appearances, propagation speed, type of abnormality, and scale of the abnormality.

[0047] For example, the calculation unit 11 sets the detection value of the sensor Sk as the first detection value and the detection value of the sensor Sm as the second detection value. The first detection value and the second detection value may be set arbitrarily and may be interchanged. When there are two or more detection values ​​including the predetermined waveform, the calculation unit 11 selects two first detection values ​​and two second detection values ​​from among the detection values ​​of the multiple sensors Sn so as to minimize the distance between the first position and the second position. When there are two or more detection values ​​including the predetermined waveform, the calculation unit 11 sets one or more combinations of the first detection value and the second detection value.

[0048] The calculation unit 11 determines whether the first detection value and the second detection value contain an abnormal waveform indicating a time-series change of a predetermined substance such as a volatile substance. If the first detection value and the second detection value contain an abnormal waveform, the calculation unit 11 sets a block Bk including a first position where the first detection value was detected as a first abnormal region where an abnormality has occurred, and sets a block Bm including the second position as a second abnormal region where an abnormality has occurred. The calculation unit 11 sets a group of blocks including the first abnormal region and the second abnormal region as an abnormal region BF. The calculation unit 11 compares the first detection value and the second detection value in time series.

[0049] The calculation unit 11 compares a first measurement time at which an abnormal waveform included in the first detection value appears with a second measurement time at which an abnormal waveform included in the second detection value appears. Based on the comparison result between the first measurement time and the second measurement time, the calculation unit 11 calculates a propagation state of the abnormality propagating from the first abnormal region to the second abnormal region. Based on the comparison result between the first measurement time and the second measurement time, for example, the calculation unit 11 calculates a propagation state including the direction and propagation speed of the abnormality based on the first measurement time at the first position at which the abnormal waveform appeared and the second measurement time at the second position at which the abnormal waveform appeared.

[0050] The calculation unit 11 calculates the direction in which the abnormality propagates based on the first measurement time, the second measurement time, and the positional relationship between the first position and the second position. The calculation unit 11 calculates the propagation speed of the abnormality based on the distance between the first position and the second position and the difference between the first measurement time and the second measurement time. In the illustrated example, when the first measurement time is earlier than the second measurement time, the abnormality propagates from block Bk, which is the first abnormal area, in the direction toward block Bm, which is the second abnormal area. When the first measurement time is later than the second measurement time, the abnormality propagates from block Bm, which is the second abnormal area, in the direction toward block Bk, which is the first abnormal area.

[0051] The calculation unit 11 may calculate an abnormality propagation region to which an abnormality will propagate from the current abnormality region BF in the future based on the calculated propagation state of the abnormality. The calculation unit 11 calculates an abnormality propagation region to which an abnormality will propagate from the abnormality region in the future based on, for example, the direction and propagation speed of the abnormality in the abnormality region BF. The calculation unit 11 calculates the abnormality propagation region based on an abnormality propagation model based on a theoretical calculation formula that can estimate the abnormality propagation region using parameters such as the direction, propagation speed, and propagation area of ​​the abnormality from the abnormality region BF. Any theoretical calculation formula may be used for the abnormality propagation model as long as it can calculate the abnormality propagation region. The calculation unit 11 may calculate the abnormality propagation region based on machine learning using data of abnormalities propagating from past abnormality regions BF as training data.

[0052] As shown in FIG. 7, the calculation unit 11 outputs a region display image 15 showing the current abnormal region BF and abnormal propagation region E to the display unit 14. For example, the calculation unit 11 sets a region including a group of blocks existing around block Bk, which is the first abnormal region, as the first abnormal propagation region E1. The calculation unit 11 generates a region display image 15 showing the first abnormal propagation region E1. For example, the calculation unit 11 sets a region including a group of blocks existing around block Bm, which is the second abnormal region, as the second abnormal propagation region E2. The calculation unit 11 generates a region display image 15 showing the second abnormal propagation region E2. The calculation unit 11 sets the combined region of the first abnormal propagation region E1 and the second abnormal propagation region E2 as the abnormal propagation region E.

[0053] The calculation unit 11 calculates the degree of propagation of the abnormal state for the first abnormal propagation region E1 and the second abnormal propagation region E2 included in the abnormal propagation region E. The calculation unit 11 performs calculations such that, for example, the earlier the measurement time at which the abnormality is detected, the greater the degree of propagation of the abnormal state. The calculation unit 11 may assign different weights to the first abnormal propagation region E1 and the second abnormal propagation region E2 based on the magnitude of the propagation degree of the abnormal state. Based on the different weights, the calculation unit 11 may display the first abnormal propagation region E1 and the second abnormal propagation region E2 in different ways, including by using different colors, shapes, patterns, and text information, in the region display image 15 or a display image described below.

