Information processing device, information processing method, and program

Abstract

To detect signs of abnormalities early based on the results of monitoring the monitored object. [Solution] The information processing device includes an acquisition unit that acquires sensor data from sensors that monitor the monitored object, a generation unit that generates an abnormality determination model for determining in advance whether an abnormality will occur in the monitored object based on multiple pieces of sensor data acquired by the acquisition unit, and a detection unit that detects signs of an abnormality that may occur in the monitored object based on the abnormality determination model generated by the generation unit when the value of the sensor data indicates an abnormal value.

Description

[Technical Field]

[0001] The present disclosure relates to an information processing device, an information processing method, and a program. [Background technology]

[0002] There are known techniques for monitoring a target and detecting abnormalities. For example, Patent Document 1 discloses a technique for detecting abnormalities in a product manufacturing process. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2020-170327 Summary of the Invention [Problem to be solved by the invention]

[0004] Disasters such as earthquakes and heavy rains can cause great damage. For this reason, it is desirable to monitor the target and detect signs of disasters early to prevent great damage from disasters.

[0005] The present disclosure aims to provide an information processing device, an information processing method, and a program that are capable of early detection of signs of an abnormality based on the monitoring results of a monitoring target. [Means for solving the problem]

[0006] The information processing device of the present disclosure includes an acquisition unit that acquires sensor data from a sensor that monitors a monitored object, a generation unit that generates an abnormality determination model for determining in advance an abnormality that may occur in the monitored object based on the multiple pieces of sensor data acquired by the acquisition unit, and a detection unit that detects signs of an abnormality that may occur in the monitored object based on the abnormality determination model generated by the generation unit when the value of the sensor data indicates an abnormal value.

[0007] The information processing method disclosed herein includes the steps of acquiring sensor data from a sensor monitoring a monitored object, generating an abnormality determination model based on the acquired plurality of pieces of sensor data to determine in advance whether an abnormality will occur in the monitored object, and detecting, based on the generated abnormality determination model, signs of an abnormality that may occur in the monitored object when the value of the sensor data indicates an abnormal value.

[0008] The program disclosed herein causes a computer to perform the following steps: acquiring sensor data from a sensor monitoring a monitored object; generating an abnormality determination model based on the acquired multiple pieces of sensor data to determine in advance whether an abnormality will occur in the monitored object; and detecting, based on the generated abnormality determination model, signs of an abnormality that may occur in the monitored object when the value of the sensor data indicates an abnormal value. [Effects of the Invention]

[0009] According to the present disclosure, signs of an abnormality can be detected early based on the monitoring results of the monitoring target. [Brief explanation of the drawings]

[0010] [Figure 1] FIG. 1 is a block diagram showing an example of the configuration of an information processing system according to an embodiment. [Figure 2] FIG. 2 is a diagram for explaining the flow of processing in the information processing system according to the embodiment. [Figure 3] FIG. 3 is a block diagram illustrating an example of the configuration of the information processing device according to the embodiment. [Figure 4] FIG. 4 is a diagram illustrating an example of sensor data according to the embodiment. [Figure 5] FIG. 5 is a diagram for explaining an example of a method for calculating a correlation according to the embodiment. [Figure 6] FIG. 6 is a diagram for explaining an example of an evaluation method according to the embodiment. [Figure 7]FIG. 7 is a diagram for explaining a first method for detecting a sign of an abnormality that may occur in a monitoring target according to the embodiment. [Figure 8] FIG. 8 is a diagram for explaining a second method for detecting a sign of an abnormality that may occur in a monitoring target according to the embodiment. [Figure 9] FIG. 9 is a diagram showing signal levels of a plurality of sensors according to the embodiment. [Figure 10] FIG. 10 is a diagram for explaining a third method for detecting a sign of an abnormality that may occur in a monitoring target according to the embodiment. [Figure 11] FIG. 11 is a diagram for explaining a fourth method for detecting a sign of an abnormality that may occur in a monitoring target according to the embodiment. [Figure 12] FIG. 12 is a flowchart showing the flow of processing for generating an abnormality determination model according to the embodiment. [Figure 13] FIG. 13 is a flowchart showing the flow of a process for providing data to a user according to the embodiment. [Figure 14] FIG. 14 is a flowchart showing the flow of processing for detecting a sign of an abnormality that may occur in a monitoring target according to the embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0011] Hereinafter, embodiments of the present disclosure will be described in detail with reference to the accompanying drawings. Note that the present disclosure is not limited to these embodiments, and in the following embodiments, the same components are designated by the same reference numerals, and redundant description will be omitted.

[0012] [Embodiment] (Information Processing System) An example of the configuration of an information processing system according to an embodiment will be described with reference to Fig. 1. Fig. 1 is a block diagram showing an example of the configuration of an information processing system according to an embodiment.

