Detection device

Abstract

A detection device (30) comprises a control unit (36) that divides, into a plurality of zones, an area in which is present an underground pipeline for which a prediction is to be made, detects the structural feature of the underground pipeline in each of the plurality of zones, transmits, to a prediction device (20), the detected structural feature of the underground pipeline in each of the zones and facility data, ground data, and disaster strike data acquired for each of the zones from a facility database (43A), causes the prediction device (20) to predict the disaster strike probability for the underground pipeline in each of the zones at the time of an earthquake, and detects, from among the plurality of zones, the zone for which the prediction device (20) has predicted a disaster strike probability exceeding a threshold for the underground pipeline.

Description

Detection Device

[0001] The present disclosure relates to a detection device.

[0002] Conventionally, techniques for predicting earthquake damage to underground pipelines have been known. Several earthquake damage prediction formulas have been developed for water supply pipes. For example, Non-Patent Document 1 discloses a damage prediction formula for earthquake motion intensity that uses pipe type, pipe diameter, topography, and liquefaction as correction factors. Non-Patent Document 2 discloses a technique for improving the accuracy of damage prediction by adding a correction factor for liquefaction of residential land development ground to the damage prediction formula disclosed in Non-Patent Document 1. Non-Patent Document 3 also discloses a technique for improving the accuracy of damage prediction by proposing a vulnerability index calculated from correction factors for pipe type and diameter classification and extension distance. Furthermore, Non-Patent Document 4 discloses a case study evaluating the correlation with damage to wooden buildings above ground.

[0003] Ryuji Isoyama and three others, "Study on Earthquake Damage Prediction of Water Supply Pipes," Journal of Japan Water Works Association 67(2) (Issue 761) (1998-02); Shigeru Nagata and two others, "Study on the Prediction Accuracy of Existing Earthquake Damage Prediction Formulas for Water Supply Distribution Pipes," Proceedings of the 34th JSCE Earthquake Engineering Research Conference, Vol. 71, No. 4, pp. I_50-I_61 (2014-10); Noburo Nojima, "Earthquake Vulnerability Assessment of Lifeline Networks Using Fragility Indices," Proceedings of the Japan Society for Regional Safety Studies, Vol. 10, pp. 137-146 (2008-11); Yoshihisa Maruyama and two others, "Evaluation of the Correlation between Water Supply Pipe Damage and Wooden Building Damage in the Northern Tokyo Bay Earthquake," Proceedings of the Japan Society of Civil Engineers A1 (Structural and Earthquake Engineering), Vol. 68, No. 4, 2012, p. I_950-I_958 (2012)

[0004] It is assumed that the force of response to an earthquake varies depending on the shape of an underground pipeline. However, there is currently no method to determine the probability of damage or vulnerability to damage based on detailed shape. Therefore, there is room for improvement in technology for predicting earthquake damage to underground pipelines.

[0005] In view of the above circumstances, an object of the present disclosure is to improve the technology relating to prediction of earthquake damage to underground pipelines.

[0006] A detection device according to one embodiment of the present disclosure divides an area in which an underground pipeline to be predicted is located into a plurality of zones, detects structural features of the underground pipeline in each of the plurality of zones, transmits the detected structural features of the underground pipeline in each of the zones, along with equipment data, ground data, and disaster data for each zone obtained from an equipment database, to a prediction device, causing the prediction device to predict the probability of damage to the underground pipeline in each zone during an earthquake, and includes a control unit that detects from among the plurality of zones an area in which the probability of damage to the underground pipeline predicted by the prediction device exceeds a threshold.

[0007] According to an embodiment of the present disclosure, it is possible to accurately predict the probability of damage to underground pipelines during an earthquake based on the structural characteristics of the underground pipelines in each area where the underground pipelines are located. Therefore, the technology for predicting earthquake damage to underground pipelines is improved.

