Information processing system, information processing method, and program

Abstract

To provide an information processing system or the like capable of detecting a breakage position even without an expert.SOLUTION: According to an aspect of the present invention, there is provided an information processing system including one or more processors. In this information processing system, in the vibration acquisition step, the processor acquires detection data indicating a combination of vibration propagating in the ground detected by the vibration inspection device in an inspection area for damage to a buried pipe and a detection position at which the vibration is detected. In the detection step, a breakage position in the investigation region is detected based on the vibration and the detection position indicated by the acquired two or more pieces of detection data, and the detection is performed using reference information indicating a relationship between the vibration and the detection position, and the breakage position.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

[0002] Patent Document 1 discloses a technology that uses a movable vibration applying device that applies low-frequency vibrations from the ground surface and a plurality of water leakage sound detection devices that are placed at a distance from each other on a pipe buried underground, to calculate a pseudo-failure position of the pipe from the low-frequency vibrations detected by the plurality of water leakage sound detection devices, and to calculate an actual failure position of the pipe from the water leakage sounds of the pipe detected by the plurality of water leakage sound detection devices. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 2019-39891 Summary of the Invention [Problem to be solved by the invention]

[0004] Even when it is not possible to attach a device to a pipe buried underground, as in the technology of Patent Document 1, the location of a pipe breakage can be detected from vibrations transmitted through the ground. In such cases, the way vibrations are transmitted varies depending on the underground structure, and various noises may be mixed into the vibrations transmitted through the ground. Therefore, it is difficult for anyone other than an expert to determine the location of the breakage, but the number of experts is limited.

[0005] In view of the above circumstances, the present invention provides an information processing system and the like that can detect the location of damage without the presence of an expert. [Means for solving the problem]

[0006] According to one aspect of the present invention, there is provided an information processing system including one or more processors. In this information processing system, in a vibration acquisition step, the processor acquires detection data indicating a combination of vibrations propagating through the ground detected by a vibration survey device in an investigation area for a break in a buried pipe and the detection positions at which the vibrations were detected. In a detection step, the processor detects the break position in the investigation area based on the vibrations and detection positions indicated by the two or more acquired detection data, and the detection is performed using reference information indicating the relationship between the vibrations and the detection positions and the break position.

[0007] According to this aspect, the location of the damage can be detected without the presence of an expert. [Brief explanation of the drawings]

[0008] [Figure 1] 1 is a diagram showing the overall configuration of a piping inspection support system 1. FIG. [Figure 2] FIG. 2 is a diagram illustrating a hardware configuration of the worker terminal 10. [Figure 3] FIG. 2 is a diagram illustrating a hardware configuration of a server device 30. [Figure 4] FIG. 10 is an activity diagram showing an example of a piping inspection support process. [Figure 5] FIG. 10 is a diagram illustrating an example of a remote support screen. [Figure 6] FIG. 10 is a diagram illustrating an example of a position image. [Figure 7] FIG. 10 is a flowchart illustrating an example of a function realization process. [Figure 8] FIG. 2 is a diagram showing an example of a survey result database DB1. [Figure 9] FIG. 10 is an activity diagram showing another example of the piping inspection support process. [Figure 10] FIG. 10 is a diagram illustrating an example of an output instruction. [Figure 11] FIG. 10 is a diagram showing another example of the survey result database DB1. [Figure 12] FIG. 10 is a flowchart showing another example of the function realization process. [Figure 13]FIG. 10 is an activity diagram showing another example of the piping inspection support process. [Figure 14] FIG. 10 is a diagram illustrating an example of an input screen for an influencing factor. [Figure 15] FIG. 10 is a diagram illustrating an example of a time slot table. [Figure 16] FIG. 10 is a diagram illustrating an example of a region table. [Figure 17] FIG. 10 is a diagram illustrating an example of a time slot table. [Figure 18] FIG. 10 is a diagram illustrating another example of a region table. DETAILED DESCRIPTION OF THE INVENTION

[0009] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described below with reference to the accompanying drawings. Various features shown in the following embodiments can be combined with each other.

[0010] Incidentally, the program for realizing the software appearing in one embodiment may be provided as a non-transitory computer-readable medium, or may be provided so that it can be downloaded from an external server, or may be provided so that the program is started on an external computer and its functions are realized on a client terminal (so-called cloud computing).

[0011] Furthermore, various information processing according to an embodiment may realize input and output corresponding to the input. Here, the form of information referenced in such information processing (hereinafter referred to as reference information) is not limited as long as an output is obtained as a result of the input. The reference information may be, for example, rule-based information such as a database, a lookup table, or a predetermined function (including a decision formula such as a regression formula constructed using a statistical method), a trained model that has previously trained the correlation between input and output, or a large-scale language model that can output a desired result by inputting a prompt.

[0012] In one embodiment, a "unit" may include, for example, a combination of hardware resources implemented by a circuit in the broad sense and software information processing that can be specifically realized by these hardware resources. In one embodiment, various information is handled, and this information is represented, for example, by physical values ​​of signal values ​​representing voltage and current, high and low signal values ​​as a binary bit set consisting of 0 or 1, or quantum superposition (so-called quantum bits), and communication and calculations can be performed on a circuit in the broad sense.

[0013] Furthermore, a circuit in the broad sense is a circuit realized by at least an appropriate combination of a circuit, circuitry, processor, memory, etc. The processor may be a general-purpose processor or a dedicated circuit. That is, it includes an application specific integrated circuit (ASIC), a programmable logic device (e.g., a simple programmable logic device (SPLD), a complex programmable logic device (CPLD), and a field programmable gate array (FPGA)), etc.

[0014] <Embodiment> 1. System Configuration The system configuration according to the embodiment will be described below. Fig. 1 is a diagram showing the overall configuration of the piping inspection support system 1. Fig. 1 shows an overview of each device provided in the piping inspection support system 1 and the users who use those devices. Each overview will be explained as needed, with reference to other figures.

[0015] The piping inspection support system 1 is an information processing system that performs information processing to support the work of inspecting pipes buried underground for damage. The piping inspection support system 1 is used by a worker U1 who inspects pipe damage on-site, and an expert U2 who can determine the location of the damage from vibrations transmitted through the ground. Damage inspections of pipes P1 can be conducted periodically, called planned inspections, or dynamically, called mobile inspections, which are conducted when a suspected damage is reported.

[0016] In either case, worker U1 sets the area to be investigated and performs damage investigation work in the set investigation area. In the example of Figure 1, worker U1 performs investigation work in investigation area R1 where pipe P1 is buried underground G1. A fluid W1 flows through pipe P1, and the fluid W1 leaks from a damaged location C1 and flows into the underground G1. The fluid W1 may be either a liquid or a gas, such as water (water supply and sewerage) or gas (city gas, etc.).

[0017] The piping inspection support system 1 includes a communication line 2, a vibration inspection machine 3, a worker terminal 10, an expert terminal 20, and a server device 30. The communication line 2 is not particularly limited, but may be configured, for example, by the Internet. The communication line 2 may also include a local area network, a mobile communication network, a VPN (Virtual Private Network), etc. The communication line 2 mediates the exchange of data between devices connected to the communication line. In the example of FIG. 1, the worker terminal 10 and the expert terminal 20 are connected to the communication line 2 wirelessly, and the server device 30 is connected to the communication line 2 by a wired connection. The expert terminal 20 and the server device 30 may be connected to the communication line 2 by a wired or wireless connection.