[0013] As shown in Fig. 1, the information processing system 1 includes an information processing device 10 and a plurality of sensors 12. In the example shown in Fig. 1, sensors 12a, 12b, and 12c are shown as the sensors 12. The information processing device 10 and the plurality of sensors 12 are communicatively connected via a network N. The network N is, for example, the Internet, but is not limited to this.

[0014] The information processing device 10 can be realized by a server device such as a cloud server. The sensor 12 is a sensor that detects various physical quantities of a monitoring target and its surroundings. The sensor 12 detects, for example, time-series data of various physical quantities of a monitoring target and its surroundings.

[0015] The information processing system 1 is a system that can detect signs of abnormalities before an abnormality occurs in a monitored object based on time-series data of physical quantities detected by a plurality of sensors 12. This allows the information processing system 1 to detect signs in advance, for example, before an earthquake or heavy rain occurs, thereby preventing damage from spreading.

[0016] (Monitored) The following describes the monitoring targets of the information processing system 1. Note that the monitoring targets listed below are examples and do not limit the present disclosure.

[0017] (First surveillance target) In this embodiment, the first monitored object is a mountain or cliff slope, including a volcano. In this case, the sensor 12 may include, for example, a rainfall sensor that detects rainfall, an inclination sensor that detects the angle of the slope, a gas sensor that detects the concentration and composition of volcanic gases, a magnetic sensor that detects magnetism generated in rocks that changes with temperature and stress, a vibration sensor that detects earthquakes and tremors occurring around the object, a visible light camera that photographs the object, and an infrared camera that detects the temperature of the object. The sensor 12 is not limited to these. Examples of abnormalities in the first monitored object include mountainside deformation and landslides caused by volcanic eruptions, large earthquakes, and heavy rain. In other words, in the first monitored object, the information processing system 1 can, for example, early recognize precursor phenomena of large-scale landslides that occur when a volcano erupts, a large earthquake, or heavy rain occurs, based on various sensor data acquired by the sensor 12.

[0018] (Second Surveillance Target) In this embodiment, the second monitoring target is, for example, a water-sealed underground rock tank for storing crude oil, LP gas, or the like. In this case, the sensor 12 may include a rainfall sensor for detecting rainfall, a tide level sensor for detecting tide levels, a water volume sensor for detecting groundwater volume, a pressure sensor for detecting groundwater pressure, and the like. The sensor 12 is not limited to these. Furthermore, the information processing system 1 may pre-store data related to legal observation holes, such as the size of legal observation holes installed in the water-sealed underground rock tank for monitoring groundwater. An abnormality in the second monitoring target is, for example, a failure of the water-sealed underground rock tank. That is, for the second monitoring target, the information processing system 1 detects signs of failure of the water-sealed underground rock tank before the water-sealed underground rock tank actually fails, based on, for example, various sensor data acquired by the sensor 12 and data related to the legal observation holes.

[0019] In the following description, the information processing system 1 is assumed to detect a sign of a possible abnormality before any abnormality occurs in a monitored object.

[0020] (Outline of information processing system processing) An overview of the processing flow of the information processing system according to the embodiment will be described with reference to Fig. 2. Fig. 2 is a diagram for explaining the processing flow of the information processing system according to the embodiment.

[0021] The information processing system 1 performs a current analysis of the monitored object (step S1). For example, the information processing system 1 determines the selection of sensor data detected by the sensor 12 and an evaluation policy in order to generate a model for detecting signs of abnormalities that may occur in the monitored object.

[0022] The information processing system 1 generates an abnormality determination model for detecting a sign of an abnormality that may occur in the monitored object based on the result of the current state analysis of the monitored object (step S2).

[0023] The information processing system 1 sets an alert for sensor data detected by the multiple sensors 12 (step S3). The information processing system 1 acquires data detected from each of the multiple sensors 12 (step S4). The information processing system 1 detects an alert if the detected sensor data exceeds a predetermined threshold (step S5). Based on the detected alert, the information processing system 1 evaluates whether or not there is an abnormality in the detected sensor data (step S6). Based on the evaluation result, the information processing system 1 reconsiders the model generated in step S2 (step S7).

[0024] In the present disclosure, for example, by repeatedly executing steps S2 to S7, the accuracy of detecting signs of abnormalities that may occur in the monitored object is improved.

[0025] (Information processing device) An example of the configuration of an information processing device according to an embodiment will be described with reference to Fig. 3. Fig. 3 is a block diagram showing an example of the configuration of an information processing device according to an embodiment.

[0026] As shown in FIG. 3, the information processing device 10 includes a communication unit 20, a storage unit 22, and a control unit 24.