[0008] FIG. 1 is a block diagram showing a schematic configuration example of a system according to an embodiment of the present disclosure. FIG. 2 is a sequence diagram showing an operation example of a system according to an embodiment of the present disclosure. FIG. 3 is a sequence diagram showing an operation example of a system according to an embodiment of the present disclosure. FIG. 4 is a schematic diagram showing a configuration example of a damage prediction system. FIG. 5 is an example of a summary table of structural features related to the construction of learning data. FIG. 6 is a schematic diagram showing an example of construction of learning data. FIG. 7 is a table showing an example of past seismic motion data in a prediction target area. FIG. 8 is a table showing an example of past seismic motion data in a prediction target area. FIG. 9 is a table showing an example of combining structural feature data with seismic motion data. FIG. 10 is a schematic diagram showing an example of combining structural feature data with seismic motion data. FIG. 11 is a schematic diagram showing an example of dividing learning data. FIG. 12 is a schematic diagram explaining an operation example of a detection device. FIG. 13 is a block diagram showing a schematic configuration example of a computer functioning as a detection device.

[0009] (Overview of the embodiment) An overview of a system 1 according to an embodiment of the present disclosure will be described with reference to Fig. 1. The system 1 includes a disaster prediction system 3, a detection device 30, and a server 40.

[0010] The damage prediction system 3 is a system for predicting the probability of damage to underground pipelines during an earthquake. The damage prediction system 3 includes a learning device 10 and a prediction device 20.

[0011] The learning device 10 is a computer that constructs (generates) a disaster prediction model M.

[0012] The prediction device 20 is a computer that inputs predetermined information into the damage prediction model M generated by the learning device 10 and predicts the probability of damage to underground pipelines in the event of an earthquake.

[0013] As shown in Fig. 7, the detection device 30 is a computer mounted on the vehicle 5. The vehicle 5 is a light vehicle that does not use a prime mover such as an engine as a power source. An example of a light vehicle is a pushcart. However, the vehicle 5 is not limited to a pushcart or other light vehicles, and may be an automobile driven by an engine and / or an electric motor.

[0014] The server 40 is a computer owned by a company that manages underground pipelines, etc. The server 40 includes an equipment database 43A that registers equipment data, topographical data, disaster data, etc. for each area in the area where earthquake damage to underground pipelines is predicted.

[0015] Next, each component of the system 1 will be described in detail.

[0016] (Configuration of Learning Device) As shown in FIG. 1 , the learning device 10 includes a communication unit 11 , a storage unit 12 , and a control unit 13 .

[0017] The communication unit 11 includes one or more communication interfaces connected to the network 2. The communication interfaces correspond to, for example, a mobile communication standard, a wired local area network (LAN) standard, or a wireless LAN standard. However, the communication interfaces are not limited to these and may correspond to any communication standard. In this embodiment, the learning device 10 communicates with the prediction device 20, the detection device 30, and the server 40 via the communication unit 11 and the network 2.

[0018] The storage unit 12 includes one or more memories. Each memory included in the storage unit 12 may function as, for example, a main storage device, an auxiliary storage device, or a cache memory. The storage unit 12 stores any information used in the operation of the learning device 10. For example, the storage unit 12 may store a system program, a database, an application program, and the generated disaster prediction model M. The information stored in the storage unit 12 may be updatable with information obtained from the network 2 via the communication unit 11, for example.

[0019] The control unit 13 includes one or more processors, one or more programmable circuits, one or more dedicated circuits, or a combination thereof. The processor is, for example, a general-purpose processor such as a CPU (Central Processing Unit) or a GPU (Graphics Processing Unit), or a dedicated processor specialized for specific processing. However, the processor is not limited to these. The programmable circuit is, for example, an FPGA (Field-Programmable Gate Array). However, the programmable circuit is not limited to an FPGA. The dedicated circuit is, for example, an ASIC (Application Specific Integrated Circuit). However, the dedicated circuit is not limited to an ASIC. The control unit 13 executes information processing related to the operation of the learning device 10.

[0020] (Configuration of Prediction Device) As shown in FIG. 1 , the prediction device 20 includes a communication unit 21 , an input unit 22 , a storage unit 23 , and a control unit 24 .

[0021] The communication unit 21 includes one or more communication interfaces connected to the network 2. The communication interfaces correspond to, for example, a mobile communication standard, a wired LAN standard, or a wireless LAN standard. However, the communication interfaces are not limited to these and may correspond to any communication standard. In this embodiment, the prediction device 20 communicates with the learning device 10, the detection device 30, and the server 40 via the communication unit 21 and the network 2.

