Patent Application
20220307831
The present invention relates to an information processing device and an information processing method, and more specifically, to an information processing device and an information processing method for recognizing a user's instrument mismatch.
As for surveying instruments, there are instrument types such as levels, theodolites, total stations, and 3D scanners, . . . etc. In general terms, a level is an instrument suitable for horizontal direction measuring and height difference measuring between object points, a theodolite is an instrument suitable for performing angle measuring of a horizontal angle and an elevation angle of an object point, a total station is an instrument suitable for measuring three-dimensional coordinates of a prism or an object point other than a prism by distance and angle measuring, and a scanner is an instrument suitable for measuring three-dimensional coordinates of a plurality of object points, and in recent years, there are sophisticated instruments including total stations with application functions such as a piling support function and an area calculating function, and a scanner with application functions such as a backward intersection support function (refer to, for example, Patent Literature 1 for the total station).
Patent Literature 1: Japanese Published Unexamined Patent Application No. 2012-202821
As described above, there are a large number of types of surveying instruments, and as surveying instruments of the same type, there are a large number of products with different specifications and functions. Therefore, in actuality, there are cases where, as compared with a purpose of use and use situation of a user, it is found that the user uses an excessively high-spec instrument, or conversely, uses a low-spec instrument, or a different instrument type is more suitable, etc. When such an instrument mismatch occurs, it is likely that a user has paid too much in user fees or performs inefficient work.
The present invention was made to solve the problem described above, and an object thereof is to recognize an instrument mismatch with respect to a use situation of a user for a surveying instrument.
In order to solve the problem described above, an information processing device according to an aspect of the present invention includes an input information creating unit configured to collect information stored in each surveying instrument from a plurality of surveying instruments, and create learning data by associating surveying instrument information, information on measuring function used, and information on measuring amount used; and a learning model generating unit configured to execute machine learning by using the learning data, and when information on object measuring function and information on object measuring amount used in an object surveying instrument owned or managed by a user are input, generate a learning model for estimating a suitable surveying instrument with respect to the information on object measuring function and the information on object measuring amount.
In the aspect described above, it is also preferable that the input information creating unit extracts, as the information on measuring function, at least functions of distance measuring, angle measuring, prism distance measuring, horizontal direction measuring, height difference measuring, and various kinds of application surveys.
In the aspect described above, it is also preferable that the input information creating unit extracts, as the information on measuring amount, at least the number of measurements, the number of measurement points, a measuring range, and an operating time.
In the aspect described above, it is also preferable that the input information creating unit extracts, as the surveying instrument information, at least a model number type of each of the surveying instruments.
In addition, in order to solve the problem described above, an information processing device according to an aspect of the present invention includes an object information acquiring unit configured to acquire information stored in an object surveying instrument owned or managed by a user as object data; an object information extracting unit configured to extract information on object measuring function and information on object measuring amount used in the object surveying instrument from the object data; an estimating unit configured to execute machine learning by using surveying instrument information, information on measuring function, and information on measuring amount as learning data from collected data collected by information stored in each surveying instrument from a plurality of surveying instruments, and when the information on object measuring function and the information on object measuring amount are input, estimate a suitable surveying instrument for an actual use situation of the user by inputting the information on object measuring function and the information on object measuring amount into a learning model for estimating a suitable surveying instrument with respect to the information on object measuring function and the information on object measuring amount; and a result providing unit configured to provide estimation results by the estimation unit to the user.
In the aspect described above, it is also preferable that the object information acquiring unit acquires the object data in a set totaling period, the estimating unit performs estimation at intervals of the totaling period, and the result providing unit provides the estimation results to the user at intervals of the totaling period.
In the aspect described above, it is also preferable that the result providing unit displays surveying instruments based on the estimation results in descending order of score on a terminal device of the user, and proposes a replacement purchase or an additional purchase of a surveying instrument.
In the aspect described above, it is also preferable that, for each customer of the user, the result providing unit displays surveying instruments based on the estimation results in descending order of score on a terminal device of the user, and proposes a replacement purchase or an additional purchase of a surveying instrument.