[0027] The communication unit 20 is a communication interface that performs communication between the information processing device 10 and various external devices. The communication unit 20 performs communication between the information processing device 10 and a plurality of sensors 12, for example. The communication unit 20 performs communication between the information processing device 10 and an external information server that provides various information such as weather information, for example. The communication unit 20 performs communication between the information processing device 10 and a terminal device used by a user, for example.

[0028] The storage unit 22 stores various types of information. The storage unit 22 stores information such as the contents of calculations performed by the control unit 24 and programs. The storage unit 22 includes at least one of a RAM (Random Access Memory), a main storage device such as a ROM (Read Only Memory), and an external storage device such as an HDD (Hard Disk Drive).

[0029] The storage unit 22 includes a sensor data storage unit 22a and a model storage unit 22b.

[0030] The sensor data storage unit 22a stores the sensor data acquired by the acquisition unit 30 from each of the plurality of sensors 12. Fig. 4 is a diagram showing an example of the sensor data according to the embodiment.

[0031] As shown in FIG. 4, the sensor data D1 includes items such as "date and time," "sensor ID," "detection item," and "detection result."

[0032] "Date and time" is time information when the sensor 12 detected a physical quantity related to the monitoring target. "Sensor ID" is identification information for identifying the sensor 12. "Detection item" is information indicating the type of physical quantity detected by the sensor 12. "Detection result" is information indicating the detection result of the sensor data detected by the sensor 12.

[0033] In Fig. 4, each item is conceptually described as "A01", but in reality, specific data is stored in each item. One example shown in Fig. 4 indicates that at date and time "A01", "C01" is detected by sensor 12 with sensor ID "B01", and the detection result is "D01". Sensor data D1 can be said to be time-series data of the detection result of sensor 12.

[0034] The model storage unit 22b stores the abnormality determination model generated by the generation unit 32.

[0035] The control unit 24 controls each unit of the information processing device 10. The control unit 24 has, for example, an information processing device such as a CPU (Central Processing Unit) or an MPU (Micro Processing Unit), and a storage device such as a RAM or a ROM. The control unit 24 executes a program that controls the operation of the information processing device 10 according to the present disclosure. The control unit 24 may be realized by, for example, an integrated circuit such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array). The control unit 24 may be realized by a combination of hardware and software.

[0036] The control unit 24 includes an acquisition unit 30, a generation unit 32, a detection unit 34, and a provision unit 36.

[0037] The acquisition unit 30 acquires various types of information. For example, the acquisition unit 30 acquires sensor data from each of the multiple sensors 12 via the communication unit 20. For example, the acquisition unit 30 acquires various types of information such as weather information from an external information server via the communication unit 20.

[0038] The generation unit 32 generates an abnormality determination model for determining an abnormality in a monitoring target. The generation unit 32 generates the abnormality determination model based on, for example, sensor data acquired by the acquisition unit 30 from the multiple sensors 12.

[0039] The generator 32 generates an anomaly determination model including, for example, a threshold value for the sensor data. The threshold value may be set based on the detected sensor data, or may be set in advance by the user. A specific example will be described later.

[0040] The generation unit 32 may generate the anomaly determination model by machine learning, for example, based on sensor data detected by the multiple sensors 12 over a predetermined period of time. In this case, the anomaly determination model can be said to be a machine learning model generated by the generation unit 32 through a learning process using machine learning. In the learning process, the generation unit 32 executes a process for learning the learning model. In the learning process, the generation unit 32 uses data detected by the multiple sensors 12 over a predetermined period of time as learning data to learn the anomaly determination model. The anomaly determination model learned by the generation unit 32 is also called a learned model.

[0041] In the present disclosure, the generation unit 32 may generate an anomaly determination model using various techniques, such as a neural network, a deep neural network, a recurrent neural network, a vector autoregression, a random forest, a gradient boosting decision tree, a convolutional neural network, a long short-term memory (LSTM), a gated recurrent units (GRU), a factor analysis model, or a Bayesian network.

[0042] For example, if the anomaly determination model is a neural network model, the anomaly determination model is composed of layers called an input layer, an intermediate layer, and an output layer, each of which includes a plurality of nodes. The plurality of nodes are connected via edges. Each layer has a function called an activation function, and each edge is weighted. The learning model includes one or more intermediate layers. In this case, the generation unit 32 learns the anomaly determination model by setting the number of intermediate layers, the number of nodes in each layer, the weight of each edge, etc.

[0043] The generation unit 32 may generate an abnormality determination model based on, for example, the degree of variation in sensor data for each of the multiple sensors 12. For example, the generation unit 32 calculates a standard deviation σ for each of the multiple sensor data based on sensor data detected by each of the multiple sensors 12 over a predetermined period. Then, for example, based on the calculated standard deviation σ, the generation unit 32 generates an abnormality determination model for each of the multiple sensors 12 that determines sensor data that falls within a range of a reference value ±3σ as normal and sensor data that falls outside the range of the reference value ±3σ as abnormal. In other words, the generation unit 32 generates an abnormality determination model for each of the multiple sensors 12, with an upper limit value set to a reference value +3σ and a lower limit value set to a reference value −3σ. A specific example of a method for calculating the standard deviation σ will be described later.

