Information processing device and information processing method

Abstract

To provide an information processing device and an information processing method capable of predicting occurrence of a claim before occurrence of a claim regarding a trouble of a surveying device in maintenance management of the surveying device.SOLUTION: An information processing device (100) generates, from a collection data (141) collected with respect to a plurality of surveying devices and containing various pieces of data regarding respective surveying devices, learning data with claim information containing contents of claims received from users and claim receiving time points, and operation state data containing error logs and device logs of the surveying device related to the claim in a prescribed period before occurrence of the claim as a set, executes machine learning by using the learning data, and generates a learning model for predicting contents and time of a claim expected to occur to an object survey device in the future when operation state data of the object surveying device is input.SELECTED DRAWING: Figure 3

Description

[Technical field]

[0001] The present invention relates to an information processing device and an information processing method, and more particularly to an information processing device and an information processing method for maintenance management of a surveying instrument. [Background technology]

[0002] Surveying instruments such as total stations and ground-mounted scanners are highly precise instruments made up of mechanical, optical and electrical components. For this reason, maintenance must be performed periodically or as needed to maintain the required measurement accuracy. For this reason, Patent Document 1 discloses a surveying instrument that monitors the operating time and occurrence of errors of the surveying instrument, and notifies the user in advance of the time when maintenance is required. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2009-139386 A Summary of the Invention [Problem to be solved by the invention]

[0004] However, if maintenance is not performed properly and a malfunction occurs in the surveying device, it will cause a complaint from the user. In the surveying device of Patent Document 1, the time required for maintenance is set as a period calculated from the operation time of the surveying device. However, the user's complaint behavior may not necessarily coincide with the maintenance time calculated from the operation time. This is because there are errors that lead to complaints, some of which occur only once and are not noticed by the user, some of which immediately lead to complaint behavior, and some of which do not immediately lead to complaint behavior but lead to complaint behavior after repeated multiple times. Also, depending on the operating environment, etc., an error may occur before the calculated maintenance time arrives. For this reason, in order to truly prevent and reduce user complaints, it has been required to predict errors that will cause user complaints in advance and respond in advance.

[0005] The present invention has been made in consideration of the above circumstances, and aims to provide a technique for predicting the occurrence of a complaint regarding a malfunction of a surveying instrument before the complaint occurs, with regard to the maintenance management of the surveying instrument. [Means for solving the problem]

[0006] In order to achieve the above object, an information processing device according to one aspect of the present invention is provided, which is configured to: Complaint information, operation status information, From the collected data, including the content of the complaint received from the user and the time of receipt of the complaint The above Claim information, Of the operational status information, The operation of the surveying equipment related to the claim for a specified period prior to the occurrence of the claim situation data Extract and Generate training data as a set, and situation The data includes an error log and an equipment log, and a learning data generation unit; and a learning model generation unit that performs machine learning using the learning data to generate a learning model that predicts the content and timing of future claims that will occur against the target surveying device when operational status data of the target surveying device is input.

[0007] In the above aspect, it is also preferable that the claim contents are tagged with a wording indicating the claim contents, which is classified according to the claim contents.

[0008] In the above aspect, the learning data generation unit checks the time when the complaint is received, situation Information from the most recent error prior to receipt of the claim of Identifying the time of occurrence and a predetermined period before the occurrence of the error Regarding the surveying instrument according to the claim It is also preferable to set the extraction period for the operational status data.

[0009] In the above aspect, it is also preferable that the operational status data further includes measurement environment data.

[0010] In the above aspect, it is also preferable to further include a re-learning unit that, when new complaint information is input, executes re-learning by using the operational status data corresponding to the new complaint.

[0011] In addition, an information processing device according to another aspect of the present invention includes a data acquisition unit that acquires operational status data of a target surveying instrument; Complaint information and operational status information From the collected data, including the content of complaints received from users and the time of complaint acceptance The above Claim information, Of the operational status information, The operation of the surveying equipment related to the claim for a specified period prior to the occurrence of the claim situation data Extract and A learning model is machine-learned using the learning data generated as a set, and when the operational status data of the target surveying device is input, the content and timing of future complaints that will occur regarding the target surveying device are predicted, and the operational status data of the target surveying device are calculated. situation The data includes an error log and a device log, a complaint prediction unit; and a prediction result of the complaint prediction unit. Offer The system includes a result providing unit that provides the results.