[0018] The vibration surveying machine 3 is a machine for surveying vibrations transmitted through the ground. The vibration surveying machine 3 is equipped with a vibration sensor 4 connected by a cable, and generates vibration data by amplifying the underground vibrations detected by the vibration sensor 4. A worker U1 wears the vibration surveying machine 3 around his / her waist or the like, and while holding the cable, grounds the vibration sensor 4 to detect vibrations transmitted through the ground G1. The worker U1 grounds the vibration sensor 4 at multiple positions in the survey area R1 to detect vibrations at the multiple positions. The vibration surveying machine 3 is directly connected to a worker terminal 10 wirelessly or via a wire, and transmits the generated vibration data to the worker terminal 10.

[0019] The vibration surveyor 3 also includes a position sensor 5. The position sensor 5 is a sensor that measures its own position using a Global Navigation Satellite System (GNSS). The higher the positioning accuracy of the position sensor 5, the better. It is desirable to use a D (Differential)-GNSS with a positioning error of 1 m or less, or an RTK (Real Time Kinematic)-GNSS with a positioning error of a few centimeters. In the example of FIG. 1, the position sensor 5 has a positioning error of a few centimeters. The position sensor 5 is also provided in the vibration sensor 4, and measures a position that is approximately the same as the vibration detection position. The position sensor 5 also has a wireless communication function, and wirelessly transmits position data (data indicating latitude, longitude, etc.) indicating the measured position to the worker terminal 10.

[0020] The worker terminal 10 is a terminal for which the worker U1 is the user. The worker terminal 10 is a portable terminal, such as a smartphone, tablet terminal, or laptop computer. The worker terminal 10 transmits the vibration data transmitted from the vibration surveyor 3 and the position data transmitted from the position sensor 5 to the expert terminal 20 via the communication line 2.

[0021] The expert terminal 20 is a terminal for the expert U2 as a user, and is, for example, a personal computer or a tablet. The expert terminal 20 outputs, by sound or image, the underground vibrations indicated by the vibration data transmitted from the worker terminal 10. The expert terminal 20 also outputs the detection position indicated by the transmitted position data. Based on the output sound or image, the expert U2 determines whether or not there is a break in the pipe P1, and if there is a break, the location of the break. The expert U2 also instructs the worker U1 on the position where the vibration should be detected so that it will be easier to determine whether there is a break. The expert U2 gives instructions by voice, for example.

[0022] The expert terminal 20 transmits instruction data indicating instructions from the expert U2 to the worker terminal 10 via the communication line 2. The worker terminal 10 outputs the instructions indicated by the transmitted instruction data. The instructions are output by voice, image, or the like. Based on the output instructions, the worker U1, for example, places the vibration sensor 4 at the instructed position to detect vibrations. By repeating the above operations, the expert U2 determines whether or not the pipe P1 is damaged and the location of the damage from the vibrations detected by the work of the worker U1.

[0023] The server device 30 is an information processing device that detects the location of damage on behalf of the expert U2. In the piping inspection support system 1, the expert U2 first detects the location of damage. The server device 30 accumulates the results of the expert U2's investigation into the presence or absence of damage to the pipe and the location of the damage in an investigation database DB1, and realizes a function for detecting the location of the damage based on the accumulated investigation results. The server device 30 has at least one AI module MD1 that modularizes the functions of AI (Artificial Intelligence), and realizes the detection function by the AI ​​module MD1. Details of the detection function will be described later.

[0024] In the example of Figure 1, both the worker terminal 10 and the expert terminal 20 have installed an application program (hereinafter referred to as the "support app") for using the piping inspection support system 1, and the functions of the support app perform information processing such as displaying information related to the piping inspection support system 1 and exchanging data between terminals.

[0025] 2. Hardware Configuration The hardware configuration according to the embodiment will be described below. 2 is a diagram showing the hardware configuration of worker terminal 10. Worker terminal 10 includes control unit 11, storage unit 12, communication unit 13, input unit 14, output unit 15, and bus 16. Bus 16 electrically connects the various units included in worker terminal 10.

[0026] (Control unit 11) The control unit 11 is, for example, a central processing unit (CPU) not shown. The control unit 11 realizes various functions related to the piping inspection support system 1 by reading out predetermined programs stored in the memory unit 12. In other words, information processing by software stored in the memory unit 12 is specifically realized by the control unit 11, which is an example of hardware, and can be executed as each functional unit included in the control unit 11. These will be described in further detail in the next section. Note that the control unit 11 is not limited to being single, and it may be implemented by having multiple control units 11 for each function. It may also be a combination of these.

[0027] (Storage unit 12) The memory unit 12 stores various pieces of information defined above. This can be implemented, for example, as a storage device such as a solid state drive (SSD) that stores various programs and the like related to the piping inspection support system 1 executed by the control unit 11, or as a memory such as a random access memory (RAM) that stores temporarily required information (arguments, arrays, etc.) related to the program calculations. The memory unit 12 stores various programs, variables, etc. related to the piping inspection support system 1 executed by the control unit 11.

[0028] (Communications Department 13) The communication unit 13 is configured to be able to transmit various electrical signals from the worker terminal 10 to external components. The communication unit 13 is also configured to be able to receive various electrical signals from the external components to the worker terminal 10. More preferably, the communication unit 13 has a network communication function, which may enable communication of various information between the worker terminal 10 and external devices via the communication line 2.

[0029] (Input section 14) The input unit 14 has keys, buttons, a touch screen, a mouse, etc., and receives input from the user. The input unit 14 may also have a microphone and have the function of receiving voice input from the user.

[0030] (Output section 15) The output unit 15 has a display, a speaker, etc., and displays visual information generated in a manner that is visible to the user, such as a screen, an image, an icon, text, etc., on the display surface of the display, and outputs sounds including voice. The output unit 15 also has various sensors and outputs measurement results of these sensors to the control unit 11. The output unit 15 has, for example, a positioning sensor and outputs position data indicating the position of the terminal itself. The output unit 15 also has a three-axis angle sensor and outputs angle data indicating a three-dimensional orientation. The output unit 15 also has imaging means such as a digital camera and outputs video data indicating a video of the surrounding scenery, etc.

[0031] 2 has the same hardware configuration as the worker terminal 10. In the expert terminal 20, the control unit 21 and output unit 25 are denoted by different reference numerals from those of the control unit 11 and output unit 15 of the worker terminal 10. The output unit 25 has a display, a speaker, etc., but does not need to have the various sensors that the output unit 15 has.

[0032] 3 is a diagram showing the hardware configuration of server device 30. Server device 30 includes a control unit 31, a storage unit 32, a communication unit 33, and a bus 34. Bus 34 electrically connects the various units included in server device 30. Control unit 31, storage unit 32, and communication unit 33 are similar hardware to control unit 11, storage unit 12, and communication unit 13 shown in FIG. 2, although their specifications, models, etc. may differ.

[0033] 3. Information Processing The following describes information processing according to the embodiment. In the following description, the worker terminal 10, the expert terminal 20, and the server device 30 are described as the subjects of each information processing, but these information processing operations are executed by at least one processor included in the piping inspection support system 1, i.e., the processor included in the control unit of each device (control units 11, 21, and 31). The following describes the piping inspection support processing executed when inspecting for damage to the piping P1 with reference to FIG. 4 etc.