[0054] 8, when the calculation unit 11 calculates the abnormality propagation region E, the calculation unit 11 displays an information display image 16 including information about the abnormality propagation region E on the display unit 14. The information display image 16 includes information such as the region R to be monitored, the type of crop H being cultivated in the region R, the block in which the abnormality occurred, the abnormality propagation region where the abnormality is estimated to be propagating, the details of the abnormality that has occurred, and countermeasures for the abnormality. The information display image 16 may be displayed based on an input operation by the user, such as clicking on the image of a block shown in the region display image 15.

[0055] As shown in FIG. 9, the calculation unit 11 outputs a command display image 17 showing a countermeasure command for suppressing the abnormality in the abnormal region (abnormal block) and the abnormality propagation region E. For example, the calculation unit 11 outputs a countermeasure appropriate to the crop, such as spraying pesticides or cutting down, to the command display image 17 according to the first abnormality propagation region E1 and the second abnormality propagation region E2. The calculation unit 11 may automatically calculate a countermeasure for the first abnormality propagation region E1 and a countermeasure for the second abnormality propagation region E2 based on the calculated magnitude of the propagation degree of the abnormal state, and output the calculated countermeasures to the command display image 17. The calculation unit 11 may arbitrarily change the countermeasure for the first abnormality propagation region E1 and the countermeasure for the second abnormality propagation region E2 in the command display image 17 based on an input operation by the user.

[0056] The calculation unit 11 causes the countermeasure unit M, which is configured to execute predetermined measures to suppress the abnormality that has occurred in the crop H, to execute the countermeasure based on the countermeasure command. For example, when pest damage occurs to the crop H, the calculation unit 11 transmits information related to the countermeasure command to the countermeasure unit M based on the magnitude of the propagation degree of the abnormal state. When the magnitude of the propagation degree of the abnormal state is within a first threshold range classified as mild, the calculation unit 11 dispatches the countermeasure unit M to the abnormal area and the abnormal propagation area, and causes the countermeasure unit M to spray a chemical such as a pesticide in the abnormal area and the abnormal propagation area.

[0057] The countermeasure unit M autonomously drives or flies in the abnormal area and the abnormal propagation area based on information regarding the countermeasure command. The countermeasure unit M sprays pesticides in the abnormal area and the abnormal propagation area. If the degree of propagation of the abnormal state is within a second threshold range classified as severe, the calculation unit 11 may dispatch the countermeasure unit M to the abnormal area and the abnormal propagation area and have the countermeasure unit cut down the crops H included in the abnormal area and the abnormal propagation area. The calculation unit 11 may select the content of the countermeasure depending on the type of abnormality occurring in the crops H, such as not only damage caused by pests but also virus propagation, environmental changes, etc. The countermeasure taken by the countermeasure unit M can prevent the abnormality occurring in the crops H from spreading from the abnormal propagation area to other normal blocks Bn.

[0058] 10 shows the process flow of a monitoring method for calculating abnormalities occurring in crops based on mutual similarity, which is executed by the monitoring device 10. The monitoring method is executed based on a computer program installed on a computer mounted on the monitoring device in an area where crops are present. The computer program causes the monitoring device 10 to perform the following processes.

[0059] The calculation unit 11 acquires a detection value for each of a plurality of different blocks in the region R (step S10). The calculation unit 11 determines whether the detection value includes a predetermined feature amount that indicates a state change occurring in the crop (step S12). If the detection value includes the feature amount, the calculation unit 11 estimates that the block is an abnormal region where an abnormality has occurred (step S14).

[0060] 11 shows the process flow of a monitoring method for calculating abnormalities occurring in crops based on mutual similarity, which is executed by the monitoring device 10. The monitoring method is executed based on a computer program installed on a computer mounted on the monitoring device in an area where crops are present. The computer program causes the monitoring device 10 to perform the following processes.

[0061] The calculation unit 11 acquires detection values ​​indicating changes in the state of crops present in multiple different blocks in the region R (step S20). The calculation unit 11 calculates the mutual similarity of the time-series waveforms of the detection values ​​detected in each different block (step S22). The calculation unit 11 calculates an abnormal region in the block where an abnormality has occurred based on the mutual similarity (step S24). The calculation unit 11 predicts an abnormal region that will occur in the future and the time of occurrence based on the monitoring results of the temporal transition of the abnormal region in the block (step S26). The calculation unit 11 outputs a countermeasure command to suppress the abnormality in the calculated abnormal region (step S28).