[0012] Further, an information processing method according to yet another aspect of the present invention is an information processing method executed by a computer, comprising: Complaint information, operation status information, From the collected data, including the content of the complaint received from the user and the time of receipt of the complaint The above Claim information, Of the operational status information, The operation of the surveying equipment related to the claim for a specified period prior to the occurrence of the claim situation data Extract and Generate training data as a set, and situation Data includes error logs and device logs, steps ;before A step of performing machine learning using the learning data to generate a learning model that predicts the content and timing of future claims that may occur regarding the target surveying device when operational status data of the target surveying device is input; The above A step of acquiring operational status data of the target surveying device; a step of inputting the operational status data of the target surveying device into the learning model and predicting the content and timing of future complaints that will occur regarding the target surveying device; and Offer The method includes the step of providing a

[0013] In this specification, the term "error" refers to a malfunction of the surveying instrument that does not function properly for surveying. It also includes not only a state in which the surveying instrument does not function, but also a state in which the surveying instrument can be used for surveying but the desired surveying accuracy cannot be obtained.

[0014] In this specification, a "complaint" refers to a "report of an error (defect)" made by a user to a manager such as a manufacturer, a distributor (dealer), or a management company. Effect of the Invention

[0015] According to the information processing device and the information processing method according to the above aspects, in regard to the maintenance management of a surveying instrument, it is possible to predict the occurrence of a complaint regarding a malfunction of the surveying instrument before the complaint actually occurs. [Brief description of the drawings]

[0016] [Figure 1] FIG. 2 is a diagram showing the overall configuration of the system. [Diagram 2] 1 is a diagram illustrating an outline of a processing flow of a system using an information processing device according to an embodiment. [Diagram 3] FIG. 2 is a diagram illustrating learning data and a learning model in the system. [Figure 4] 1 is a block diagram showing an example of a configuration of an information processing device according to an embodiment; [Diagram 5] FIG. 2 is a block diagram showing an example of the configuration of a terminal device used in the above system. [Figure 6] FIG. 2 is a block diagram showing an example of the configuration of a surveying device used in the system. [Figure 7] 13 is a flowchart of a process of the information processing device in a learning phase. [Figure 8] 13 is a flowchart showing detailed processing for generating learning data in a learning phase. [Figure 9] FIG. 13 is a diagram showing an example of complaint information. [Figure 10] FIG. 11 is a diagram illustrating an example of operational status data extracted from collected data. [Figure 11] 1 is a table explaining the types and relationships of errors that occur in a surveying instrument. [Figure 12] 13 is a flowchart of a process of the information processing device in a prediction phase. [Figure 13] FIG. 13 is a diagram showing an example of an output screen of a complaint prediction result. [Figure 14] FIG. 13 is a diagram showing another example of an output screen of a complaint prediction result. [Figure 15] FIG. 13 is a block diagram showing an example of the configuration of an information processing device according to a modified example. [Figure 16] 13 is a flowchart showing a re-learning process of the information processing device according to the modified example. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0017] Hereinafter, preferred embodiments of the information processing device and information processing method of the present disclosure will be described with reference to the drawings, but the present invention is not limited thereto. In addition, in each embodiment, the same components are given the same reference numerals, and duplicated descriptions will be omitted as appropriate.

[0018] (Embodiment) 1. Overview of the information processing system and processing by the information processing system First, an overview of an information processing system (hereinafter also simply referred to as a system) 1 according to the present embodiment will be described with reference to FIGS. 1 to 3. FIG.

[0019] Fig. 1 is a diagram showing a schematic configuration of the system 1. Fig. 2 is a diagram explaining an outline of a processing flow in the system 1. Fig. 3 is a diagram explaining learning data and a learning model.

[0020] As shown in FIG. 1, the system 1 includes at least one information processing device 100, at least one terminal device 10, and at least one surveying device S (S 1~n ) (n is a natural number.) The information processing device 100, the terminal device 10, and the surveying device S are connected to each other wirelessly or by wire so as to be able to communicate with each other via a communication network N. The communication network N is, for example, a LAN (Local Area Network) or a WAN (Wide Area Network) such as the Internet.

[0021] The information processing device 100 is, for example, a management server owned by an administrator M of the surveying device S, such as a manufacturer, an agent (dealer), or a management company of the surveying device S. The terminal device 10 is a terminal device owned by the administrator M.

[0022] The surveying instrument S is a surveying instrument owned or used by a user U. The data provided for processing is data related to the surveying instrument S. The information processing device 100 is configured to process a plurality of surveying instruments S of a plurality of users. 1~nIn the following explanation, when referring to an unspecified surveying instrument, it is referred to as surveying instrument S. When referring to a surveying instrument that is the subject of a claim prediction, it is referred to as subject surveying instrument S. x He said.