[0034] FIG. 4 is an activity diagram showing an example of the piping inspection support process. The piping inspection support process is initiated when the expert U2 activates the expert terminal 20 and waits in a responsive state, and when the worker U1 puts on the vibration inspection device 3, picks up the worker terminal 10, and starts work in the inspection area R1. In FIG. 4, activities performed by people are represented by dotted lines. For example, first, the worker U1 grounds the vibration sensor 4 of the vibration inspection device 3 (activity A1).

[0035] Next, the vibration surveying machine 3 detects vibrations transmitted through the underground G1 using the vibration sensor 4 (activity A11). Hereinafter, the vibrations detected by the vibration surveying machine 3 will be referred to as "detected vibrations." Next, the vibration surveying machine 3 amplifies the detected vibrations (activity A12) and transmits vibration data indicating the amplified detected vibrations to the worker terminal 10. The vibration surveying machine 3 also measures the detected position using the position sensor 5 (activity A13) and transmits position data indicating the measured detected position to the worker terminal 10. Although the operations of A11 and A12 and A13 are shown aligned in a row in FIG. 3, in reality, these operations are performed in parallel and are continuously repeated until the operation of the vibration surveying machine 3 is stopped.

[0036] Next, a description will be given of the operations performed in parallel with A11 to A13 by the worker terminal 10, the expert terminal 20, and the server device 30. First, the expert terminal 20 displays a remote assistance screen in response to an operation by the expert U2 (activity A31). Fig. 5 is a diagram showing an example of a remote assistance screen. The remote assistance screen D1 shown in Fig. 5 displays a video display field E11, a map display field E12, and a vibration display field E13. The expert terminal 20 displays a map M12 including the investigation area R1 in the map display field E12 (activity A32).

[0037] The map M12 can be displayed in various ways. For example, the server device 30 stores map data for various regions in advance. When the expert U2 specifies a map including the survey area R1 based on the address of the survey area, the server device 30 reads out the specified map and displays it on the expert terminal 20 as the map M12. In the example of FIG. 5, the worker terminal 10 measures the terminal position using a positioning sensor (activity A21) and transmits position data indicating the measured terminal position to the server device 30. The server device 30 identifies a map including the terminal position indicated by the transmitted position data from the maps indicated by the stored map data (activity A41). The server device 30 transmits the identified map to the server device 30. The expert terminal 20 displays the transmitted map as the map M12.

[0038] Next, the worker terminal 10 starts capturing images using the imaging means in response to an operation by the worker U1, and transmits the captured images to the expert terminal 20 in real time (activity A22). The expert terminal 20 displays the image L11 transmitted from the worker terminal 10 in the image display field E11 (activity A33). The image L11 shows the investigation area R1 as seen by the worker U1 and the vibration sensor 4 grounded at the detection position. The transmission and display of the images may be performed using the functions of a so-called video conference service, or a similar function may be implemented in the assistance app using well-known streaming technology.

[0039] Next, the worker terminal 10 transfers the vibration data and position data transmitted from the vibration inspection machine 3 to the server device 30 (activity A23). The server device 30 generates on-site data indicating data obtained at the site based on the transferred vibration data and position data (activity A42). The on-site data is, for example, data indicating the detected vibration indicated by the vibration data in association with the position indicated by the position data. The server device 30 transmits the generated on-site data to the expert terminal 20 and stores it in its own inspection result database DB1 (activity A43). The expert terminal 20 outputs the detected vibration and detected position indicated by the transmitted on-site data (activity A34).

[0040] For example, as shown in FIG. 5, the expert terminal 20 displays a position image F11 representing the detection position indicated by the site data, superimposed on a map displayed in the map display field E12. The expert terminal 20 also outputs a sound (hereinafter referred to as "detected sound") representing the detected vibration indicated by the site data from a speaker. The expert terminal 20 also outputs a waveform image N13 representing the waveform of the detected sound in the vibration display field E13. While the detected sound is output as a sound in the audible range, the waveform image N13 may also represent a waveform of a sound including sounds outside the audible range. Such a waveform image N13 can represent the characteristics of the sound even if the fluid W1 leaking from the damaged location C1 emits a sound outside the audible range. Based on the output detected sound and waveform image N13, the expert U2 determines whether the pipe P1 is damaged and, if so, the location of the damage.

[0041] The expert terminal 20 displays an operation image for controlling the output of the detected sound in the vibration display field E13. The expert terminal 20 adjusts the volume of the detected sound by operating the operation image. Furthermore, by saving the output site data, the expert terminal 20 outputs the detected sound in the following ways: pausing, rewinding, fast-forwarding, rewinding a predetermined time ago (five seconds ago in the example of FIG. 5), and advancing a predetermined time later (five seconds later in the example of FIG. 4). This allows the user to listen to the detected sound again or return the output sound to the currently detected sound.

[0042] Next, the expert terminal 20 accepts input of instructions from the expert U2 (activity A35). The instructions from the expert U2 are given, for example, by voice. The content of the instructions is, for example, an instruction on the position where the vibration sensor 4 should be grounded. The expert U2 looks at the map to determine the direction and movement distance of the next position where vibration should be detected, and gives an instruction such as, for example, "Measure the position 30 cm eastward." Furthermore, the expert U2 can determine the direction the worker U1 is facing by looking at the image L11, so he can also give an instruction such as, for example, "Measure the position 30 cm diagonally forward to the right."

[0043] When the expert terminal 20 receives an instruction, it generates instruction data indicating the received instruction (activity A36). The expert terminal 20 generates, for example, voice data indicating the voice of the instruction as the instruction data. The expert terminal 20 transmits the generated instruction data to the worker terminal 10. The worker terminal 10 outputs the instruction indicated by the transmitted instruction data (activity A24). The worker terminal 10 outputs, for example, the voice of the instruction indicated by the instruction data.

[0044] The worker U1 performs the work in accordance with the output instructions. For example, the worker U1 determines the position where vibration should be detected next from the output voice instruction (activity A2). Then, the worker U1 returns to A1 and places the vibration sensor 4 on the determined position. Thereafter, the operations of A1, A11, A12, A13, A22, A23, A42, A43, A33, A34, A35, A36, A24, and A2 are repeated. When the detection position moves, the expert terminal 20 displays a position image as follows:

[0045] FIG. 6 is a diagram showing an example of a position image. In the map display field E12 shown in FIG. 6, a position image F11, position images F21 and F22, etc. are displayed. The position image F11 indicates the position where underground G1 vibrations are currently detected. The position images F21 and F22, etc. indicate positions where underground G1 vibrations have been detected in the past. In this way, by displaying position images showing past detection positions, it is easier to perform the task of detecting underground G1 vibrations more comprehensively than when only a position image showing the current detection position is displayed.

[0046] The expert U2 identifies the location of the damage in the pipe P1 through the above work. The expert U2 inputs the identified location of the damage into the expert terminal 20. The expert terminal 20 accepts the input of the identified location of the damage (activity A37) and notifies a predetermined destination of the input location of the damage (activity A38). The predetermined destination is, for example, a construction company that repairs the damage in the pipe P1 in the investigation area R1. When the construction company that received the notification checks for the presence or absence of a damage at the notified location of the damage in the investigation area R1 and repairs it, the damage disappears and the leakage of the fluid W1 stops.

[0047] The expert terminal 20 also transmits damage location data indicating the input damage location to the server device 30. The server device 30 stores the transmitted damage location data in the investigation results database DB1 (activity A44). As described above, data related to the water leak investigation (site data, instruction data, and damage location data) is stored in the investigation results database DB1, and the server device 30 uses this stored data to realize the function of detecting the damage location on behalf of the expert U2.