[0044] The generation unit 32 may generate an abnormality determination model based on, for example, the degree of variation in sensor data among the multiple sensors 12. For example, the generation unit 32 calculates a standard deviation σ tilde between the multiple sensor data based on sensor data detected by each of the multiple sensors 12 over a predetermined period. Then, for example, based on the calculated standard deviation σ tilde, the generation unit 32 generates an abnormality determination model that determines, among the multiple sensors 12, sensor data that falls within a range of a reference value ±3σ tilde as normal and sensor data that falls outside the range of the reference value ±3σ tilde as abnormal. In other words, the generation unit 32 generates an abnormality determination model among the multiple sensors 12, in which the upper limit is the reference value +3σ tilde and the lower limit is the reference value −3σ tilde. A specific method for calculating the standard deviation σ tilde will be described later.

[0045] The generation unit 32 may generate the anomaly determination model based on, for example, the correlation of the sensor data of each of the multiple sensors 12. The generation unit 32, for example, calculates the correlation of the sensor data detected by each of the multiple sensors 12 over a predetermined period. The generation unit 32, for example, calculates the correlation of the sensor data detected by each of the multiple sensors 12 over a predetermined period. The generation unit 32, for example, generates the anomaly determination model including a predicted value of the sensor data of each of the multiple sensors 12 based on the calculated correlation. In other words, the generation unit 32 generates an anomaly determination model that determines that the sensor data is abnormal when the sensor data deviates from the predicted value based on, for example, the correlation of the multiple sensor data. A specific example will be described later.

[0046] 5 is a diagram illustrating an example of a method for calculating a correlation according to an embodiment. As shown in FIG. 5, the correlation information D2 includes items such as a "detection item," a "correlation target," and a "correlation coefficient."

[0047] The "detection item" is information indicating the type of physical quantity detected by the sensor 12. The "correlation target" is information indicating the multiple sensors 12 for which the correlation is calculated. The "correlation coefficient" is information indicating the correlation coefficient between the multiple sensors 12 included in the "correlation target."

[0048] In Fig. 5, each item is conceptually described as "E01", but in reality, specific data is stored for each item. One example shown in Fig. 5 indicates that the correlation coefficient of the detection results of "F01", which is one of the multiple sensors 12 that detected "E01", is "G01".

[0049] For example, based on the "correlation coefficient," the generation unit 32 determines a predicted value for each "inspection item" of the multiple sensors 12. Then, the generation unit 32 generates an anomaly determination model that determines that the sensor data is abnormal when the predicted value is not met.

[0050] The detection unit 34 evaluates the monitoring target. The detection unit 34 detects signs of an abnormality that may occur in the monitoring target. The detection unit 34, for example, determines whether an abnormality may occur in the monitoring target. The detection unit 34 determines whether an abnormality may occur in the monitoring target, for example, based on at least one of the sensor data stored in the sensor data storage unit 22a and the abnormality determination model stored in the model storage unit 22b. The detection unit 34 determines that an abnormality may occur in the monitoring target when it is determined that the sensor data is abnormal, for example, based on at least one of the sensor data acquired in real time by the acquisition unit 30 and the abnormality determination model.

[0051] 6 is a diagram illustrating an example of an evaluation method according to an embodiment. As shown in FIG. 6, evaluation information D3 includes items such as "start time," "end time," "elapsed time," "sensor ID," "inspection item," and "evaluation result."

[0052] "Start time" is information indicating the time when monitoring of the monitored object started. "End time" is information indicating the time when monitoring of the monitored object ended. "Elapsed time" is information indicating the time during which the monitored object was monitored. "Sensor ID" is identification information for identifying the sensor 12. "Inspection item" is information indicating the type of physical quantity detected by the sensor 12. "Evaluation result" is information indicating the evaluation result of the sensor data.

[0053] In Fig. 6, each item is conceptually described as "H01", but in reality, specific data is stored in each item. One example shown in Fig. 5 indicates that the evaluation result of "L01" detected by sensor ID "K01" during the elapsed time "J01" from start time "H01" to end time "J01" is "M01".

[0054] For example, if the sensor data detected by the sensor 12 is not between the upper limit and lower limit indicated by the abnormality determination model, the detection unit 34 determines that the sensor data is abnormal. In this case, "M01" indicates information that the sensor data is abnormal.