[0023] The administrator M provides a web page for managing the surveying device S. Information related to the surveying device S can be viewed and managed on a web page displayed on the display unit of the terminal device 10 from an administrator screen linked to the administrator M. Such a communication management system for the surveying device S can have a known configuration, for example, as disclosed in JP 2019-7903 A and the like.

[0024] As shown in the upper left of FIG. 2, the surveying device S transmits various information including basic information, operating status information, surveying data, etc. to the information processing device 100 at a predetermined timing.

[0025] The "basic information" includes at least identification information of the surveying instrument S (hereinafter referred to as "ID of the surveying instrument S"). It may also include environmental information data (temperature, humidity, etc.), location information, software version information, etc. The ID of the surveying instrument S is information that includes model information, manufacturing lot, individual number, etc. of the surveying instrument.

[0026] "Operation status information" is information that includes an "error log" and an "equipment log." The "error log" is history information such as an error code, the date of occurrence, the time of occurrence, the number of occurrences, the state of the instrument when the error occurred, the horizontal angle measurement value, the vertical angle measurement value, the distance measurement value, etc.

[0027] The "instrument log" is information about the operation of the surveying instrument S, such as the cumulative number of distance measurements, the number of programs started, and the type of programs started, as well as historical information about the instrument's status during operation. For example, the motor rotation speed during execution of the distance measurement program, the output value from each board, etc.

[0028] The predetermined timing is, for example, always, every certain period (one hour, one day, one week, one month, etc.), every certain operation, etc. The predetermined timing may be set, for example, every certain period when there is no special event, or for each event when an event occurs.

[0029] On the other hand, as shown in the upper right of Figure 2, when a problem occurs during use of the surveying device S and a complaint is made by a user U, an operator of the manager M accepts the complaint. Specifically, a complaint is usually accepted by the operator of the manager M via telephone, e-mail, facsimile, posting on a web page, etc.

[0030] As shown in FIG. 3, the "complaint information" includes the "time of receipt of the complaint" and the "content of the complaint". The "content of the complaint" is the symptom of the malfunction of the surveying device S that the user U conveyed to the operator, and includes the user U's subjective evaluation. The "complaint information" is linked to the "ID of the surveying device S" that is the subject of the complaint, and is input from the terminal device 10 to the information processing device 100 via a web page, and is stored in the auxiliary storage device 140 as collected data 141. The time of receipt of the complaint is the time when the complaint is received from the user U. It may be specified to the hour and minute, or may be specified by the day. In addition, the time when the error that caused the complaint occurred, which is understood from the operation status information, may be approximately the time of receipt of the complaint.

[0031] In this way, the information processing device 100 collects various data related to the surveying instrument S, including "operation status information" and "complaint information", as big data, and manages the data in association with the ID of the surveying instrument S. The collected data may not be stored in the information processing device 100, but may be stored in another database server or cloud storage.

[0032] When sufficient data is accumulated, the information processing device 100 generates learning data as shown in FIG. 3 using the accumulated complaint information and operational status information regarding a large number of surveying devices S in step S1. And learn A learning model 144 is generated.

[0033] Specifically, the information processing device 100 generates learning data by extracting "complaint information" and "operational status data" from the collected data according to a process described below.

[0034] Here, the "operational status data" is In the learning phase, which will be described later, Among the operational status information, the specified period before the error that caused the complaint occurred of This predetermined period may be a fixed period such as three months or one year. Alternatively, it may be the period from the previous maintenance to the time the error occurs. Furthermore, if the maintenance period has not yet elapsed after the first operation, it may be the period from the first operation to the time the error occurs. In addition, in the prediction phase described later, the "operational status data" is data for a predetermined period prior to the present, out of the operational status information, for the surveying instrument to be predicted. In other words, the "operational status data" is data for a predetermined period out of the operational status information.

[0035] Then, machine learning is performed using the learning data thus acquired to generate a learning model 144. This learning model 144 is a model that, when the operational status data of the target surveying device Sx is input, x This is a trained model that predicts the occurrence of claims regarding

[0036] Returning to FIG. 2, in step S2, the information processing device 100 adds the target surveying device S to the generated learning model 144. x The operation status data for a specified period up to the present time is input, and future complaints (contents and timing) are output.

[0037] Then, the information processing device 100 provides the estimation result to the administrator M in step S3.