[0022] The input unit 22 acquires structural characteristic data, equipment data, seismic motion data, and ground data (liquefaction data, etc.) required to predict the probability of damage from the detection device 30 via the communication unit 21 and the network 2.

[0023] The storage unit 23 includes one or more memories. Each memory included in the storage unit 23 may function as, for example, a main storage device, an auxiliary storage device, or a cache memory. The storage unit 23 stores any information used in the operation of the prediction device 20. For example, the storage unit 23 may store a system program, a database, an application program, and a damage prediction result (damage probability). The information stored in the storage unit 23 may be updatable with information obtained from the network 2 via the communication unit 21, for example.

[0024] The control unit 24 includes one or more processors, one or more programmable circuits, one or more dedicated circuits, or a combination thereof. The control unit 24 executes information processing related to the operation of the prediction device 20.

[0025] (Configuration of the Detection Device) As shown in FIG. 1 , the detection device 30 includes a communication unit 31 , a first detection unit 32 , a second detection unit 33 , a display unit 34 , a storage unit 35 , and a control unit 36 ​​.

[0026] The communication unit 31 includes at least one communication module connectable to the network 2. The communication module is a communication module compatible with mobile communication standards such as LTE (Long Term Evolution), 4G (4th Generation), or 5G (5th Generation). However, the communication module is not limited to these. The communication module may be compatible with any communication standard. In this embodiment, the detection device 30 communicates with the learning device 10, the prediction device 20, and the server 40 via the communication unit 31 and the network 2.

[0027] The first detection unit 32 includes one or more sensors. The one or more sensors include, for example, a camera 32A and a gyro sensor 32B. However, the one or more sensors are not limited to these.

[0028] The second detection unit 33 includes one or more sensors. The one or more sensors include, for example, a GPS (Global Positioning System) sensor 33A and a distance sensor 33B. However, the one or more sensors are not limited to these. The one or more sensors may include a sensor that measures position information, such as a three-axis acceleration sensor or a geomagnetic sensor.

[0029] The display unit 34 includes at least one display interface capable of displaying data. The display interface is, for example, a display. The display is, for example, an LCD (Liquid Crystal Display) or an organic EL (Electro Luminescent) display. However, the output interface is not limited to these.

[0030] The storage unit 35 includes one or more memories. Each memory included in the storage unit 35 may function as, for example, a main storage device, an auxiliary storage device, or a cache memory. The storage unit 35 stores any information used in the operation of the detection device 30. For example, the storage unit 35 may store a system program, a database, an application program, information on the structural characteristics of each of multiple zones into which an area where an underground pipeline to be predicted exists is divided, detected by the first detection unit 32, and location information of each zone detected by the second detection unit 33. The information stored in the storage unit 35 may be updatable with information obtained from the network 2 via, for example, the communication unit 31.

[0031] The control unit 36 ​​includes one or more processors, one or more programmable circuits, one or more dedicated circuits, or a combination thereof. The control unit 36 ​​executes information processing related to the operation of the detection device 30.

[0032] (Configuration of Server) As shown in FIG. 1 , the server 40 includes a communication unit 41 , a display unit 42 , a storage unit 43 , and a control unit 44 .

[0033] The communication unit 41 includes one or more communication interfaces connected to the network 2. The communication interfaces correspond to, for example, a mobile communication standard, a wired LAN standard, or a wireless LAN standard. However, the communication interfaces are not limited to these and may correspond to any communication standard. In this embodiment, the server 40 communicates with the learning device 10, the prediction device 20, and the detection device 30 via the communication unit 41 and the network 2.

[0034] The display unit 42 includes at least one display interface capable of displaying data. The display interface is, for example, a display. The display is, for example, an LCD or an organic EL display. However, the display interface is not limited to these.