In addition, in order to solve the problem described above, an information processing method according to an aspect of the present invention is an information processing method to be executed by a computer, and includes a step of collecting information stored in each surveying instrument as collected data from a plurality of surveying instruments; a step of extracting surveying instrument information, information on measuring function, and information on measuring amount, from the collected data, and creating a set of the surveying instrument information, the information on measuring function, and the information on measuring amount as learning data; a step of executing machine learning by using the learning data, and when information on object measuring function and information on object measuring amount used in an object surveying instrument owned or managed by a user are input, generating a learning model for estimating a suitable surveying instrument with respect to the information on object measuring function and the information on object measuring amount; a step of acquiring information stored in the object surveying instrument as object data; a step of estimating a suitable surveying instrument for an actual use situation of the user by inputting the information on object measuring function and the information on object measuring amount into the learning model; and a step of providing estimation results to the user.
According to the present invention, a technique to recognize an instrument mismatch with respect to a use situation of a user for a surveying instrument can be provided.
Next, a preferred embodiment of the present invention will be described with reference to the drawings.
1. Outline of Information Processing
First, an outline of information processing according to the present embodiment will be described.
An information processing device 100 illustrated in
The information processing device 100 collects data from the surveying instruments M1, M2, . . . MN through the communication network N. As illustrated in
As illustrated in
As illustrated in
Then, according to the output data (estimation results), the information processing device 100 proposes an additional purchase or replacement purchase of a surveying instrument to the user C. These are the outline of information processing to be performed in the present embodiment. Hereinafter, the information processing will be described in detail in a divided manner into a learning phase and an estimation phase.
2. Information Processing in Learning Phase
2-1. Configuration of Information Processing Device
A detailed configuration of the information processing device 100 in a learning phase will be described.
The communication unit 101 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 103 transmits and receives various information to and from the surveying instruments M1, M2, . . . MN (
The main storage device 102A is a semiconductor memory device such as a RAM (Random Access Memory) or a flash memory, or a storage medium such as an HDD (Hard Disc Drive) or an optical disc.
In the main storage device 102A, “collected data 120” collected from the plurality of surveying instruments M1, M2, . . . MN is stored. The collected data 120 includes various data such as, in addition to the “surveying instrument information,” “information on measuring function,” and “information on measuring amount” described later, measurement data, image data, audio data, environmental data, error logs, machine logs of components, and data on a maintenance period and a rental period. Each time new data is accepted, the main storage device 102A associates collected data with an identification ID provided for an individual number of each surveying instrument so that the data can be identified by measurement or date. The collected data is also used for the purpose of measurement data analysis and error analysis, etc., other than in the present embodiment. The collected data 120 may be stored not in the main storage device 102A but in a server or a cloud storage different from the information processing device 100.
For extraction of the “surveying instrument information” described later, the main storage device 102A stores an instrument identification table 123 for identifying an “instrument type” of a surveying instrument such as a level/a theodolite/a total station/a scanner, etc., and a “model number type (model number)” of surveying instruments of the same type based on individual numbers of the surveying instruments. The instrument identification table 123 may also be stored not in the main storage device 102A but in a server or a cloud storage different from the information processing device 100.
The auxiliary storage device 102B is a storage medium such as an SRAM, a flash memory, or an HDD. In the auxiliary storage device 102B, the learning model 121 and a learning data DB 122 are stored. These may also be stored not in the auxiliary storage device 102B but in a server or a cloud storage different from the information processing device 100.
The learning data DB 122 stores a plurality of learning dataset created by an input information creating unit 131 described later. The learning model 121 is generated by a learning model generating unit 132 described later, and functions as a classifier made as a result of machine learning. This will be described in detail in the description of the learning model generating unit 132.
The control unit 103 consists of one or a plurality of CPUs (Central Processing Units), multicore CPUs, or GPUs (Graphics Processing Units), etc. The control unit 103 is connected to respective hardware units constituting the information processing device 100 through a bus.
The control unit 103 includes, as functional units, the input information creating unit 131, the learning model generating unit 132, an object information acquiring unit 135, an object information extracting unit 136, an estimating unit 137, and a result providing unit 138. Among these, the input information creating unit 131 and the learning model generating unit 132 function in a learning phase. The remaining functional units 135, 136, 137, and 138 function in an estimation phase.