[0038] 2. Detailed configuration of information processing system 1 2.1 Information processing device 100 The detailed configuration of the information processing system 1 will be described below. Fig. 4 is a diagram showing an example of the configuration of the information processing device 100. The information processing device 100 is a so-called server computer. The information processing device 100 includes a communication unit 110, a control unit 120, a main storage device 130, and an auxiliary storage device 140.

[0039] The communication unit 110 is a communication control device such as a network adapter, a network interface card, or a LAN card, and connects the information processing device 100 to the communication network N by wire or wirelessly. The control unit 120 can transmit and receive various information to and from the surveying device S and the terminal device 10 via the communication unit 110 and the communication network N.

[0040] The control unit 120 is configured with one or more CPUs (Central Processing Units), a multi-core CPU, a GPU (Graphics Processing Unit), etc. The control unit 120 is connected to each hardware unit constituting the information processing device 100 via a bus.

[0041] The control unit 120 includes, as functional units, a learning data generation unit 121, a learning model generation unit 122, an operating status data acquisition unit 123, a complaint occurrence prediction unit 124, and a result provision unit 125.

[0042] The learning data generation unit 121 extracts from the collected data 141 a set of corresponding "claim information" linked to the same surveying device ID and "operational status data" for a predetermined period before the occurrence of the error that caused the claim, performs a predetermined process, generates learning data, and stores it in the auxiliary storage device 140 as a learning data DB 143. The process of generating learning data will be described later.

[0043] The learning model generation unit 122 performs machine learning using the "claim information" and the "operation status data" as a learning data set. Specifically, the learning model generation unit 122 ,workmanThe learning model generation unit 122 performs machine learning by qualitatively evaluating (probability distribution-based) changes in the operation status data for a predetermined period before the occurrence of the error. The learning model generation unit 122 further performs machine learning to predict the time and content of the complaint, taking into account the time difference between the time the error occurs and the time the complaint is received, to generate a learning model 144. The generated learning model 144 is stored in the auxiliary storage device 140. The generated learning model 144 is a learning model that predicts the content of the complaint that will occur and the time when the complaint will occur when the operation status data of the target surveying device Sx for a predetermined period before the current time is input. In other words, the output of the learning model 144 is a prediction of what kind of complaint will occur, when, and with what probability. The generation of the learning model 144 is realized, for example, by a neural network that uses one or more layers of nonlinear units to predict the output for the input.

[0044] Specifically, the machine learning can be performed by any method such as logistic regression, SVM (Support Vector Machine), Random Forest, CNN (Convolutional Neural Network), RNN (Recurrent Neural Network), XGBoost (eXtreme Gradient Boosting), etc. Furthermore, the machine learning can be supervised learning, semi-supervised machine learning, or unsupervised learning.

[0045] RNN is a method capable of learning time-series data. When time-series data is used as the operational status data, it is preferable to use the RNN method, such as MTRNN (Multi Timescale RNN) or LSTM (Long Short Term Memory).

[0046] The operation status data acquisition unit 123 acquires the operation status data of the target surveying device S x Obtain operational status data for

[0047] The complaint occurrence prediction unit 124 inputs the target surveying device S acquired by the operation status data acquisition unit 123 into the learning model 144 stored in the auxiliary storage device 140. x Enter the operation status data of the target surveying device S x Regarding the above, the content of claims that are expected to arise and the timing of such claims are predicted.

[0048] The result providing unit 125 provides the prediction result to the administrator. The prediction result may be displayed as a Web page on the display unit 12 of the terminal device 10 of the administrator M. Alternatively, the prediction result may be provided in the form of being sent to the terminal device 10 of the administrator M by e-mail.

[0049] The main memory device 130 is a storage device such as a static random access memory (SRAM), a dynamic random access memory (DRAM), a flash memory, etc. The main memory device 130 temporarily stores information required for processing performed by the control unit 120 and programs being executed by the control unit 120.

[0050] The auxiliary storage device 140 is a storage device such as an SRAM, a flash memory, or an HDD (Hard Disc Drive). The auxiliary storage device 140 stores collected data 141, a learning data DB 143, and a learning model 144. In addition, a program for executing the functions of each functional unit of the control unit 120 and various data required for executing the program are stored. The collected data 141, the learning data DB 143, and the learning model 144 may be stored in an external large-capacity storage device connected to the information processing device 100, or in a data server (not shown), etc.

[0051] 2.2 Terminal device 10 The terminal device 10 is realized by a desktop PC, a notebook PC, a tablet terminal, a mobile phone, a PDA (Personal Digital Assistant), or the like. The terminal device 10 has a plurality of applications, such as a Web browser application. An operator of the administrator M can input information and instructions to the information processing device 100 from the terminal device 10 via a Web page. The prediction result by the information processing device 100 is displayed on the Web page.