[0035] The storage unit 43 includes one or more memories. Examples of the memories include, but are not limited to, semiconductor memory, magnetic memory, or optical memory. Each memory included in the storage unit 43 may function as, for example, a main storage device, an auxiliary storage device, or a cache memory. The storage unit 43 stores any information used in the operation of the server 40. For example, the storage unit 43 may store system programs, application programs, embedded software, and the like. Furthermore, the storage unit 43 includes an equipment database 43A that stores equipment data, topographical data, disaster data, and images of plan views and cross sections of the area where the underground pipeline is located, for each of multiple zones into which the earthquake damage prediction target area for underground pipelines is divided. The information stored in the equipment database 43A is provided in response to a request from the detection device 30 or the like when predicting the probability of damage in the prediction target area. The information stored in the storage unit 43 may be updated, for example, with information acquired from the network 2 via the communication unit 41.

[0036] The control unit 44 includes one or more processors, one or more programmable circuits, one or more dedicated circuits, or a combination thereof. The control unit 44 executes information processing related to the operation of the server 40.

[0037] 2A and 2B are sequence diagrams illustrating an example of operation of the system 1 according to an embodiment of the present disclosure. The operation of the detection device 30 will be described in relation to the operation of the disaster prediction system 3 and the server 40 with reference to FIGS.

[0038] S101: The control unit 13 of the learning device 10 generates a disaster prediction model M.

[0039] Fig. 3 is a schematic diagram showing an example configuration of the damage prediction system 3. As shown in Fig. 3, the prediction device 20 associated with the damage prediction system 3 inputs (i) structural characteristics of the underground pipeline, (ii) equipment data such as the pipe type and age, (iii) disaster data such as earthquake motion intensity, and (iv) ground data such as susceptibility to liquefaction into a damage prediction model M generated by the learning device 10, and predicts the probability of damage to the underground pipeline during an earthquake.

[0040] The structural features include the magnitude of bends in underground pipelines, connections between different types of pipelines, the presence or absence of branches, etc. Other structural features, such as the presence or absence of accessories such as valves, may also be used, or some of these may be used.

[0041] As shown in FIG. 3 , the damage prediction system 3 includes a learning device 10 and a prediction device 20. First, the learning device 10 will be described. The learning device 10 constructs learning data and generates a damage prediction model M by machine learning the learning data. When constructing the learning data, learning data that can learn damage patterns by machine learning is constructed from data obtained from past disaster cases (past disaster data). For this purpose, the learning data is shaped.

[0042] FIG. 4A shows an example of a summary table of structural features used in building training data as structural feature data D1. FIG. 4B is a schematic diagram illustrating an example of building training data. The training data targets data from an area where underground pipelines that have been damaged by earthquakes in the past are located. The area where the underground pipelines to be trained are located is divided into N×N meter zones, where N is a real number. The number of structural features and damaged locations in each zone is then tallied. As shown on the vertical axis of FIG. 4A , the number of damaged locations and the number of structural features are tallied for each pipe type and diameter. The number of bends is tallied for locations greater than a predetermined angle. In this case, the number of bends may be tallied for multiple angles. For example, when x and y are integers, the number of bends is tallied as x for bends between 10 and 20 degrees, y for bends greater than 20 degrees, etc. Alternatively, numerical information such as the maximum, average, or minimum bend angles within each zone may be used. The number of branching points may be counted for each branching shape, such as a T-shape or a cross shape.

[0043] Furthermore, the number of dissimilar pipe connections is tallied for each combination of pipe types. Examples of such combinations include a combination of threaded steel pipes and threaded cast iron pipes, or a combination of plug-in PVC pipes and threaded cast iron pipes. The combinations of pipe types are not limited to these, and may also be tallied for each combination of materials, such as a combination of resin and metal. The pipe types are determined using a geographic information system (GIS) based on facility drawings (plan and profile).