[0048] The following describes a function implementation process in which the server device 30 implements a damage location detection function, and an automatic detection process in which the server device 30 detects the damage location using the implemented detection function. 7 is a flow diagram showing an example of the function realization process. The function realization process is executed in a state where sufficient investigation results have been accumulated in the investigation result database DB1. The server device 30 first acquires detection data for the teacher (step S11). The server device 30 acquires, for example, data stored in the investigation result database DB1 as the detection data for the teacher.

[0049] FIG. 8 is a diagram showing an example of the investigation result database DB1. The investigation result database DB1 stores site data and damage location data in association with the investigation area. The site data includes the detection date and time, the detection location, and the detected vibration transmitted through the ground. The damage location data stores the latitude and longitude values ​​of the damage location identified by the expert U2 or the damage location measured by the construction company that repaired the damage.

[0050] The server device 30 first acquires the detected positions and detected vibrations associated with a certain investigation area as detection data for teacher. Next, the server device 30 acquires damage position data for the same investigation area as the detection data as damage position data for teacher (step S12). This detection data and damage position data is teacher data representing pairs of "example questions" and "correct answers," and is also called training data or training data.

[0051] Next, the server device 30 executes machine learning processing to enable the AI ​​module MD1 shown in Fig. 1 to be used as an AI for detecting the location of damage (step S13). The machine learning processing is processing to instruct the detection AI to execute machine learning, for example, processing to instruct the execution of machine learning in which a combination of detected vibrations and detection locations detected in the same investigation area is used as input and the location of damage detected in that investigation area is used as output. The detection AI executes machine learning according to instructions from the server device 30 and generates a learning model that represents the correlation between the input and the output.

[0052] Next, the server device 30 determines whether there is unlearned data (step S14). Unlearned data is, for example, detection data and damage location data of the investigation area that have not yet been acquired from the investigation result database DB1. If the server device 30 determines that there is unlearned data (YES), it returns to step S11 and continues operation. By repeating the operations from S11 to S13 in this manner, machine learning is performed using the detection data and damage location data of the multiple investigation areas stored in the investigation result database DB1.

[0053] 7, training data is acquired for each survey area and machine learning is performed, but training data for multiple survey areas may be acquired collectively and machine learning may be performed as long as pairs of example questions and correct answers are associated with each other. In addition, the server device 30 may perform preprocessing such as removing noise or outliers on the training data used for machine learning.

[0054] If the server device 30 determines in step S14 that there is no unlearned data (NO), it completes the machine learning. Note that if new detection data and damage location data are accumulated after the machine learning is completed, the server device 30 may perform additional machine learning processing using the data. The server device 30 then ends the function realization processing. Next, the server device 30 performs automatic detection processing to detect the damage location using the detection AI for which machine learning has been completed. The automatic detection processing is performed, for example, as part of the piping inspection support processing.

[0055] Fig. 9 is an activity diagram showing another example of the piping inspection support process. Unlike the example of Fig. 4, the example of Fig. 9 does not use the expert terminal 20. However, as in the example of Fig. 4, a worker U1 grounds the vibration sensor 4 in the inspection area R1 (activity A1) to detect vibrations transmitted through the ground G1. Next, the vibration inspection machine 3 performs the operations A11, A12, and A13 shown in Fig. 4, and vibration data indicating the detected vibrations and position data indicating the detected positions are transmitted to the server device 30 via the worker terminal 10.

[0056] The server device 30 acquires the transmitted vibration data and position data as detection data (activity A51). The detection data is data indicating a combination of vibrations detected by the vibration inspection device 3 in the investigation area R1 and the detection positions at which the vibrations were detected. Next, the server device 30 inputs the acquired detection data to the detection AI (activity A52). More specifically, the server device 30 inputs all of the detection data acquired in the same investigation area R1 into the detection AI. The detection AI has a learning model generated by the machine learning process shown in Figure 7, and outputs the damage position based on the detected vibrations and detection positions indicated by the input detection data. The server device 30 acquires the damage position output by the detection AI (activity A53).

[0057] Here, the detection AI calculates and outputs a score for the output damage location along with the damage location. The damage location score is a value that indicates the likelihood of the output damage location, and hereinafter, the higher the score, the more likely the damage location is. The server device 30 determines whether the score calculated by the detection AI is equal to or greater than a threshold value (activity A54). In the following, for ease of explanation, an example will be described in which the damage location score is calculated as a value from 0 to 100, and the damage location is detected with a threshold value of 90.

[0058] If the server device 30 determines that the score is less than the threshold, it generates instruction data instructing the worker to change the detection position (activity A55) and transmits the generated instruction data to the worker terminal 10. The worker terminal 10 outputs the instruction indicated by the transmitted instruction data (activity A24). Fig. 10 is a diagram showing an example of an output instruction. In the example of Fig. 10, the worker terminal 10 displays a field work screen D2. The field work screen D2 displays a status display field E21, a score display field E22, a map display field E23, and an instruction display field E24.

[0059] The status display field E21 displays the detection status of the damage position. In the example of FIG. 10, a character string indicating the detection status, "Not found yet," is displayed. The score display field E22 displays the score of the damage position calculated by the detection AI. In the example of FIG. 10, 65 points out of 100 points is displayed as the score. By looking at this score, worker U1 can get an idea of ​​how close he or she is to the detection position where the damage position can be detected.

[0060] The map display field E23 displays a map M12 including the survey area, a position image F11 indicating the current detection position, and position images F21 and F22 indicating past detection positions. The instruction display field E24 displays a character string indicating an instruction, "Please move the detection position." The worker U1 moves the detection position in accordance with the displayed instruction. In this case, by looking at the map display field E23, the worker U1 can determine that a position where vibration has not yet been detected is the next position where vibration should be detected.

[0061] In this way, detection of vibrations in the investigation area R1 is repeated until the score of the damage location output by the detection AI becomes equal to or greater than the threshold. When the server device 30 determines in A54 that the score of the damage location is equal to or greater than the threshold, it generates completion notification data indicating that the investigation of the damage to the pipe has been completed (activity A56) and transmits the generated completion notification data to the worker terminal 10. The worker terminal 10 outputs a notification indicated by the transmitted completion notification data (activity A25). The completion notification data is, for example, data indicating a character string indicating the end of the detection work. Then, the server device 30 notifies a predetermined destination (such as a construction company) of the identified damage location (activity A57) and ends the piping investigation support process.

[0062] The determination of A54 may be made by the detection AI. In this case, the detection AI outputs the crushing position when the score is equal to or greater than the threshold, and outputs, for example, only the score when the score is less than the threshold. The server device 30 performs A56 (generates and transmits completion notification data) when the crushing position is output, and performs A55 (generates and transmits instruction data) when only the score is output. In this case, too, the position output by the detection AI is detected as the damage position.

[0063] As described above, the server device 30 functions as an example of a vibration acquisition unit that acquires detection data indicating detected vibrations and detection positions. Specifically, the detection data is data indicating a combination of vibrations traveling through the ground detected by the vibration survey device 3 in an area surveyed for damage to buried pipes and the detection positions at which the vibrations were detected. In the example of FIG. 9, the server device 30 acquires detection data for the survey area R1 at A51, and two or more pieces of detection data are acquired by having the worker U1 perform detection at two or more positions.