[0052] 5 shows an example of the configuration of the terminal device 10. The terminal device 10 includes a communication unit 11, a display unit 12, an input unit 13, a main storage device 14, a control unit 15, and an auxiliary storage device 16, and these units are connected via a bus.

[0053] The communication unit 11 is a communication control device such as a network adapter, a network interface card, or a LAN card. The terminal device 10 is connected to the communication network N by wire or wirelessly. The control unit 15 can transmit and receive various information to and from the information processing device 100 via the communication unit 11 and the communication network N.

[0054] The display unit 12 is an organic EL display or a liquid crystal display. Under the control of the control unit 15, the display unit 12 displays various information on a Web page.

[0055] The input unit 13 is a keyboard including letter keys, numeric keys, an enter key, etc., a mouse, a power button, etc. An operator can input various information via the input unit 13. The display unit 12 and the input unit 13 may be integrated into one unit as a touch panel display.

[0056] The main memory device 14 is a storage device such as an SRAM, a DRAM, or a flash memory. The main memory device 14 temporarily stores information required for the processing performed by the control unit 15 and programs being executed by the control unit 15. 16 SRAM, flash memory 、A storage device such as an HDD, etc., which stores in advance the control program executed by the control unit 15 and various data required for executing the control program.

[0057] The control unit 15 includes, for example, a microcomputer having a CPU, a ROM, a RAM, an input / output port, and various other circuits. The control unit 15 reads out and executes various programs stored in the main storage device 14 and the RAM. The control unit 15 includes a result acquisition unit 51.

[0058] 2.3 Surveying equipment S In this example, the surveying device S is a total station. As shown in FIG. 6, the surveying device S includes a surveying unit 21, a rotary drive unit 22, a control unit 23, a communication unit 24, a display unit 25, and an environmental sensor 26.

[0059] The surveying unit 21 includes a distance measuring unit that transmits distance measuring light and receives reflected light to measure the distance to a target, and an angle measuring unit that detects the collimation angle of the distance measuring light.

[0060] The rotation drive unit 22 is a motor, and includes one that rotates the collimating telescope vertically and one that rotates the housing horizontally.

[0061] The control unit 23 is a control unit having at least a CPU and a memory (ROM, RAM, etc.). The control unit 23 executes a survey application program and realizes the functions of the surveying device S. The control unit 23 performs an operation at a predetermined interval. situation The information is linked to the ID of the surveying device S and transmitted to the information processing device 100.

[0062] The communication unit 24 is a communication control device similar to the communication unit 110 of the information processing device 100 .

[0063] The display unit 25 has a liquid crystal screen, on which survey conditions and the like can be input, and on which various information relating to the survey is displayed.

[0064] The environmental sensor 26 is a sensor for acquiring measured environmental data, such as a temperature sensor, a humidity sensor, or the like.

[0065] Surveying equipment S is not limited to total stations, situation A control unit that links information with the ID of the surveying device S and transmits it to the information processing device 100, and a control unit that controls the operation of the measurement environment data, etc. situation As long as it is equipped with a sensor or the like for acquiring the data contained in the information, it may be a surveying device such as a 3D laser scanner or a theodolite.

[0066] 3. Information Processing Next, the details of information processing by this system will be described. The information processing includes the use of a learning model 144 by machine learning. Generate and a prediction phase in which the occurrence of a complaint is predicted based on the learning model 144. The processing of the information processing device 100 in each phase will be described below.

[0067] 3.1 Learning Phase FIG. 7 is a flowchart of the processing of the information processing device 100 in the learning phase. When the process starts, in step S01, the learning data generation unit 121 generates a set of "claim information" and "operating status data" related to the same surveying instrument from the collected data 141 as learning data, and stores the set in the learning data DB 143. The details of step S01 will be described later.

[0068] Next, in step S02, the learning model generation unit 122 performs machine learning on the set of "claim information" and "operation status data" stored in the learning data DB 143 as learning data, and generates a model for the target surveying device S X When the current operating status data of the target surveying device S is entered, X Then, the learning model generation unit 122 stores the generated learning model 144 in the auxiliary storage device 140, and the process ends.

[0069] Details of step S01 will now be described with reference to Fig. 8. Fig. 8 is a flowchart showing details of step S01, that is, processing related to pre-processing of learning data.