[0044] Figures 5A and 5B show examples of past seismic motion data for the prediction target area as seismic motion data D2 and D3, respectively. Figure 5C is a table showing an example of combining structural feature data D1 with seismic motion data D2 and D3. Figure 5D is a schematic diagram showing an example of constructing learning data D4 by combining structural feature data D1 with seismic motion data D2 and D3. As shown in Figure 5D, the structural feature data D1 shown in Figure 4A is combined with past seismic motion data D2 and D3 for the area shown in Figures 5A and 5B. For the seismic motion data D2 and D3, a seismic motion map provided by Quiet+ or the like may be used. If the intensity of the seismic motion is provided in a mesh, a mesh different from the N x N meter mesh used to construct the structural feature data D1 may be used. As shown in Figures 5A and 5B, the seismic motion data D2 and D3 include measured seismic intensity, maximum ground surface velocity, and maximum ground surface acceleration. As shown in Fig. 5C, the structural characteristic data D1 is combined with the seismic motion data D2 and D3. Learning data D4 is constructed by adding ground data, such as the microtopography classification published by the National Research Institute for Earth Science and Disaster Prevention, the average S-wave velocity, or the PL value indicating the susceptibility to liquefaction.

[0045] FIG. 6 is a schematic diagram showing an example of division of training data. As shown in FIG. 6, the training data is classified by pipe type and diameter. Machine learning is performed as classification to estimate the presence or absence of damage using the intensity of earthquake motion, ground data, structural characteristics, and the number of years elapsed for each N x N meter zone, or as regression to estimate the number of damaged areas. Algorithms used for machine learning include logistic regression, random forest, gradient boosting decision tree, and deep neural network. In this way, a damage prediction model M is constructed for each pipe type and diameter.

[0046] S102: The control unit 36 ​​of the detection device 30 divides the area where the underground pipeline to be predicted exists into a plurality of zones.

[0047] S103: The control unit 36 ​​detects the structural characteristics of each of the multiple zones.

[0048] FIG. 7 is a schematic diagram illustrating an example of the operation of the detection device 30. The detection device 30 is mounted on a vehicle 5 and travels through each of the multiple zones into which an earthquake damage prediction target area is divided. The detection device 30 detects structural features of the underground pipeline 4 in each zone, such as the magnitude of the pipeline bends, the connection of different types of pipelines, and the presence or absence of branches. The detection device 30 then counts these features, as shown in FIG. 4A . The magnitude of the pipeline bends is detected by a gyro sensor 32B provided in the first detection unit 32. The connection of different types of pipelines and the presence or absence of branches are detected by image analysis or deep learning of images captured by a camera 32A provided in the first detection unit 32. The detected structural features are stored in the memory unit 35. The control unit 36 ​​reads the detected structural features from the memory unit 35 and transmits them to the prediction device 20. The transmitted structural features are provided to the input unit 22 of the prediction device 20.

[0049] S104: The control unit 36 ​​detects the position information of each area where the structural characteristics of the underground pipeline 4 are detected.

[0050] The position information is measured using a GPS-compatible receiver (GPS sensor 33A) provided in the second detection unit 33. The control unit 36 ​​may detect the position information using a distance sensor 33B provided in the second detection unit 33 that measures distance from the distance traveled by the wheels 6 of the vehicle 5.

[0051] S105: The control unit 36 ​​transmits to the prediction device 20 the structural characteristics of the underground pipeline 4 in each area where the position information has been detected.

[0052] The control unit 36 ​​may transmit to the prediction device 20 the structural features of the underground pipeline 4 as well as the location information of each area in which the structural features of the underground pipeline 4 are detected.

[0053] S106: The control unit 36 ​​requests the server 40 to transmit the facility data, ground data, and disaster data registered in the facility database 43A.

[0054] S107: In response to the request from the control unit 36, the control unit 44 of the server 40 transmits the facility data, ground data, and disaster data registered in the facility database 43A to the detection device 30.

[0055] S108: The control unit 36 ​​transmits the facility data, ground data, and disaster data received from the server 40 to the prediction device 20.

[0056] S109: The control unit 24 of the prediction device 20 predicts the probability of disaster for each area.

[0057] The prediction device 20 predicts the probability of damage to each area by inputting the structural characteristics, equipment data, ground data, and disaster data of each area transmitted from the control unit 36 ​​into a damage prediction model M generated by machine learning.