[0055] The detection unit 34 may detect signs of an abnormality that may occur in the monitored object, for example, by using a statistical method on the sensor data stored in the sensor data storage unit 22a. Specifically, the detection unit 34 may detect signs of an abnormality that may occur in the monitored object, for example, by using a method such as multivariate analysis on the sensor data stored in the sensor data storage unit 22a. Specifically, the detection unit 34 may detect signs of an abnormality that may occur in the monitored object, for example, by using a method such as principal component analysis, canonical correlation analysis, linear discriminant analysis, multidimensional scaling, cluster analysis, partial least squares, or covariance structure analysis.

[0056] The providing unit 36 ​​provides, for example, various types of information to a terminal device used by the user. The providing unit 36 ​​provides, for example, a plurality of pieces of sensor data acquired from a plurality of sensors 12 by the acquiring unit 30 to a terminal device used by the user. The providing unit 36 ​​provides, for example, an evaluation result by the detecting unit 34 to a terminal device used by the user. The providing unit 36 ​​provides the plurality of pieces of sensor data and the evaluation result to a terminal device used by the user, for example, in accordance with an instruction from the user. That is, the providing unit 36 ​​transmits, via the communication unit 20, the plurality of pieces of sensor data acquired from a plurality of sensors 12 by the acquiring unit 30 and the evaluation result by the detecting unit 34 to the user's terminal device.

[0057] [Detection method] (First method) A first method for detecting a sign of an abnormality that may occur in a monitoring target according to the embodiment will be described with reference to Fig. 7. Fig. 7 is a diagram for explaining the first method for detecting a sign of an abnormality that may occur in a monitoring target according to the embodiment.

[0058] In FIG. 7, the horizontal axis represents time and the vertical axis represents signal level. Line 101 represents the signal level of sensor data detected in real time by sensor 12a, one of the multiple sensors 12. Line 102 represents the signal level of sensor data detected in real time by sensor 12b, another of the multiple sensors 12, which is different from sensor 12a. Sensor 12a and sensor 12b may detect the same physical quantity, or may detect different physical quantities. Line 103 represents the threshold value of the sensor data. The threshold value may be set based on an anomaly determination model generated by generation unit 32, for example, or may be set in advance by the user. The first method is a method of detecting a sign of an anomaly that may occur in the monitored object depending on whether the sensor data exceeds a threshold value.

[0059] 7, the line 102 is below the threshold indicated by the line 103 over the entire section. The line 101 exceeds the threshold indicated by the line 103 at the timing t1. In this case, the detection unit 34 detects a sign of an abnormality that may occur in the monitored object at the timing t1.

[0060] (Second method) A second method for detecting a sign of an abnormality that may occur in a monitoring target according to the embodiment will be described with reference to Fig. 8. Fig. 8 is a diagram for explaining the second method for detecting a sign of an abnormality that may occur in a monitoring target according to the embodiment.

[0061] In Figure 8, the horizontal axis represents time and the vertical axis represents signal level. Line 111 represents sensor data from one of the multiple sensors 12. Line 112 represents the upper limit value of the sensor data. Line 113 represents the upper limit value of the sensor data. The second method is a method of detecting signs of an abnormality that may occur in the monitored object depending on whether the sensor data exceeds the upper limit value or falls below the lower limit value.

[0062] The generation unit 32 generates an abnormality determination model based on, for example, the degree of variation in the sensor data indicated by the line 111. The generation unit 32 calculates the standard deviation σ for each of the multiple sensor data based on, for example, the sensor data during the evaluation period from timing t11 to timing t12, using the following formula (1):

[0063]

number

[0064] In equation (1), n ​​is the number of sampling data, X i is the signal level of the sensor data, and X bar is the average value of the signal levels of the sensor data from timing t11 to timing t12. The generation unit 32 then generates an abnormality determination model in which a signal level of the reference value + 3σ is the upper limit and a signal level of the reference value + 3σ is the lower limit. The reference value is, for example, but is not limited to, the average value of the signal levels of the sensor data from timing t11 to timing t12. The detection unit 34 determines that an abnormality may occur in the monitoring target if, after timing t12, line 111 exceeds line 112 or falls below line 113.

[0065] (Third Method) A third method for detecting a sign of an abnormality that may occur in a monitoring target according to the embodiment will be described with reference to Fig. 9 and Fig. 10. Fig. 9 is a diagram showing signal levels of a plurality of sensors according to the embodiment. Fig. 10 is a diagram for explaining the third method for detecting a sign of an abnormality that may occur in a monitoring target according to the embodiment.