Functions of the respective units are realized by, for example, reading and executing programs stored in the ROM or the main storage device 102A by the CPU. Part of the respective units may consist of hardware such as ASIC (Application Specific Integrated Circuit) or FPGA (Field-Programmable Gate Array).
The input information creating unit 131 extracts “surveying instrument information,” “information on measuring function,” and “information on measuring amount” from the collected data 120 collected in the main storage device 102A with respect to each surveying instrument.
The input information creating unit 131 extracts an “instrument type” and a “model number type” of each surveying instrument as “surveying instrument information.” When the “instrument type” and “model number type” are not included in the collected data 120, based on an individual number of the surveying instrument, at least the “model number type” is extracted by referring to the instrument identification table 123.
The input information creating unit 131 extracts, as “information on measuring function,” a function used among at least a distance-measuring function, an angle-measuring function, a prism distance-measuring function, a horizontal direction measuring function, a height difference measuring function, and application functions. More specifically, as for application functions, for example, in the case of a total station, a function used is extracted among radiation observation, coordinate observation, pair of observations, piling support, opposite side measurement, traverse calculation, area calculation, and topographical survey, etc.
The input information creating unit 131 extracts, as “information on measuring amount,” amounts used among at least the number of measurements, the number of measurement points, a measurement range, and an operating time, in the form of “numerical values.”
The input information creating unit 131 associates the sets of these “surveying instrument information,” “information on measuring function,” and “information on measuring amount” with an identification ID of each surveying instrument, and stores these as “learning data” in the learning data DB 122. The input information creating unit 131 performs this creating work for the collected data 120 at predetermined timings, that is, each time new data is accepted, or at regular time intervals (hourly, every several hours, daily, etc.).
A detailed example of the learning data creation by the input information creating unit 131 is described.
The learning model generating unit 132 reads learning data from the learning data DB 122 and executes machine learning to generate the learning model 121. The learning model 121 is realized by a neural network using one or a plurality of layers of nonlinear units for predicting an output responding to an input. The learning model 121 generated by the learning model generating unit 132 is stored in the auxiliary storage device 102B.
As an example, the learning model generating unit 132 uses teacherless learning such as clustering or a known statistical procedure so that samples are grouped by “model number type” of the surveying instruments, and generates the learning model 121 for grasping, as characteristics of the respective samples included in a certain surveying instrument (group of a certain model number type), a “general use model” including a measuring function generally used in this surveying instrument and a measuring amount of the measuring function. The “general use model” is created to have content in which, for example, in a total station (model No. C-1000), 100 prism measurements and 100 non-prism measurements are performed on monthly average.
When input data of the user C is input into this learning model 121, by comparison with the “general use model” of each surveying instrument (group of a certain model number type), based on similarity or statistical numerical values such as averages, medians, modes, accumulated values, and standard deviation, or based on shape matching with a shape obtained by defining the general use model as a graphic, or based on a combination of these, one or some surveying instruments having use patterns similar to the input data of the user C are output.
The learning model generating unit 132 may perform the above-described processing for each “instrument type” of surveying instruments. However, the above-described processing is just an example of the learning model generating unit 132, and the learning model generating unit 132 may generate the learning model 121 by using other methods of teacherless machine learning such as a principal component analysis.
The object information acquiring unit 135, the object information extracting unit 136, the estimating unit 137, and the result providing unit 138 will be described in an estimation phase.
2-2. Information Processing Method in Learning Phase
Next, in Step S02, the learning model generating unit 132 executes machine learning to generate a learning model 121. After the learning model 121 is stored in the auxiliary storage device 102B, the processing is ended.
3. Information Processing in Estimation Phase
3-1. Configuration of Terminal Device
First, the terminal device 20 (
The communication unit 21 is a communication control device such as a network adapter, a network interface card, or a LAN card. The communication unit 21 connects the terminal device 20 to the communication network N by wire or wirelessly. The control unit 23 can transmit and receive various information to and from the information processing device 100 through the communication unit 21 and the communication network N.