[0064] The server device 30 functions as an example of a detection unit that detects the location of a break in a pipe in an investigation area based on the vibration and detection position indicated by the two or more pieces of acquired detection data. Specifically, the server device 30 (an example of a detection unit) inputs the vibration and detection position indicated by the two or more pieces of acquired detection data to a detection AI, and detects the location output by the detection AI as the break location in the investigation area. The AI ​​module MD1 shown in Figure 1 is an example of a detection AI.

[0065] For machine learning of the detection AI, first, the server device 30 functions as an example of a correct answer acquisition unit that acquires correct answer data indicating the location of damage in the investigation area. The server device 30 acquires the training damage location data shown in FIG. 7 as correct answer data. Next, the server device 30 functions as an example of a detection learning unit that causes the detection AI to perform machine learning to generate a detection learning model. More specifically, the server device 30 (an example of a detection learning unit) causes the detection AI to perform machine learning that takes as input the vibrations and detection positions indicated by two or more pieces of detection data acquired in the investigation area and outputs the damage location indicated by the acquired correct answer data. According to this embodiment, it is possible to detect the location of fracture by utilizing the results of previous detection of the location of damage.

[0066] By performing machine learning as described above, the detection AI is provided with a detection learning model that has learned the relationship (correlation) between vibrations detected in multiple investigation areas, the detection positions of those vibrations, and the damage positions in the multiple investigation areas. Various AI engines are available for the detection AI. In recent years, AI engines with various performance and cost have been provided, so according to the above aspect, it is possible to select and use the most appropriate one from these various AI engines and detect the damage position using AI.

[0067] As described above, the server device 30 (an example of a detection unit) detects the location of damage using reference information that represents the relationship between the detected vibration and the detected position and the location of damage. For example, the detection learning model is an example of reference information because it represents the relationship between the detected vibration and the detected position and the location of damage. By using such reference information, the location of damage can be detected without the expert U2.

[0068] <Variation: Reference Information> The reference information is not limited to the above-described detection learning model. For example, the reference information may be information that associates sample data indicating vibrations detected at a plurality of detection positions on the ground surface around the damage position with the distance from each detection position to the damage position.

[0069] In this case, the server device 30 calculates the similarity between the detected vibration and the vibration indicated by the sample data, and if there is sample data for which the calculated similarity is equal to or greater than a threshold, determines that the damage location is in the vicinity of the detection location of the detected vibration. More specifically, the server device 30 determines that the damage location is in a position on a circumference centered on the detection location of the detected vibration, a distance away from the detection location of the detected vibration that is associated with the sample data of the vibration similar to the detected vibration. When the server device 30 also draws this circumference for the surrounding detection locations, it detects the position where the circumference overlaps more as the damage location.

[0070] In addition to the above, the correlation between the detected vibrations at two or more detection positions indicated by two or more sample data and the damage position may be analyzed, and an algorithm may be created to calculate the damage position from the two or more detected vibrations based on the analyzed correlation, and the created algorithm may be used as the reference information. In short, any information may be used as the reference information as long as it represents the correlation between the damage position and a combination of two or more detected vibrations and detection positions.

[0071] <Variation: Element Information> In the above example, the detected vibration and detected position were used to detect the location of the damage, but if there are factors (hereinafter referred to as "influencing factors") that affect the detected vibration due to the damage to the pipe, taking these influencing factors into consideration will improve detection accuracy. Influences on vibration include effects that change the amplitude, wavelength, or phase of the vibration, effects on the path the vibration travels (reflection, blocking, etc.), and effects on the transmission speed of the vibration.

[0072] Influencing factors include, for example, the depth to which the pipe is buried. The deeper the pipe, the greater the three-dimensional distance (including depth) even if the two-dimensional distance (the distance between locations indicated by latitude and longitude) is close, resulting in vibration attenuation. Other influencing factors include the pipe's material and diameter. When fluid leaks from a damaged area, the fluid vibrates the pipe, generating vibrations underground. The pipe's material and diameter affect the vibrations, for example, the harder the pipe's material and the smaller its diameter, the higher the frequency. Another influencing factor is the pavement material surrounding the pipe. The density and moisture content of the pavement, such as asphalt, concrete, resin-based mixture, or natural soil pavement (each of which is further classified into smaller types), vary, affecting the vibrations by varying the speed at which vibrations are transmitted underground.

[0073] The server device 30 detects the location of the damage based on the influencing factors as described above. First, the server device 30 functions as an example of an element acquisition unit that acquires element information indicating influencing factors related to the pipe to be inspected for damage. The element information is assumed to be stored in advance in, for example, the investigation result database DB1.

[0074] FIG. 11 is a diagram showing another example of the investigation result database DB1. In the example of FIG. 11, the investigation result database DB1 stores, as element information, information indicating the pipe depth, pipe material, pipe diameter, and pavement material. For example, the pipe depth determined by an expert U2 from detected vibrations or the depth measured by the contractor who performed the work on the damaged area is input into the server device 30 and stored in the investigation result database DB1. Furthermore, element information such as the pipe material, diameter, and pavement material around the pipe is input into the server device 30 based on, for example, publicly or commercially provided design drawings of the site, and stored in the investigation result database DB1. Note that while it is desirable for all element information to be available, partial information may be sufficient if the necessary information is not stored and cannot be collected.

[0075] FIG. 12 is a flow diagram showing another example of the function realization processing. In the example of FIG. 12, the server device 30 acquires element information for the teacher (step S15) after S11 (acquisition of detection data for the teacher) shown in FIG. 7. Note that S11 and S15 may be performed simultaneously or in reverse order. Next, the server device 30 acquires damage location data for the teacher in S12. In the example of FIG. 12, the server device 30 acquires, as damage location data for the teacher, information indicating the depth of the pipe included in the element information (i.e., data indicating the damage location in three dimensions) in addition to the latitude and longitude indicated by the damage location data stored in the investigation result database DB1.

[0076] In S13, the server device 30 executes a machine learning process instructing the execution of machine learning that inputs the combination of detected vibrations and detection positions detected in the same survey area as well as the acquired element information and outputs the damage position including the depth indicated by the acquired damage position data. As a result, the detection AI generates a learning model that represents the correlation between two or more detected vibrations and detection positions on the ground surface, the element information, and the three-dimensional damage position of the damage area. The server device 30 executes the automatic detection process described in FIG. 9 using the detection AI having the learning model thus generated.

[0077] Fig. 13 is an activity diagram showing another example of the piping inspection support process. In the example of Fig. 13, in addition to the operations shown in Fig. 9, the worker terminal 10 displays an input screen for inputting influencing factors (activity A26). The input screen is displayed before the start of vibration detection work in the inspection area R1.

[0078] FIG. 14 is a diagram showing an example of an input screen for influencing factors. The on-site work screen D3 shown in FIG. 14 displays input fields E31, E32, E33, and E34 for impact information and a confirm button B31. The input fields E31, E32, E33, and E34 are used to input element information, such as "pipe depth," "pipe material," "pipe diameter," and "pavement material," respectively. It is assumed that the worker U1 has previously obtained information indicating the influencing factors before conducting an investigation in the investigation area R1. When the worker U1 inputs the information into each input field and operates the confirm button B31, the worker terminal 10 accepts the input of the influencing factors (activity A27) and notifies the server device 30 of the input influencing factors.