[0066] In FIG. 9, the horizontal axis represents time and the vertical axis represents signal level. Line 121 represents the signal level of sensor data detected in real time by sensor 12a of the multiple sensors 12. Line 122 represents the signal level of sensor data detected in real time by sensor 12b, which is different from sensor 12a, of the multiple sensors 12. Line 123 represents the signal level of sensor data detected in real time by sensor 12c, which is different from sensors 12a and 12b, of the multiple sensors. Sensors 12a to 12c may detect the same physical quantity or different physical quantities. A third method is a method of detecting a sign of a possible abnormality in a monitored object based on the variation among multiple sensor data and depending on whether the standard deviation among the multiple data exceeds an upper limit or falls below a lower limit.

[0067] In Figure 10, the horizontal axis represents time and the vertical axis represents signal level. Line 131 represents the standard deviation between the sensor data from sensors 12a to 12c. Line 132 represents the upper limit of the standard deviation between the multiple sensor data. Line 133 represents the lower limit of the standard deviation between the multiple sensor data. The third method is a method of detecting signs of an abnormality that may occur in the monitored object depending on whether the standard deviation between the sensor data exceeds the upper limit or falls below the lower limit.

[0068] The generation unit 32 generates an abnormality determination model based on, for example, the degree of variation in standard deviation between the plurality of sensor data indicated by line 131. In other words, the generation unit 32 generates an abnormality determination model based on, for example, the degree of disruption of the relationship between the plurality of sensor data. For example, the generation unit 32 calculates the standard deviation σ tilde between the plurality of sensor data at the current point in time (for example, timing t21) using the following equation (2):

[0069]

number

[0070] In equation (2), k is the number of sensors 12, and X j The tilde represents the signal level of the sensor data at the current time, and the X tilde bar represents the average value of the multiple sensor data at the current time. Then, the generation unit 32 generates an abnormality determination model in which the signal level of the reference value + 3σ is the upper limit and the signal level of the reference value + 3σ is the lower limit. In the third method, the reference value is the current standard deviation σ tilde. If, after timing t21, line 131 exceeds line 132 or falls below line 133, the detection unit 34 determines that an abnormality may occur in the monitored object.

[0071] (Fourth Method) A fourth method for detecting a sign of an abnormality that may occur in a monitoring target according to the embodiment will be described with reference to Fig. 11. Fig. 11 is a diagram for explaining the fourth method for detecting a sign of an abnormality that may occur in a monitoring target according to the embodiment.

[0072] In FIG. 11, the horizontal axis represents time and the vertical axis represents signal level. Line 141 represents the signal level of sensor data detected in real time by sensor 12a of the multiple sensors 12. Line 142 represents the signal level of sensor data detected in real time by sensor 12b, which is different from sensor 12a, of the multiple sensors 12. Line 143 represents the signal level of sensor data detected in real time by sensor 12c, which is different from sensors 12a and 12b, of the multiple sensors. Sensors 12a to 12c may each detect the same physical quantity, or may detect different physical quantities. A fourth method is a method of detecting signs of possible abnormalities in a monitored object based on correlations between multiple sensor data.

[0073] The generation unit 32 calculates, for example, a correlation coefficient between the sensor 12a indicated by line 141, the sensor 12b indicated by line 142, and the sensor 12c indicated by line 143. In the example shown in FIG. 11, the sensors 12a, 12b, and 12c are assumed to have a correlation higher than a predetermined value. In this case, the detection unit 34 calculates a predicted value of the sensor data of each of the sensors 12a to 12c using a statistical method such as multivariate analysis based on the sensor data detected by the sensors 12a to 12c, respectively. The line 144 shown in FIG. 11 indicates the predicted value of the sensor data of the sensor 12a predicted by the detection unit 34. As shown in FIG. 11, the line 141 and the line 144 are misaligned. In this case, the detection unit 34 determines that an abnormality may occur in the monitored object.

[0074] (Abnormality determination model generation process) The flow of processing for generating an abnormality determination model according to the embodiment will be described with reference to Fig. 12. Fig. 12 is a flowchart showing the flow of processing for generating an abnormality determination model according to the embodiment.

[0075] The acquiring unit 30 acquires sensor data from each of the plurality of sensors 12 (step S10). The acquiring unit 30 stores the acquired sensor data in the sensor data storage unit 22a (step S12).

[0076] The generation unit 32 generates an abnormality determination model based on the sensor data stored in the sensor data storage unit 22a (step S14). The generation unit 32 stores the generated abnormality determination model in the model storage unit 22b (step S16). Then, the processing of FIG. 12 ends.

[0077] (Data provision processing) The flow of processing for providing data to a user according to the embodiment will be described with reference to Fig. 13. Fig. 13 is a flowchart showing the flow of processing for providing data to a user according to the embodiment.

[0078] The providing unit 36 ​​determines whether or not to provide the data to the user (step S20). Specifically, the providing unit 36 ​​determines whether or not to provide the sensor data acquired by the acquiring unit 30 and the detection result detected by the detecting unit 34 to the user. For example, when the providing unit 36 ​​receives a data request from the user's terminal device, it determines to provide the data to the user. If it is determined to provide the data (step S20; Yes), the process proceeds to step S22. If it is not determined to provide the data (step S20; No), the process of FIG. 13 ends.