The display unit 24 is an organic EL display or a liquid crystal display. The display unit 24 displays various information on a webpage based on control of the control unit 23.
The input unit 25 is a keyboard including character keys, numeric keys, and an enter key, etc., a mouse, a power supply button, etc. The user C can operate the webpage through the input unit 25. The display unit 24 and the input unit 25 may be configured integrally as a touch panel display.
The storage unit 22 is, for example, a semiconductor memory device such as a RAM or a flash memory, or a storage medium such as an HDD or an optical disc. The storage unit 22 stores software of applications to be executed by the terminal device 20. The storage unit 22 may store information received from the object surveying instruments M101 to M105.
The control unit 23 includes a microcomputer including a CPU, a ROM, a RAM, and I/O ports, etc., and various circuits. The control unit 23 reads and executes various programs stored in the storage unit 22 and the RAM. The control unit 23 includes an object information transmitting unit 231 and a result display unit 232. The functions of the respective functional units are realized by, for example, reading and executing programs stored in the ROM or the storage unit 22.
The object information transmitting unit 231 acquires information on the object surveying instruments M101 to M105 of the user C from the storage unit 22 or a management server, etc., of the user C and transmits the information to the information processing device 100 on a request from the object information acquiring unit 135 of the information processing device 100. It is also possible that the user C selects an object surveying instrument, information on which is to be transmitted to the information processing device 100, through the webpage.
The result display unit 232 receives estimation results from the result providing unit 138 described later, and displays the estimation results on the display unit 24. The estimation results are displayed in response to push notification or on a request from the user C. The estimation results that the result display unit 232 receives from the result providing unit 138 will be described in detail later with reference to
3-2. Configuration of Information Processing Device
The configuration of the information processing device 100 has already been illustrated in
The object information acquiring unit 135 acquires information stored in the object surveying instruments M101 to M105 of the user C as “object data.” The “object data” includes, as with the collected data 120, in addition to information on measuring function and information on measuring amount of the object surveying instruments, various information such as measurement data, image data, audio data, error code data, operating time data of components, driving data of components, and data on maintenance periods and rental periods.
Here, it is preferable that the object information acquiring unit 135 acquires the “object data” at intervals of a set “totaling period (unit period).” The totaling period can be set by minutes, hours, days, weeks, months, quarters, seasons, years, decade by decade, etc., or designated as a period such as “from Jan. 5, 2021 to Feb. 20, 2021.” As the totaling period, a default value is determined in the information processing device 100, however, it is preferable that the totaling period can be changed by the user C through the webpage.
The object information extracting unit 136 extracts information on measuring function and information on measuring amount used in the object surveying instruments M101 to M105 as “information on object measuring function” and “information on object measuring amount” from the “object data” acquired by the object information acquiring unit 135 in the same manner as in the input information creating unit 131, and uses these as input data of the user C. The object information extracting unit 136 functions each time the object information acquiring unit 135 acquires object data.
The estimating unit 137 inputs the input data of the user C into the learning model 121, and estimates “suitable surveying instruments” with respect to the “information on object measuring function” and “information on object measuring amount” of the user C (that is, an actual use situation of the user C). The estimating unit 137 functions each time the object information extracting unit 136 performs the extraction. That is, the estimating unit 137 performs one estimation at intervals of the “totaling period.”
“Estimation examples” of the estimating unit 137 are as follows.
(i) From the “information on object measuring function” and the “information on object measuring amount,” when it is found that a utilization rate of angle measuring with respect to a single point by the user C is high, the estimating unit 137 is likely to estimate any one of model number types of “theodolites.”
(ii) From the “information on object measuring function” and the “information on object measuring amount,” when it is found that a utilization rate of horizontal direction measuring and height difference measuring by the user C is high, the estimating unit 137 is likely to estimate any one of model number types of “levels.”
(iii) From the “information on object measuring function” and the “information on object measuring amount,” when it is found that the number of measurement points is large or the measurement range is wide at one site of the user C, the estimating unit 137 is likely to estimate any one of model number types of “scanners.”