[0062] 12 shows the process flow of a monitoring method executed by the monitoring device 10 for calculating abnormalities occurring in crops based on a predetermined waveform. The monitoring method is executed based on a computer program installed in a computer mounted on the monitoring device in an area where crops are present. The computer program causes the monitoring device 10 to execute the following processes.

[0063] The calculation unit 11 acquires a first detection value indicating a state change occurring in a crop at a first position in a first block in the region R (step S100). The calculation unit 11 acquires a second detection value indicating a state change at a second position in a second block in the region R (step S102). The calculation unit 11 compares the first detection value and the second detection value in chronological order (step S104). The calculation unit 11 determines whether the first detection value and the second detection value include a predetermined state (step S106).

[0064] When the first detection value and the second detection value include a predetermined state, the calculation unit 11 determines whether or not an abnormality has occurred in the crop H based on the appearance of the predetermined state in the region R (step S108). When the calculation unit 11 determines that an abnormality has occurred in the crop H, it outputs a countermeasure command to suppress the abnormality (step S110).

[0065] As described above, the monitoring system 1 can estimate an abnormality occurring in the crop H based on the detection values ​​detected by the detection unit S. The monitoring system 1 can estimate the propagation state of the abnormality occurring in the crop H by monitoring the first detection value and the second detection value. The monitoring system 1 can cause the countermeasure unit M to take countermeasures depending on the nature of the abnormality occurring in the crop H. The monitoring system 1 can estimate an abnormality that will occur in the region R in the future by comparing the first detection value and the second detection value in chronological order and calculating the propagation state of the abnormality. The monitoring system 1 can realize a labor-saving smart farm by managing the region R based on the detection values ​​detected by the detection unit S.

[0066] The following describes modified examples of the monitoring system 1. In the following description, the same names and symbols are used for components having similar functions to those in the above embodiment, and duplicate descriptions will be omitted as appropriate.

[0067] [Variation 1] The sensor Sn may be configured to acquire substances or detection values ​​other than volatile substances produced by crops. In the sensor Sn, the substance sensor Tn may be configured to detect substances other than leaf alcohol. For example, the substance sensor Tn may not only detect volatile substances in the atmosphere as described above, but may also detect predetermined substances released from crops into the ground or water in the event of an abnormality. The substance sensor Tn may detect predetermined substances harmful to crops in the ground or water in the event of an abnormality in the crops. The substance sensor Tn may be configured to detect predetermined substances related to abnormalities in the crops H.

[0068] The calculation unit 11 may acquire first and second detection values ​​from the two substance sensors, compare the first and second detection values ​​in time series, and determine whether or not a predetermined state indicating the occurrence of an abnormality is included in the first and second detection values. The calculation unit 11 may determine whether or not an abnormality has occurred in the crop based on the appearance of the predetermined state in the area. The predetermined state may include a change over time in a predetermined substance related to the abnormality of the crop H. The calculation unit 11 may be configured to be able to determine the appearance of the predetermined state based on the change over time in the predetermined substance based on machine learning.

[0069] [Variation 2] The sensor Sn may be configured to acquire a detection value occurring in the crop when an abnormality occurs. The sensor Sn may detect a predetermined factor related to an abnormality in the crop, such as a change in the potential occurring in the crop, a change in a specific substance in the crop, or an environmental factor that causes an abnormality in the crop in the region R. The environmental factor includes factors that cause an abnormality in the crop, such as a change in temperature, a change in the amount of water, a state of nutrients, a change in the amount of light, or a change in the amount of wind.

[0070] The calculation unit 11 may acquire first and second detection values ​​from the two sensors Sn, compare the first and second detection values ​​in time series, and determine whether the first and second detection values ​​include a predetermined state indicating the occurrence of an abnormality. The predetermined state may include a change over time in a predetermined factor related to a crop abnormality. The calculation unit 11 may acquire first and second detection values ​​from the two substance sensors, compare the first and second detection values ​​in time series, and determine whether the first and second detection values ​​include a predetermined state indicating the occurrence of an abnormality.

[0071] The calculation unit 11 may determine whether or not an abnormality has occurred in the crop based on the appearance of a predetermined state in the region. The calculation unit 11 may be configured to be able to determine the appearance of a predetermined state based on a change over time in a predetermined factor related to an abnormality in the crop based on machine learning.