[0079] The server device 30 acquires and stores element information indicating the notified influencing element (activity A61). After that, when the server device 30 acquires the detection data (A51), it inputs the acquired element information to the detection AI in addition to the acquired detection data (activity A62). The detection AI outputs the damage location based on the detected vibration and detected position indicated by the input detection data and the influencing element indicated by the element information. Subsequently, the same operation as in the example of FIG. 9 is performed, and the damage location is detected.

[0080] As described above, in the example of FIG. 13 etc., the server device 30 (an example of a detection unit) detects the location of damage based on the vibration and detection position indicated by the two or more pieces of acquired detection data, as well as the elements indicated by the acquired element information. In this case, the reference information used by the server device 30 indicates the relationship between the detected vibration and detection position, the location of damage in the pipe, and the elements that affect the vibration detected due to damage in the pipe. In the example of FIG. 13 etc., the detection learning model generated by the function realization process shown in FIG. 12 is an example of reference information. According to this aspect, the location of damage is detected based on the influencing elements as well, thereby improving detection accuracy compared to when the influencing elements are not used.

[0081] The element information includes information indicating the depth at which the pipe is buried. As described above, when the pipe is deep, vibrations are attenuated. Therefore, if the pipe depth is not used as element information, a nearby pipe is likely to be detected as if it were far away. In comparison with this case, a detection mode based on element information can reduce detection errors due to the pipe depth.

[0082] It is desirable that both the depth information used in machine learning and the depth information used to detect the damage position be accurate, but it is possible to detect the damage position even if there is some error.At the very least, compared to detecting the damage position assuming that the depth of the damage position is 0, by taking the depth into consideration, the distance from the detection position to the damage position (position including depth) is more accurately reflected in the correlation, thereby improving detection accuracy.

[0083] <Variation: Reference information according to time zone and region> The vibrations that occur in the survey area change depending on the time of day. For example, during the day, there are cars and people passing by, so the vibrations they generate are likely to be included as noise, and in the morning and evening, the flow rate in the pipes increases due to meals and baths, so the vibrations generated by the fluid tend to become larger. Therefore, the correlation indicated by the above-mentioned reference information changes depending on the time of day. Therefore, different reference information may be used depending on the time of day.

[0084] In this case, the server device 30 functions as an example of a time zone identification unit that identifies the time zone in which vibration was detected by the vibration inspection device 3. Then, the server device 30 (an example of a detection unit) performs detection using different reference information depending on the identified time zone. More specifically, the server device 30 acquires detection data indicating the detected vibration for each time zone, and causes the detection AI to perform machine learning using the detection data for each time zone to generate a detection learning model. Then, the server device 30 detects the location of damage using a time zone table that associates time zones with detection learning models.

[0085] FIG. 15 is a diagram showing an example of a time period table. In the time period table TB1 shown in FIG. 15, the time periods "morning," "daytime," "dinnertime," and "late night" are associated with "detection learning model α1," "detection learning model α2," "detection learning model α3," and "detection learning model α4." These detection learning models are generated in advance using the detection data and damage location data for each time period and stored in server device 30. When detection data is transmitted to server device 30, server device 30 first identifies the time period based on the current time.

[0086] Next, the server device 30 reads out the detection learning model associated with the identified time period in the time period table TB1. The server device 30 then instructs the detection AI to output the damage location using the read out detection learning model. According to this embodiment, the damage location is detected based on the detected vibration that is more suited to the time period in which the detected vibration was detected and the correlation between the detected position and the damage location, thereby improving the accuracy of damage location detection compared to when the same reference information is used for all time periods.

[0087] Furthermore, the vibrations occurring in the survey area vary depending on the region. For example, noise tends to be louder in busy areas due to the higher volume of traffic and foot traffic than in residential areas, and noise caused by cars tends to be louder along major roads. Furthermore, noise caused by rivers or subways tends to be louder in areas with nearby rivers or subways. Therefore, the correlation indicated by the reference information described above also varies depending on the region. Therefore, different reference information may be used depending on the region.

[0088] In this case, the server device 30 functions as an example of a region identification unit that identifies a region where vibrations have been detected by the vibration surveyor 3. The server device 30 (an example of a detection unit) then performs detection using reference information that differs depending on the identified region. More specifically, the server device 30 acquires detection data indicating detected vibrations for each region, and causes the detection AI to perform machine learning using the detection data for each region to generate a detection learning model. The server device 30 then detects the location of damage using a region table that associates regions with detection learning models.

[0089] FIG. 16 is a diagram illustrating an example of a region table. In the region table TB2 illustrated in FIG. 16, "detection learning model β1," "detection learning model β2," "detection learning model β3," and "detection learning model β4" are associated with the regions of "downtown," "residential area," "along main roads," and "along rivers." These detection learning models are generated in advance using the detection data and damage position data for each region and stored in the server device 30. In the example of FIG. 16, for example, the worker terminal 10 transmits its own terminal's position data to the server device 30, and the server device 30 identifies the region including the detection position based on the position indicated by the transmitted position data.

[0090] Next, the server device 30 reads out the detection learning model associated with the identified region in the region table TB2. The server device 30 then instructs the detection AI to output the damage location using the read out detection learning model. According to this aspect, the damage location is detected based on the detected vibration that is more suited to the region where the detected vibration is detected and the correlation between the detected position and the damage location, thereby improving the accuracy of damage location detection compared to when the same reference information is used for all regions.

[0091] Furthermore, the server device 30 may perform detection using different reference information depending on the combination of time period and region. In this case, the server device 30 may detect the damage location using a time period / region table that associates the combination of time period and region with the detection learning model. Note that the time period and region described above are merely examples and are not limiting. The above-described element information may also be included in the reference information. In short, it is sufficient that reference information is prepared for each time period or region in which there is a common correlation between the detected vibration and detection position and the damage location.

[0092] <Variation: Pretreatment> The server device 30 detects the location of the damage based on the detected vibrations indicated by the vibration data, and may perform preprocessing on the detected vibrations. Specifically, the server device 30 may function as an example of a preprocessing unit that performs preprocessing to reduce noise-indicating portions of the vibrations indicated by the acquired detection data. The noise referred to here refers to vibrations other than those generated by fluid flowing through the pipe or fluid leaking from the location of the damage, such as vibrations generated by cars or people passing nearby or water flowing in a river.

[0093] A specific example of preprocessing will be described. Preprocessing is performed, for example, by a preprocessing AI. In this case, the server device 30 acquires, for example, vibration data acquired in the same survey area, vibration data containing a lot of noise and vibration data with little noise. The vibration data containing a lot of noise is, for example, vibration data acquired when there are cars or people around, and the vibration data with little noise is, for example, vibration data acquired when there are no cars or people around. Furthermore, for noise that is always included, such as noise from rivers, for example, vibration data with little noise is generated by manually performing a process to identify and remove the noisy parts from the vibrations.

[0094] The server device 30 inputs the acquired vibration data and causes the preprocessing AI to perform machine learning, which defines vibration data with a noise ratio below a threshold as correct data. Through this machine learning, the preprocessing AI generates a preprocessing learning model that represents the correlation between vibration data with a lot of noise and vibration data with little noise, among vibration data indicating vibrations detected in the same survey area. The server device 30 inputs the acquired detection data to the preprocessing AI having the preprocessing learning model thus generated, and acquires the vibration data output by the preprocessing AI as vibration data indicating preprocessed vibrations.

[0095] Then, the server device 30 (an example of a detection unit) detects the damage position based on the preprocessed vibration. Specifically, the server device 30 (an example of a preprocessing unit) inputs the acquired detection data to a preprocessing AI equipped with the generated preprocessing learning model, and sets the vibration indicated by the vibration data output by this preprocessing AI as the preprocessed vibration.