[0070] When step S01 starts, in step S11, the learning data generation unit 121 selects one complaint from the collected data 141. Fig. 9(A) shows an example of complaint information accumulated as the collected data 141. In addition to the "complaint content", the complaint information includes the ID of the surveying instrument S that is the subject of the complaint and the "time of receipt of the complaint".

[0071] Next, in step S12, the learning data generating unit 121 tags each type of symptom reported by the user U based on the text described in the complaint, and quantifies the tags to create features. The tagging rules may be, for example, a definition of the contents of the text and the tags in a table format. For example, the tagging rules are set such that "the screen sometimes goes blank only on the other side" is tagged as "contact failure / panel failure", "the motor stops by itself while rotating" is tagged as "rotation control abnormality", and "the motor rotation does not stop at the specified angle" is tagged as "rotation control abnormality". Note that these tagging rules are based solely on the reports of the user U's physical experience, and do not necessarily match the classification by error code.

[0072] FIG. 9(B) is an example of tagged claim information. Such tagging may be performed by an operator when accepting a claim, or may be achieved by a known text annotation tool. Alternatively, it may be achieved by a known text mining technique. In addition, it may be achieved by using a learning model generated by supervised machine learning, in which the combination of the content of each claim and the tag is used as learning data.

[0073] Next, in step S13, the learning data generation unit 121 checks the ID of the surveying instrument S to which the complaint information extracted in step S11 is linked and the time of receipt of the complaint.

[0074] Next, in step S14, the learning data generation unit 121 acquires, from the collected data 141, operation status information about the ID of the surveying instrument S acquired in step S13.

[0075] Next, in step S15, the learning data generation unit 121 identifies, from the operational status information of the surveying device S, the time at which the most recent error occurred going back to the time the complaint was received in step S14, as the cause of the event that caused the complaint.

[0076] Next, in step S16, data for a predetermined period going back from the time when the error occurred is extracted as operation status data. situation This is an example of the data. situation The data includes the period (2020 / 12 / 15~2021 / 1 / 15) for which the operation status data was extracted, determined from the date of receipt of the complaint (2021 / 1 / 17) to the time of the most recent error (2021 / 01 / 15 11:23), as well as the error log for the extraction period, as chronological data. situation The data does not necessarily have to be time-series data, and may be converted into data that allows for evaluation of fluctuations within the specified period in the form of, for example, cumulative values, average values, standard deviations, or the like.

[0077] Next, in step S17, the learning data generation unit 121 stores the tagged complaint content in step S12 and the operational status data extracted in step S15 in the learning data DB 143 of the auxiliary storage device 140 as one set of learning data.

[0078] Next, in step S18, the learning data generating unit 121 determines whether processing for all necessary complaints stored in the collected data 141 has been completed.

[0079] If all claims have been processed (Yes), step S01 ends and the process proceeds to step S02. If not (No), the process returns to step S11 and steps S11 to S18 are repeated for the next claim until all claims have been processed.

[0080] In this way, if the wording of the complaint content is tagged, even if the received complaints have various expressions, the complaint content can be easily classified, and the learning accuracy is improved. In particular, by having the learning data generating unit 141 automatically tag the complaints, it is possible to eliminate the variation between operators and human error and perform uniform classification.

[0081] Incidentally, there are two types of errors: serious errors, such as error code Y in Figure 11, which can cause the surveying device S to stop functioning with just one error and lead to a complaint; and minor errors, such as error code 01, which are not serious enough to cause a complaint individually, but which can cause a serious error like error code X if they occur multiple times. For example, as in the example in Figure 11, a minor error such as "measurement error due to oil leakage" is repeatedly generated, causing a major error called "unit failure." Minor errors are not detected by warnings such as error codes, and even if there is a warning, the use of the surveying device can be continued by canceling the warning state.

[0082] According to the information processing device of this embodiment, the time of occurrence of the error that caused the complaint is specified, and the operation status data (instrument log and error log) for a predetermined period before the occurrence is used as learning data. Therefore, it is possible to detect minor errors in the target surveying device S before the occurrence of a serious error. x If this occurs, the target survey device S x From the pattern of occurrence of minor errors, it becomes possible to predict the occurrence of serious errors. As a result, it is possible to quickly and accurately provide the user with a method for eliminating minor errors. Furthermore, it is possible to prevent the occurrence of serious errors.

[0083] In addition, even in the case of a serious error that could lead to a complaint if it occurs just once, the information processing device 100 can predict when a complaint will occur by taking into account the device operating status before the error occurred, which cannot be determined by the error code alone, because the device log and error log are used as learning data.