[0072] [Variation 3] The sensor Sn may detect the occurrence of a predetermined state including factors other than the crop, such as sound waves or infrared rays, which may affect pests or animals that affect crops. The calculation unit 11 may acquire first and second detection values ​​from the two sensors Sn, compare the first and second detection values ​​in time series, and determine whether the first and second detection values ​​include a predetermined state indicating the occurrence of an abnormality. The predetermined state includes factors other than the crop that affect the crop.

[0073] The calculation unit 11 may acquire first and second detection values ​​from the two substance sensors, compare the first and second detection values ​​in time series, and determine whether the first and second detection values ​​include a predetermined state indicating the occurrence of an abnormality. The calculation unit 11 may determine whether an abnormality has occurred in the crop based on the appearance of the predetermined state in the area.

[0074] [Variation 4] The sensor Sn may be configured with a camera. The sensor Sn may, for example, capture an image of the crop and acquire a predetermined condition of the crop, such as deformation, discoloration, poor growth, damage by pests, damage by animals, damage by viruses, damage by weather, or proliferation of unintended plants, based on the captured image. The calculation unit 11 may be configured to be able to determine, based on machine learning using the captured image, how a predetermined condition appears based on a change over time in a predetermined factor related to an abnormality in the crop.

[0075] The calculation unit 11 may acquire first detection values ​​based on the captured images and second detection values ​​based on the captured images from the two sensors Sn, compare the first detection values ​​with the second detection values ​​in time series, and determine whether or not the first detection values ​​and the second detection values ​​include a predetermined state indicating the occurrence of an abnormality. The calculation unit 11 may determine whether or not an abnormality has occurred in the crop based on the appearance of the predetermined state in the area.

[0076] The calculation unit 11 according to each of the above-described modified examples may be configured to determine whether or not an abnormality has occurred in the crops based on the appearance of a predetermined state corresponding to different detection values ​​in the region. The countermeasure unit M according to each of the above-described modified examples may be configured to adjust the amount of water, nutrients, light, temperature, air volume, etc. for the crops in the region R. The countermeasure unit M may be configured to repel animals harmful to the crops based on sound waves, light emission, vibration, movement, or morphological changes.

[0077] In the above-described embodiment, the computer program executed in each component of the monitoring device 10 may be provided in a form recorded on a computer-readable, portable, non-transitory recording medium such as a semiconductor memory, a magnetic recording medium, or an optical recording medium. [Explanation of symbols]

[0078] 1 monitoring system, 10 monitoring device, 11 calculation unit, 12 memory unit, 13 communication unit, 14 display unit, 15 area display image, 16 information display image, 17 command display image, 51 first feature unit, 52 second feature unit, 53 third feature unit, BF abnormal area, Bk, Bm, Bn blocks, Cn communication unit, Dn memory unit, E abnormal propagation area, E1 first abnormal propagation area, E2 second abnormal propagation area, Gn position sensor, H crop, Kn supply unit, M countermeasure unit, M1 countermeasure device, M2 control unit, M3 communication unit, M4 moving device, M5 position sensor, Pn control unit, R area, S detection unit, Sk, Sm, Sn sensor, Ta substrate unit, Tb sensitive layer, Tc support unit, Td piezoresistor, Te substrate, Tf electrical path, Tg Output section, Th opening, Tj sensitive section, Tn material sensor, Ts input section, Tt frame section, W network

Claims

1. a calculation unit that estimates an abnormality occurring in a crop in an area where the crop is present, The calculation unit obtaining a first detection value that detects a state change occurring in the crop at a first position in a first block among two or more blocks included in the area, and a second detection value that detects the state change at a second position in a second block included in the area; comparing the first detection value with the second detection value in time series; determining whether the first detection value and the second detection value include a predetermined state; If the predetermined state is included, the block is estimated to be an abnormal area where the abnormality occurs. monitoring equipment.

2. The calculation unit calculating a mutual similarity between time-series waveforms included in the first detection value and the second detection value; calculating whether the predetermined state is included or not based on the mutual similarity; The monitoring device of claim 1 .

3. The calculation unit calculating a cross-correlation coefficient indicating a correlation between the time-series waveforms of different blocks; calculating the mutual similarity based on the calculation result of the cross-correlation coefficient; The monitoring device according to claim 2 .