[0096] The preprocessing method is not limited to the method using preprocessing AI. For example, the server device 30 may store noise data indicating only pre-measured noise, and perform preprocessing to reduce portions of the detected vibrations indicated by the vibration data acquired in the investigation area that are similar to the vibrations indicated by the noise data. Other well-known methods for removing noise from detected vibrations may also be used. In either case, noise is reduced from the detected vibrations, thereby improving detection accuracy compared to when preprocessing is not performed.

[0097] <Variation: Preprocessing according to time zone and region> In the above example, the reference information is changed depending on the time period or region, but the pre-processing may also be changed in a similar manner. When the time period is used, the server device 30 (an example of a time period identification unit) first identifies the time period in which the vibration was detected by the vibration inspection device 3. Then, the server device 30 (an example of a pre-processing unit) performs different processing as pre-processing depending on the identified time period.

[0098] More specifically, the server device 30 acquires detection data indicating detected vibrations for each time period, and causes the preprocessing AI to perform machine learning using the detection data for each time period to generate a preprocessing learning model. The server device 30 then performs preprocessing using a time period table that associates time periods with preprocessing learning models.

[0099] Fig. 17 is a diagram showing an example of a time period table. In the time period table TB3 shown in Fig. 17, the time periods "morning," "daytime," "dinnertime," and "late night" are associated with "preprocessing learning model γ1," "preprocessing learning model γ2," "preprocessing learning model γ3," and "preprocessing learning model γ4." These preprocessing learning models are generated in advance using detection data for each time period and stored in the server device 30. When detection data is transmitted to the server device 30, the server device 30 first identifies the time period based on the current time.

[0100] Next, the server device 30 reads out the preprocessing learning model associated with the identified time period in the time period table TB3. The server device 30 then instructs the preprocessing AI to execute preprocessing using the read preprocessing learning model. According to this embodiment, preprocessing that is more suited to the time period in which the detected vibration is detected is executed. This more effectively reduces noise than when the same preprocessing is executed for all time periods. As a result, the accuracy of detecting the damage location can be improved.

[0101] Next, a case where a region is used will be described. In this case, first, the server device 30 (an example of a region identification unit) identifies the region where vibration was detected by the vibration surveyor 3. Then, the server device 30 (an example of a detection unit) performs different processing as preprocessing depending on the identified region. More specifically, the server device 30 acquires detection data indicating the detected vibration for each region, and causes the preprocessing AI to perform machine learning using the detection data for each region to generate a preprocessing learning model. Then, the server device 30 performs preprocessing using a time period table that associates regions with preprocessing learning models.

[0102] FIG. 18 is a diagram showing another example of a region table. In the region table TB4 shown in FIG. 18, "preprocessing learning model δ1," "preprocessing learning model δ2," "preprocessing learning model δ3," and "preprocessing learning model δ4" are associated with the regions "downtown," "residential area," "along main roads," and "along rivers." These preprocessing learning models are generated in advance using detection data for each region and stored in the server device 30. The server device 30 identifies the region that includes the detected position, similar to the example of FIG. 16.

[0103] Next, the server device 30 reads out the preprocessing learning model associated with the identified region in the region table TB4. The server device 30 then instructs the preprocessing AI to execute preprocessing using the read preprocessing learning model. According to this embodiment, preprocessing that is more suited to the region where the detected vibration is detected is executed, so noise is reduced more effectively than when the same preprocessing is executed in every region, thereby improving the accuracy of detecting the damage location.

[0104] Furthermore, the server device 30 may perform different preprocessing processes depending on the combination of time zones and regions. In this case, the server device 30 may perform preprocessing using a time zone / region table that associates combinations of time zones and regions with preprocessing learning models. Note that the time zones and regions described above are merely examples and are not limiting. Different preprocessing processes may also be performed depending on the element information described above. In short, it is sufficient that different preprocessing processes are performed for each time zone or region where the noise contained in the detected vibrations has commonality.

[0105] <Example of variation: Variation of composition> The configuration (overall configuration, hardware configuration, functional configuration, etc.) shown in FIG. 1 and elsewhere is an example, and other configurations may be used as long as they are not inconvenient for implementation. For example, the worker terminal 10, the expert terminal 20, and the server device 30 may each be distributed across two or more devices. The server device 30 may also be provided in the form of SaaS (Software as a Service) or a cloud computing system. In the example of FIG. 1, the server device 30 includes the AI ​​module MD1, but an AI included in an external device may be used, or an AI service provided by an external business may be used.

[0106] Furthermore, the information processing performed by each information processing device may be performed by or controlled by another information processing device. For example, the worker terminal 10 and the expert terminal 20 log in to the server device 30 and display a system screen provided by the server device 30. The worker terminal 10 functions as a UI (User Interface) for the worker U1, and the expert terminal 20 functions as a UI for the expert U2, but the displayed screen, information, and data exchange are all controlled by the server device 30. Furthermore, instead of the server device 30, either the worker terminal 10 or the expert terminal 20 may control the information processing of the entire piping inspection support system 1.

[0107] Furthermore, some of the processing that was previously performed by the worker terminal 10 (for example, sending vibration data and position data to the expert terminal 20 or server device 30) may be performed by the vibration surveyor 3 or the position sensor 5. Also, a device that integrates the vibration surveyor 3, the position sensor 5, and the worker terminal 10 may be used. Furthermore, video may be captured not by the worker terminal 10 but by a separate digital camera or a wearable camera, etc. In short, as long as the necessary information processing is performed by the entire piping inspection support system 1, the devices that perform this information processing may have any configuration.

[0108] The output destination of information or data (hereinafter referred to as "information, etc.") may be another device, a display, a memory unit (including an internal memory unit and an external memory unit), an email address, an account of another system, etc. Acquisition of information, etc. includes not only acquiring information, etc. transmitted from another device, but also acquiring information, etc. generated by the device itself. Furthermore, the table, etc. (table, database, etc.) in which parameters are associated is not limited to the illustrated table, etc., and the number of parameters may be reduced or increased. Furthermore, information, etc. corresponding to parameters may be obtained using a mathematical formula, a conditional formula, etc., without using a table, etc.

[0109] The above-described embodiments are information processing devices such as the worker terminal 10, the expert terminal 20, and the server device 30, and information processing systems such as the piping inspection support system 1 that include these devices and terminals (including those configured in a single housing as well as those configured in multiple housings as long as they include one or more processors), but they may also be information processing methods. The information processing method includes the same steps as those executed by the information processing system. The above-described embodiments may also be programs. The program causes a computer to execute the same steps as those executed by the information processing system.

[0110] <Additional Notes> Furthermore, it may be provided in the following aspects.

[0111] (1) An information processing system having one or more processors, wherein, in a vibration acquisition step, the processor acquires detection data indicating a combination of vibrations traveling through the ground detected by a vibration survey device in an area to be investigated for damage to buried pipes and the detection position at which the vibrations were detected, and in a detection step, detects the damage position in the investigation area based on the vibrations and detection positions indicated by the two or more acquired detection data, and the detection is performed using reference information indicating the relationship between the vibrations and the detection positions and the damage position.

[0112] According to this aspect, the location of the damage can be detected without the presence of an expert.