[0084] In the former case, the number of times a minor error has occurred since the last maintenance is also an important factor in predicting when a complaint will occur, so it is preferable to set the specified period for extracting operation information data to the period from the last maintenance or the start of initial operation to the time the error occurred.

[0085] Furthermore, in this embodiment, since the "complaint content," which is the symptom of the malfunction that the user U conveys to the operator, is used as the learning data, the content of the complaint that will occur (i.e., the error that will occur) can be accurately predicted. This is because the symptoms and solutions described in the instruction manual, etc. are theoretical, and may not necessarily match the symptoms that occur during actual measurement. By including the complaint content that corresponds to the symptoms that the user intuitively grasps in the learning data, it becomes possible to predict the complaint content in accordance with the actual operating situation.

[0086] 3.2 Prediction Phase 12 is a flowchart of the process of the information processing device 100 in the prediction phase. x is set in advance. Target survey device S x The setting may be set so as to be executed sequentially for all surveying instruments S under the management of the administrator M. Alternatively, the setting may be set at a predetermined interval for a part of the surveying instruments S. Alternatively, the setting may be set so as to be executed sequentially for all surveying instruments S under the management of the administrator M. Alternatively, the setting may be set so as to be executed sequentially for a part of the surveying instruments S at a predetermined interval. Alternatively, the setting may be set so as to be executed sequentially for all surveying instruments S under the management of the administrator M. x Alternatively, the setting may be made by individually specifying the above.

[0087] When the process starts, in step S21, the operation status data acquisition unit 123 acquires the operation status data of the surveying instrument S x For example, the operation status data for the target surveying device S within a predetermined period before the current time is acquired from the collected data 141. x If the ID of the surveying device S is TS1234 and the current date is February 20, 2021, and the specified period is January, the operation status data acquisition unit 123 acquires operation status data from January 20, 2021 to February 20, 2021 for the surveying device S with ID TS1234 from the operation status information DB 142. do.

[0088] Next, in step S22, the complaint prediction unit 124 inputs the operational status data acquired in step S21 into the learning model 144 to predict complaints that may occur in the near future regarding the target surveying instrument S (e.g., ID: TS1234).

[0089] Next, in step S23, the result providing unit 125 provides the output result, and the process ends.

[0090] As mentioned above, the output is, for example, the probability of the content of a complaint based on the operational status data of the target surveying device S, and the time when the complaint is expected to occur. FIG. 13 shows an example of the output result displayed on the display unit 12, and a complaint prediction result output screen. The complaint prediction result output screen displays the type of complaint that will occur, when, and with what probability. This display enables the administrator M to carry out maintenance on the user U before the user U takes action to make a complaint.

[0091] However, since the manager M manages a large number of surveying instruments, it is difficult for him to constantly monitor the complaint prediction result output screen for each instrument. In such a case, a push notification may be displayed on the display unit 12 of the terminal device 10 as shown in Fig. 14, to alert the user that there is a surveying instrument that is predicted to cause a complaint soon.

[0092] The display of the output results output to the terminal device 10 is controlled by the result acquisition unit 51 of the terminal device 10. As described above, the output results may be provided by e-mail or other means.

[0093] In this way, in the information processing device 100 according to the above embodiment, complaint information and the operating status of the surveying device S are stored in a database, and the learning model 144 that has been machine-learned is configured to predict future occurrence of user complaints based on the current operating status of the surveying device S, so that an administrator who sees the output can prevent complaints from users in advance and enables an agency or the like to take appropriate action against the user. Also, for the user, the administrator can respond appropriately before the surveying device S malfunctions to the extent that a complaint must be made, improving customer satisfaction.

[0094] In this embodiment, the operating status of the surveying device S at the time of an actual complaint is taken into consideration, so there is no need to leave any leeway in terms of timing, etc., and it enables agencies, etc. to respond appropriately to users.

[0095] In this embodiment, the learning phase and the prediction phase are configured to be realized by the same information processing device 100. However, the learning phase and the prediction phase may be realized by different information processing devices 100. In this case, the information processing device that executes the learning phase and the prediction phase includes a learning unit 120A or a prediction unit 120B in the control unit 120, respectively.

[0096] In this embodiment, the prediction result is notified to the manager M, but the present invention is not limited to this and may be provided to the user.

[0097] 4. Variations 15 is a configuration block diagram of an information processing device 100a according to a modification. The information processing device 100a has roughly the same configuration as the information processing device 100, but differs in that a re-learning unit 126 is further provided in a control unit 120a.

[0098] The re-learning unit 126 of the information processing device 100a executes the process of the flowchart in FIG.