[0058] As described above, the learning device 10 generates the damage prediction model M by machine learning the learning data. The prediction device 20 inputs the data input to the input unit 22 into the damage prediction model M generated by machine learning and predicts the probability of damage in each area. At this time, the input unit 22 must input all of the same data items as those used when generating the damage prediction model M: structural features, facility information, ground information, and disaster information. For example, if the damage prediction model M was generated using maximum speed and maximum acceleration in the disaster information (earthquake information), the same maximum speed and maximum acceleration must also be used in the input information. Furthermore, the prediction unit for information input must be the N x N meter area unit used when constructing the structural feature data. The earthquake motion used for prediction may be either expected earthquake motion information or earthquake motion information after an earthquake has occurred. When this data is input into the damage prediction model M, the probability of damage to the underground pipeline 4 in each area is predicted. Furthermore, when the location information of the point where the structural characteristics of the underground pipeline 4 are detected is transmitted to the prediction device 20, the prediction device 20 becomes able to predict which locations of the underground pipeline 4 in each area will be more likely to be affected by the earthquake.

[0059] S110: The control unit 36 ​​acquires the probability of damage to the underground pipes 4 in each area predicted by the prediction device 20.

[0060] S111: The control unit 36 ​​detects areas where the probability of damage to the underground conduits 4 in each area during an earthquake exceeds a threshold value α.

[0061] When the control unit 36 ​​transmits the location information of the point where the structural characteristics of the underground pipeline 4 are detected to the prediction device 20, it detects areas where the probability of damage to the underground pipeline 4 in each area during an earthquake exceeds the threshold value α based on the prediction results by the prediction device 20, and displays the location of the detected area on a map displayed on the display provided in the display unit 34.

[0062] S112 : The control unit 36 ​​transmits the detection result to the server 40 .

[0063] S113: The control unit 44 of the server 40 registers the detection result sent from the control unit 36 ​​in the equipment database 43A.

[0064] S114: The control unit 36 ​​displays on the map the locations of areas where the probability of damage to the underground conduits 4 in each area during an earthquake exceeds the threshold value α.

[0065] The control unit 36 ​​may display the location of the area where the damage probability exceeds the threshold value α on a map displayed on a display provided in the display unit 34. Note that the display device that displays the location of the area on the map is not limited to the display unit 34 of the detection device 30. The location information of the area may be configured to be displayed on any display device that can communicate with the server 40 via the network 2 (for example, the display unit 42 of the server 40).

[0066] As described above, the detection device 30 of this embodiment divides the area where the underground pipeline to be predicted is located into multiple zones, detects the structural characteristics of the underground pipeline in each of the multiple zones, transmits the detected structural characteristics of the underground pipeline in each zone, as well as the equipment data, ground data, and disaster data for each zone obtained from the equipment database 43A, to the prediction device 20, causes the prediction device 20 to predict the probability of damage to the underground pipeline in each zone during an earthquake, and detects from among the multiple zones the zone where the probability of damage to the underground pipeline predicted by the prediction device 20 exceeds a threshold.

[0067] This configuration makes it possible to predict with high accuracy the probability of damage to underground pipelines during an earthquake based on the detailed structural characteristics of the underground pipelines in each area where they exist, thereby improving the technology for predicting earthquake damage to underground pipelines.

[0068] Although the present disclosure has been described based on the drawings and examples, it should be noted that those skilled in the art may make various modifications and alterations based on the present disclosure. Therefore, it should be noted that these modifications and alterations are included in the scope of the present disclosure. For example, the functions included in each component or step can be rearranged so as not to be logically inconsistent, and multiple components or steps can be combined or divided into one.

[0069] For example, in the above-described embodiment, the configuration and operation of the detection device 30 may be distributed among multiple computers that can communicate with each other. Also, for example, an embodiment in which some of the components of the detection device 30 are provided in the prediction device 20 may be possible.

[0070] In the above-described embodiment, the detection device 30 detects the structural features of the underground pipeline 4 using the camera 32A and the gyro sensor 32B provided in the first detection unit 32. However, the method of detecting the structural features by the detection device 30 is not limited to this. The control unit 36 ​​of the detection device 30 may acquire, from the server 40, images of plan views and / or cross-sectional views of the area where the underground pipeline 4 is located, which are stored in the equipment database 43A, and detect the structural features of the underground pipeline 4 included in the acquired images by image recognition.