[0079] If the determination in step S20 is Yes, the providing unit 36 ​​provides the data to the user (step S22). Specifically, the providing unit 36 ​​transmits the data to the user's terminal device via the communication unit 20. In this case, the providing unit 36 ​​may, for example, statistically analyze the trends in the sensor data for each of the multiple sensors 12 and provide the results. The providing unit 36 ​​may, for example, provide the sensor data to the user in the form of a histogram, a scatter diagram, or the like. By being provided with the sensor data in various forms, the user can appropriately grasp the state and trends of the monitored object.

[0080] (Abnormality detection process) The flow of processing for detecting a sign of an abnormality that may occur in a monitoring target according to the embodiment will be described with reference to Fig. 14. Fig. 14 is a flowchart showing the flow of processing for detecting a sign of an abnormality that may occur in a monitoring target according to the embodiment.

[0081] The acquiring unit 30 acquires real-time sensor data from each of the multiple sensors 12 (step S30). The detecting unit 34 evaluates the sensor data acquired by the acquiring unit 30 based on the abnormality determination model stored in the model storage unit 22b (step S32).

[0082] The detection unit 34 determines whether an abnormality has occurred in the sensor data (step S34). If it is determined that an abnormality has occurred in the sensor data (step S34; Yes), the process proceeds to step S36. If it is not determined that an abnormality has occurred in the sensor data (step S34; No), the process proceeds to step S38.

[0083] If the determination in step S34 is Yes, the detection unit 34 detects a sign of an abnormality that may occur in the monitored object (step S36). In this case, the detection unit 34 may notify the user via the communication unit 20 that an abnormality may occur in the monitored object.

[0084] If the determination in step S34 is No, the detection unit 34 determines that the monitoring target is stable (step S38).

[0085] The control unit 24 determines whether or not to end the process (step S40). The control unit 24 determines to end the process when, for example, an instruction to end the process is received. If it is determined to end the process (step S40; Yes), the process of Fig. 14 ends. If it is not determined to end the process (step S40; No), the process returns to step S30, and the above-described process is repeated.

[0086] As described above, in this embodiment, signs of an abnormality that may occur in an observation target are detected based on the detection results of a plurality of sensors. As a result, this embodiment can detect signs of an earthquake early, for example, before a disaster such as an earthquake occurs, and can prevent damage from spreading by evacuating users, for example.

[0087] The components of each device shown in the figure are conceptual functional components and do not necessarily have to be physically configured as shown. In other words, the specific form of distribution and integration of each device is not limited to that shown in the figure, and all or part of them can be functionally or physically distributed and integrated in any unit depending on various loads and usage conditions. This distribution and integration configuration may also be performed dynamically.

[0088] Although the embodiments of the present disclosure have been described above, the present disclosure is not limited to the contents of these embodiments. Furthermore, the above-described components include those that can be easily imagined by a person skilled in the art, those that are substantially the same, and those that are within the so-called equivalent range. Furthermore, the above-described components can be combined as appropriate. Furthermore, various omissions, substitutions, or modifications of the components can be made without departing from the spirit of the above-described embodiments.