4. The calculation unit normalizing the cross-correlation coefficients; and / or calculating an approximation of the time derivative of the cross-correlation coefficient based on finite differences; calculating the mutual similarity based on the calculation result; The monitoring device according to claim 3.

5. The calculation unit predicting the abnormal region that will occur in the future and the time of occurrence based on the transition of the abnormal region over time; A monitoring device according to any one of claims 1 to 4.

6. The calculation unit When the abnormality region that will occur in the future is calculated, a countermeasure command for suppressing the abnormality in the calculated abnormality region is output. The monitoring device according to claim 5.

7. The calculation unit calculating the abnormal region based on the clustering result of the mutual similarity; A monitoring device according to any one of claims 2 to 4.

8. The calculation unit determining whether or not the abnormality has occurred in the crop based on the appearance of the predetermined state in the region; The monitoring device of claim 1 .

9. The calculation unit calculating the abnormal region based on a first measurement time at which the predetermined state included in the first detection value appears and a second measurement time at which the predetermined state included in the second detection value appears; Calculating a propagation state of the anomaly propagating from the anomaly region; The monitoring device according to claim 8.

10. The calculation unit calculating an anomaly propagation region to which the anomaly will propagate from the anomaly region in the future based on the propagation state; The monitoring device of claim 9.

11. The calculation unit acquiring the first detection value for detecting a predetermined substance produced from the crop and the second detection value for detecting the predetermined substance; comparing the first detection value with the second detection value in time series; determining whether or not the first detection value and the second detection value contain a predetermined waveform that indicates a time-series change of the predetermined substance; estimating the propagation state based on the appearance of the predetermined waveform; The monitoring device of claim 10.

12. The calculation unit obtaining the first detection value detecting a volatile substance generated from the crop and the second detection value detecting the volatile substance; determining whether the first detection value and the second detection value contain the predetermined waveform of the volatile substance; The monitoring device of claim 11.

13. The calculation unit calculating the propagation state including a direction and a propagation speed of the anomaly based on the first position where the predetermined waveform occurs and the second position where the predetermined waveform occurs; Calculating the abnormal propagation region and occurrence time based on the calculation result; 13. The monitoring device of claim 12.

14. The calculation unit obtaining a detection value of the volatile substances including leaf alcohol emitted from the crop; determining whether the detected value includes the predetermined waveform; 13. The monitoring device of claim 12.

15. The calculation unit When the abnormal propagation region is calculated, outputting a countermeasure command for suppressing the abnormality to the abnormal region and the abnormality propagation region; The monitoring device of claim 11.

16. The calculation unit a countermeasure unit configured to execute a predetermined measure to suppress the abnormality based on the countermeasure command, and spraying an agent into the abnormal area and the abnormality propagation area; 16. The monitoring device of claim 15.

17. The calculation unit causing the countermeasure unit to cut down the crops included in the abnormal area and the abnormality propagation area based on the countermeasure command; 17. The monitoring device of claim 16.

18. a detection unit that detects detection values ​​that indicate state changes occurring in the crops present in a plurality of different blocks in an area where the crops are present; and a calculation unit that acquires a first detection value that detects a state change occurring in the crop at a first position of a first block among two or more blocks included in the area and a second detection value that detects the state change at a second position of a second block included in the area, compares the first detection value and the second detection value in time series, and determines whether or not a predetermined state is included in the first detection value and the second detection value, and if the predetermined state is included in the detection value, estimates the block to be an abnormal area where an abnormality is occurring. Surveillance system.

19. A monitoring method executed by a computer mounted on a monitoring device that calculates abnormalities occurring in crops in an area where the crops are present, comprising: The computer obtaining a first detection value that detects a state change occurring in the crop at a first position in a first block among two or more blocks included in the area, and a second detection value that detects the state change at a second position in a second block included in the area; comparing the first detection value with the second detection value in time series; determining whether the first detection value and the second detection value include a predetermined state; If the predetermined state is included, the block is estimated to be an abnormal area where an abnormality occurs. Execute the process, Monitoring method.

20. A program installed on a computer mounted on a monitoring device that calculates abnormalities occurring in crops in an area where the crops are present, obtaining a first detection value that detects a state change occurring in the crop at a first position in a first block among two or more blocks included in the area, and a second detection value that detects the state change at a second position in a second block included in the area; comparing the first detection value with the second detection value in time series; determining whether the first detection value and the second detection value include a predetermined state; If the predetermined state is included, the computer is caused to execute a process of estimating that the block is an abnormal area where an abnormality is occurring. program.

Images