[0071] Also, an embodiment is possible in which, for example, a general-purpose computer functions as the detection device 30 according to the above-described embodiment. FIG. 8 is a block diagram showing a schematic configuration example of a computer functioning as the detection device 30. Specifically, a program describing the processing content for realizing each function of the detection device 30 according to the above-described embodiment is stored in the memory of the general-purpose computer, and the program is read and executed by a processor. Therefore, the present disclosure can also be realized as a program executable by a processor, or a non-transitory computer-readable medium storing the program.

[0072] The following additional notes are provided regarding the above-described embodiments.

[0073] (Supplementary Item 1) A detection device comprising a control unit that divides an area where an underground pipeline to be predicted is located into a plurality of zones, detects structural features of the underground pipeline in each of the plurality of zones, transmits the detected structural features of the underground pipeline in each of the zones, as well as facility data, ground data, and disaster data for each zone obtained from an facility database to a prediction device, causing the prediction device to predict a damage probability of the underground pipeline in each of the zones in the event of an earthquake, and detects from among the plurality of zones an area where the damage probability of the underground pipeline predicted by the prediction device exceeds a threshold. (Supplementary Item 2) The detection device according to Supplementary Item 1, wherein the structural features include a magnitude of a bend in the underground pipeline, a connection of different types of pipelines, and the presence or absence of a branch, and the magnitude of the bend in the pipeline is detected by a gyro sensor, and the connection of different types of pipelines and the presence or absence of the branch are detected by image analysis of an image captured by a camera. (Supplementary Item 3) The detection device according to Supplementary Item 1 or 2, wherein the control unit further displays on a map the positions of areas where the damage probability of the detected underground pipeline exceeds the threshold. (Supplementary Item 4) The detection device according to any one of Supplementary Items 1 to 3, wherein the prediction device predicts the damage probability of the underground pipeline in each of the areas during an earthquake by inputting the structural features of the underground pipeline, the facility data, the ground data, and the disaster data transmitted from the control unit into a damage prediction model generated by machine learning.

[0074] REFERENCE SIGNS LIST 1 System 2 Network 3 Disaster prediction system 4 Underground pipeline 5 Vehicle (light vehicle) 6 Wheel 10 Learning device 11 Communication unit 12 Memory unit 13 Control unit 20 Prediction device 21 Communication unit 22 Input unit 23 Memory unit 24 Control unit 30 Detection device 31 Communication unit 32 First detection unit 32A Camera 32B Gyro sensor 33 Second detection unit 33A GPS sensor 33B Distance sensor 34 Display unit 35 Memory unit 36 ​​Control unit 40 Server 41 Communication unit 42 Display unit 43 Memory unit 43A Equipment database (DB) 44 Control unit

Claims

1. A detection device comprising a control unit that divides an area where an underground pipeline to be predicted is located into a plurality of zones, detects structural characteristics of the underground pipeline in each of the plurality of zones, transmits the detected structural characteristics of the underground pipeline in each of the zones, along with equipment data, ground data, and disaster data for each zone obtained from an equipment database, to a prediction device, causing the prediction device to predict the probability of damage to the underground pipeline in each of the zones during an earthquake, and detects from among the plurality of zones an area where the probability of damage to the underground pipeline predicted by the prediction device exceeds a threshold.

2. A detection device according to claim 1, wherein the structural features include the magnitude of the bends in the underground pipeline, the connections of different types of pipelines, and the presence or absence of branches, the magnitude of the bends in the pipelines being detected by a gyro sensor, and the connections of different types of pipelines and the presence or absence of branches being detected by image analysis of images captured by a camera.

3. A detection device according to claim 1, wherein the control unit further displays on a map the location of an area where the probability of damage to the detected underground pipeline exceeds the threshold.

4. A detection device as described in claim 1, wherein the prediction device predicts the probability of damage to the underground pipeline in each area during an earthquake by inputting the structural characteristics of the underground pipeline, the equipment data, the ground data, and the disaster data transmitted from the control unit into a damage prediction model generated by machine learning.

Images