[0089] The present disclosure can also be configured as follows. (1) an acquisition unit that acquires sensor data from a sensor that monitors the monitoring target; a generation unit that generates an abnormality determination model for determining in advance an abnormality that will occur in the monitoring target based on the plurality of pieces of sensor data acquired by the acquisition unit; a detection unit that detects a sign of an abnormality that may occur in the monitoring target when the value of the sensor data indicates an abnormal value, based on the abnormality determination model generated by the generation unit; An information processing device comprising: (2) the acquisition unit acquires sensor data from each of a plurality of sensors that monitor the monitoring target; the generation unit calculates a standard deviation of the sensor data acquired during a predetermined period, and generates the anomaly determination model including an upper limit value and a lower limit value of the sensor data based on the calculated standard deviation; the detection unit determines that an abnormality may occur in the monitoring target when the sensor data acquired by the acquisition unit exceeds the upper limit value or falls below the lower limit value; The information processing device according to (1) above. (3) the acquisition unit acquires sensor data from each of a plurality of sensors that monitor the monitoring target; the generation unit generates the anomaly determination model including an upper limit value and a lower limit value of a standard deviation between the plurality of sensor data; the detection unit calculates a standard deviation between the plurality of pieces of sensor data, and determines that an abnormality may occur in the monitoring target when the calculated standard deviation exceeds the upper limit value or falls below the lower limit value; The information processing device according to (1) above. (4) the acquisition unit acquires sensor data from each of a plurality of sensors that monitor the monitoring target; the generation unit generates the anomaly determination model including predicted values ​​of the plurality of sensor data based on a correlation between the plurality of sensor data; The detection unit determines that an abnormality may occur in the monitoring target when the sensor data acquired by the acquisition unit includes sensor data that deviates from the predicted value. The information processing device according to (1) above. (5) a providing unit that provides the sensor data acquired by the acquiring unit and the detection result detected by the detecting unit, The information processing device according to any one of (1) to (4). (6) the monitoring target is a slope of a mountain or a cliff, The sensor includes at least an inclination sensor that detects an inclination angle of the slope. The information processing device according to any one of (1) to (5). (7) The monitoring target is a water-sealed underground rock tank, The sensor includes at least a water pressure sensor that detects the water pressure around the area where the water-sealed underground rock tank is installed. The information processing device according to any one of (1) to (5). (8) acquiring sensor data from a sensor monitoring the monitoring target; generating an anomaly determination model for determining in advance an anomaly occurring in the monitoring target based on the acquired plurality of sensor data; detecting a sign of an abnormality that may occur in the monitored object when the value of the sensor data indicates an abnormal value based on the generated abnormality determination model; An information processing method, including: (9) acquiring sensor data from a sensor monitoring the monitoring target; generating an anomaly determination model for determining in advance an anomaly occurring in the monitoring target based on the acquired plurality of sensor data; detecting a sign of an abnormality that may occur in the monitored object when the value of the sensor data indicates an abnormal value based on the generated abnormality determination model; A program that causes a computer to execute the following. [Explanation of symbols]

[0090] 1. Information Processing Systems 10. Information processing equipment 12 sensors 20 Communications Department 22 Memory section 22a Sensor data storage unit 22b Model memory section 24 Control Unit 30 Acquisition Department 32 Generation part 34 Detection unit 36 Providing Department

Claims

1. an acquisition unit that acquires sensor data from a sensor that monitors the monitoring target; a generation unit that generates an abnormality determination model for determining in advance an abnormality that will occur in the monitoring target based on the plurality of pieces of sensor data acquired by the acquisition unit; a detection unit that detects a sign of an abnormality that may occur in the monitoring target when the value of the sensor data indicates an abnormal value, based on the abnormality determination model generated by the generation unit; An information processing device comprising:

2. the acquisition unit acquires sensor data from each of a plurality of sensors that monitor the monitoring target; the generation unit calculates a standard deviation of the sensor data acquired during a predetermined period, and generates the anomaly determination model including an upper limit value and a lower limit value of the sensor data based on the calculated standard deviation; the detection unit determines that an abnormality may occur in the monitoring target when the sensor data acquired by the acquisition unit exceeds the upper limit value or falls below the lower limit value; The information processing device according to claim 1 .

3. the acquisition unit acquires sensor data from each of a plurality of sensors that monitor the monitoring target; the generation unit generates the anomaly determination model including an upper limit value and a lower limit value of a standard deviation between the plurality of sensor data; the detection unit calculates a standard deviation between the plurality of pieces of sensor data, and determines that an abnormality may occur in the monitoring target when the calculated standard deviation exceeds the upper limit value or falls below the lower limit value; The information processing device according to claim 1 .

4. the acquisition unit acquires sensor data from each of a plurality of sensors that monitor the monitoring target; the generation unit generates the anomaly determination model including predicted values ​​of the plurality of sensor data based on a correlation between the plurality of sensor data; The detection unit determines that an abnormality may occur in the monitoring target when the sensor data acquired by the acquisition unit includes sensor data that deviates from the predicted value. The information processing device according to claim 1 .

5. a providing unit that provides the sensor data acquired by the acquiring unit and the detection result detected by the detecting unit, The information processing device according to claim 1 .

6. the monitoring target is a slope of a mountain or a cliff, The sensor includes at least an inclination sensor that detects an inclination angle of the slope. The information processing device according to claim 1 .

7. The monitoring target is a water-sealed underground rock tank, The sensor includes at least a water pressure sensor that detects the water pressure around the area where the water-sealed underground rock tank is installed. The information processing device according to claim 1 .

8. acquiring sensor data from a sensor monitoring the monitoring target; generating an anomaly determination model for determining in advance an anomaly occurring in the monitoring target based on the acquired plurality of sensor data; detecting a sign of an abnormality that may occur in the monitored object when the value of the sensor data indicates an abnormal value based on the generated abnormality determination model; An information processing method, including:

9. acquiring sensor data from a sensor monitoring the monitoring target; generating an anomaly determination model for determining in advance an anomaly occurring in the monitoring target based on the acquired plurality of sensor data; detecting a sign of an abnormality that may occur in the monitored object when the value of the sensor data indicates an abnormal value based on the generated abnormality determination model; A program that causes a computer to execute the following.

Images