[0113] (2) In the information processing system described in (1) above, in the detection step, the processor inputs the vibrations and detection positions indicated by the two or more acquired detection data into a detection AI, and performs the detection using the position output by the detection AI as the damage position in the investigation area, and the detection AI is provided with a detection learning model as the reference information that has learned the relationship between the vibrations detected in the multiple investigation areas and the detection positions of the vibrations and the damage positions in the multiple investigation areas.

[0114] According to this embodiment, the location of damage can be detected using AI.

[0115] (3) In the information processing system described in (2) above, in the correct answer acquisition step, the processor acquires correct answer data indicating the location of damage in the investigation area, and in the learning step, causes the detection AI to perform machine learning using the vibrations and detection positions indicated by two or more of the detection data acquired in the investigation area as input and the location of damage indicated by the acquired correct answer data as output, thereby generating the detection learning model.

[0116] According to this aspect, it is possible to make use of the results of previous detection of the damage position.

[0117] (4) In the information processing system described in any one of (1) to (3) above, the reference information represents the relationship between the vibration and the detection position, and the damage position, as well as elements that affect the vibration detected due to damage to the pipe, and in the element acquisition step, the processor acquires element information indicating the elements related to the pipe to be inspected for damage, and in the detection step, performs the detection based on the elements indicated by the acquired element information in addition to the vibration and detection position indicated by the acquired two or more detection data.

[0118] According to this aspect, it is possible to improve the detection accuracy.

[0119] (5) In the information processing system described in (4) above, the element information includes information indicating the depth at which the pipe is buried.

[0120] According to this aspect, it is possible to reduce detection errors due to the depth of the pipe.

[0121] (6) In the information processing system described in any one of (1) to (5) above, in the time zone identification step, the processor identifies the time zone in which vibration was detected by the vibration inspection device, and in the detection step, performs the detection using the reference information that differs depending on the identified time zone.

[0122] According to this aspect, it is possible to reduce detection errors caused by differences in noise between time periods.

[0123] (7) In the information processing system described in any one of (1) to (6) above, in the area identification step, the processor identifies the area where vibrations were detected by the vibration surveying device, and in the detection step, performs the detection using the reference information that differs depending on the identified area.

[0124] According to this embodiment, it is possible to reduce detection errors caused by differences in noise between regions.

[0125] (8) In the information processing system described in any one of (1) to (7) above, in the preprocessing step, the processor performs preprocessing to reduce the noise-indicating portion from the vibration indicated by the acquired detection data, and in the detection step, performs the detection based on the vibration on which the preprocessing has been performed.

[0126] According to this aspect, it is possible to improve the detection accuracy.

[0127] (9) In the information processing system described in (8) above, in the time zone identification step, the processor identifies the time zone in which vibration was detected by the vibration inspection device, and in the preprocessing step, performs different processing as the preprocessing depending on the identified time zone.

[0128] According to this aspect, noise can be effectively reduced regardless of the time of day.

[0129] (10) In the information processing system described in (8) or (9) above, in the area identification step, the processor identifies the area where vibrations were detected by the vibration surveying device, and in the preprocessing step, performs different processing as the preprocessing depending on the identified area.

[0130] According to this embodiment, noise can be effectively reduced regardless of the region.

[0131] (11) An information processing method, in which a processor included in an information processing system executes each step of the information processing system described in any one of (1) to (10) above.

[0132] According to this aspect, the location of the damage can be detected without the presence of an expert.

[0133] (12) A program that causes a computer to execute each step of the information processing system according to any one of (1) to (10) above.

[0134] According to this aspect, the location of the damage can be detected without the presence of an expert. Of course, this is not the case. Furthermore, the above-described embodiments and modifications may be combined in any desired manner.

[0135] Finally, while various embodiments of the present invention have been described, these are presented by way of example only and are not intended to limit the scope of the invention. The novel embodiments may be embodied in various other forms, and various omissions, substitutions, and modifications may be made without departing from the spirit of the invention. The embodiments and their modifications are intended to be included within the scope and spirit of the invention, as well as within the scope of the inventions and their equivalents as defined in the appended claims. [Explanation of symbols]

[0136] 1: Piping inspection support system 2: Communication line 3: Vibration survey machine 4: Vibration sensor 5: Position sensor 10: Worker terminal 11: Control section 20: Expert terminal 21: Control unit 30: Server device 31: Control unit

Claims

1. An information processing system comprising one or more processors, the processor: In the vibration acquisition step, detection data indicating a combination of vibrations transmitted through the ground detected by the vibration surveying device in an area to be investigated for damage to the buried pipe and a detection position where the vibrations were detected is acquired; In the detection step, a damage position in the investigation area is detected based on the vibrations and detection positions indicated by the two or more pieces of detection data obtained, and the detection is performed using reference information that indicates a relationship between the vibrations and the detection positions and the damage position. Information processing system.

2. 2. The information processing system according to claim 1, the processor: In the detection step, the vibration and the detection position indicated by the two or more pieces of detection data acquired are input to a detection AI, and the detection is performed by determining the position output by the detection AI as the damage position in the investigation area; The detection AI includes, as the reference information, a detection learning model that has learned the relationship between vibrations detected in the plurality of investigation areas, the detection positions of the vibrations, and the damage positions in the plurality of investigation areas. Information processing system.

3. 3. The information processing system according to claim 2, the processor: In the correct answer acquisition step, correct answer data indicating a damage position in the investigation area is acquired, In the learning step, the detection AI is caused to perform machine learning using the vibrations and detection positions indicated by the two or more pieces of detection data acquired in the investigation area as inputs and the damage position indicated by the acquired correct answer data as output, thereby generating the detection learning model. Information processing system.

4. 2. The information processing system according to claim 1, the reference information represents a relationship between the vibration and the detection position, and the breakage position, as well as factors that affect the vibration detected due to the breakage of the pipe; the processor: In the element acquisition step, element information indicating the element related to the pipe to be inspected for damage is acquired; In the detecting step, the detection is performed based on the vibration and the detection position indicated by the two or more pieces of detection data obtained, as well as the element indicated by the obtained element information. Information processing system.

5. 5. The information processing system according to claim 4, The element information includes information indicating the depth at which the pipe is buried. Information processing system.

6. 2. The information processing system according to claim 1, the processor: In the time period identification step, a time period in which vibration was detected by the vibration surveyor is identified, In the detecting step, the detection is performed using the reference information that differs depending on the identified time period. Information processing system.

7. 2. The information processing system according to claim 1, the processor: In the area identification step, an area where vibration is detected by the vibration surveyor is identified, In the detecting step, the detection is performed using the reference information that differs depending on the identified region. Information processing system.

8. 2. The information processing system according to claim 1, the processor: In the preprocessing step, a preprocessing is performed to reduce noise from the vibrations indicated by the acquired detection data; In the detecting step, the detection is performed based on the vibration that has been subjected to the preprocessing. Information processing system.

9. 9. The information processing system according to claim 8, the processor: In the time period identification step, a time period in which vibration was detected by the vibration surveyor is identified, In the pre-processing step, different processing is performed as the pre-processing depending on the identified time period. Information processing system.

10. 9. The information processing system according to claim 8, the processor: In the area identification step, an area where vibration is detected by the vibration surveyor is identified, In the pre-processing step, different processing is performed depending on the identified region as the pre-processing. Information processing system.

11. An information processing method, comprising: The processor of the information processing system Executing each step of the information processing system according to any one of claims 1 to 10. Information processing methods.

12. A program, A computer is caused to execute each step of the information processing system according to any one of claims 1 to 10. program.

Images