[0099] That is, in step S31, it is determined whether new complaint information is to be added to collected data 141. If it is added (Yes), in step S32, learning data is generated, and in step S33, re-learning of learning model 144 is performed. In this way, re-learning is performed every time new data is input, and the accuracy of prediction can be improved.

[0100] As another variation, the learning data may be “claim information,” “operation status,” data In addition to the "measurement environment data," the learning data may be configured to include "measurement environment data" such as temperature and humidity. The occurrence of malfunctions is often related to the environment during use, and by taking the environment into consideration, a more accurate prediction of a method of resolving the malfunction can be made. Therefore, if the learning data includes measurement environment information, a more accurate prediction of a method of resolving the error can be made.

[0101] In another modified example, the surveying device is equipped with a GNSS device that acquires position information, and the learning data is "claim information," "operation status data " may be configured to include "location information" in addition to the "location information." As a result, the estimated results obtained using the learning model can be made to be results for each region or country. This is because, depending on the region or national character, there may be cases where many users immediately complain to the administrator when an error occurs, and cases where many users are slow to complain to the administrator even when an error occurs. In this way, by including location information in the learning data, it becomes possible to predict when a complaint will occur depending on the region or national character of the user.

[0102] The above describes preferred embodiments of the present invention. However, the above embodiments are merely examples of the present invention, and these can be combined based on the knowledge of those skilled in the art, and such combinations are also included in the scope of the present invention. [Explanation of symbols]

[0103] 100: Information processing device 100a: Information processing device 120A: Learning Department 120B: Prediction section 121: Learning data generation unit 122: Learning model generation unit 123: Operation status data acquisition unit 124: Complaint Prediction Department 125:Result provision department 126: Re-learning section 141 :Collected data 143 :Learning data DB 144: Learning model

Claims

1. a learning data generating unit which extracts, from collected data including complaint information and operational status information about each of a plurality of surveying instruments, the complaint information including the content of the complaint received from a user and the time of receipt of the complaint, and from the operational status information, operational status data for the surveying instrument related to the complaint for a predetermined period before the occurrence of the complaint, and generates learning data as a set, the operational status data including an error log and an instrument log; and An information processing device characterized by having a learning model generation unit that performs machine learning using the learning data and generates a learning model that predicts the content and timing of future complaints that will occur regarding the target surveying device when operating status data of the target surveying device is input.

2. 2 . The information processing apparatus according to claim 1 , wherein the complaint contents are tagged with a wording indicating the complaint contents, which is classified according to the complaint contents.

3. The information processing device described in claim 1 or 2, characterized in that the learning data generation unit checks the time when the complaint is received, identifies the time when the most recent error occurred before the complaint is received from the operation status information, and sets a predetermined period before the error occurred as the extraction period for the operation status data for the surveying device related to the complaint.

4. 4. The information processing apparatus according to claim 1, wherein the operational status data further includes measurement environment data.

5. An information processing device according to any one of claims 1 to 4, further comprising a re-learning unit that, when new complaint information is input, performs re-learning using the operational status data corresponding to the new complaint.

6. a data acquisition unit for acquiring operation status data of the target surveying device; a complaint occurrence prediction unit which extracts complaint information including the complaint content of the complaint received from a user and the time of complaint receipt, and from the operational status information, operational status data for the surveying instrument related to the complaint for a predetermined period before the complaint occurred, from collected data including complaint information and operational status information for each surveying instrument collected for a plurality of surveying instruments, and uses a learning model trained by machine learning using the learning data generated as a set to predict the content and timing of a complaint that will occur in the future for the target surveying instrument when the operational status data of the target surveying instrument is input, the operational status data including an error log and an instrument log; and The information processing device further comprises a result providing unit that provides a prediction result of the complaint occurrence prediction unit.

7. 1. A computer-implemented information processing method, comprising: A step of extracting, from collected data including complaint information and operational status information about each of a plurality of surveying instruments, the complaint information including the content of the complaint received from a user and the time of receipt of the complaint, and operational status data for the surveying instrument related to the complaint for a predetermined period before the complaint occurred from the operational status information, and generating learning data as a set, the operational status data including an error log and an instrument log; A step of performing machine learning using the learning data to generate a learning model that predicts the content and timing of future claims that may occur regarding the target surveying device when operational status data of the target surveying device is input; acquiring operational status data of the target surveying device; A step of inputting the operational status data of the target surveying device into the learning model to predict the content and timing of future claims that may arise regarding the target surveying device; and 11. A method comprising providing a predicted outcome.

Images