(iv) From the “information on object measuring function” and the “information on object measuring amount,” when it is found that the number of prism measurements is large at one site of the user C, the estimating unit 137 is likely to estimate any one of model number types of “total stations” resistant to motor driving.
(v) From the “information on object measuring function” and the “information on object measuring amount,” when it is found that the user C performs distance and angle measuring but rarely uses the application functions, the estimating unit 137 is likely to estimate any one of model number types of inexpensive “total stations” equipped with no application functions.
Based on the learning model 121, the estimating unit 137 estimates one or several types of surveying instruments close to the input data of the user C (the “information on object measuring function” and the “information on object measuring amount”). The estimating unit 137 scores surveying instruments whose “general use model” patterns based on the learning model 121 are close to the input data according to numerical values of, for example, similarities, averages, medians, modes, accumulated values, and standard deviation, etc., or numerical values of matching ratios, etc., of shape matching with the “general use model,” or a combination of these, and determines surveying instruments with high scores as estimation results.
The result providing unit 138 presents the surveying instruments estimated by the estimating unit 137 on the result display unit 232 of the terminal device 20 in descending order of score together with reasons for recommendation. The reason for recommendation includes content proposing a replacement purchase or additional purchase of the surveying instrument.
For example, when the user C uses a sophisticated total station with a large number of application functions, the result providing unit 138 respectively proposes additional purchases of a theodolite, a level, and a scanner in cases where the estimating unit 137 presents the estimation examples (i) to (iii). When the estimating unit 137 presents the estimation examples (iv) to (v), the result providing unit 138 proposes a replacement purchase of a total station of a different model number (specifications).
With reference to
3-3. Information Processing Method in Estimation Phase
Next, in Step S12, the object information extracting unit 136 extracts “information on object measuring function” and “information on object measuring amount” from the “object data.”
Next, in Step S13, the estimating unit 137 inputs the “information on object measuring function” and the “information on object measuring amount” extracted in Step S12 into the learning model 121, and estimates surveying instruments suitable for an actual use situation of the user C.
Next, in Step S14, the estimation results are output by the result providing unit 138 to the terminal device 20 of the user C, and the processing is ended.
4. Effect
As described above, according to the present embodiment, the information processing device 100 is configured to extract “surveying instrument information,” “information on measuring function,” and “information on measuring amount” from big data of the surveying instruments, and by inputting an actual use situation (“information on object measuring function” and “information on object measuring amount”) of a user into the learning model 121 obtained by machine learning by using the sets of these “surveying instrument information,” “information on measuring function,” and “information on measuring amount” as learning data, estimate surveying instruments suitable for the actual use situation of the user. Accordingly, by comparison with the actual use situation, the user can know use of an excessively high-spec instrument or, conversely, use of a low-spec instrument, or can know the fact that a different instrument type is more suitable, etc. That is, according to the present embodiment, the user can recognize such an instrument mismatch, and can consider a replacement purchase of an inexpensive model or a change in contract plan according to estimation results. When the user is a dealer, the dealer can obtain materials for a proposal of a replacement purchase or an additional purchase to a customer.
According to the present embodiment, the information processing device 100 is configured to perform one estimation at intervals of a “totaling period,” and configured so that the “totaling period” can be arbitrarily changed by the user C. Accordingly, the user C can recognize an instrument mismatch according to the user's or customer's needs weekly, monthly, quarterly, or seasonally. Therefore, the user can more specifically consider a purchase or rental, and this leads to an improvement in customer satisfaction.
The present embodiment is configured so that the learning phase and the estimation phase are realized by the same information processing device 100, however, the learning phase and the estimation phase may be realized by different information processing devices. In this case, the information processing device that executes the estimation phase is configured to store the learning model 121 in the storage unit, or made accessible to a storage medium storing the learning model 121.
The preferred embodiment and modifications of the present invention have been described above, and the embodiment and modifications described above are examples of the present invention, and the embodiment and modifications can be combined based on the knowledge of a person skilled in the art, and such a combined embodiment is also included in the scope of the present invention.
| Number | Date | Country | Kind |
|---|---|---|---|
| 2021-054860 | Mar 2021 | JP | national |
