Demand prediction device and demand prediction method

Abstract

To provide a demand prediction device and demand prediction method that can appropriately predict the number of requests and request contents.SOLUTION: A demand prediction device 1 for predicting the demand for maintenance work on a predictive date includes: a storage device 1F that stores request information related to each past maintenance work request as a request history; and a processor 1A that predicts the number of maintenance work requests and request contents as the demand for maintenance work on the predictive date based on the request history stored in the storage device. The request information includes date attributes and external factor information of the day when the request was made, and the request contents. The processor executes steps of acquiring the date attributes and external factor information of the predictive date, and predicting the number of requests for each time zone on the predictive date and corresponding request contents based on the acquired date attributes and external factor information and the request history.SELECTED DRAWING: Figure 7

Description

Technical Field

[0001] The present disclosure relates to a demand prediction device that predicts the demand for maintenance work from customers in a maintenance service that causes workers to perform work in response to requests from customers.

Background Art

[0002] Conventionally, in load services, electric power companies, water and gas utilities, etc., a maintenance service has been widely performed in which, in response to requests from customers, workers are sent out from vehicles or the like and the workers are made to perform maintenance work.

[0003] In such a maintenance service, from the viewpoints of improving work efficiency and customer satisfaction, optimization of worker allocation, accurate grasping and management of work situations, and creation of an optimal schedule have become issues.

[0004] In order to address such issues, for example, a quantification unit that quantifies skill information representing the skills required for work and work information representing the amount of work of the work, and based on the skill information and the work information quantified by the quantification unit, an optimization algorithm execution unit that calculates a work schedule by calculating a circuit connecting work sites, and a presentation unit that presents the work schedule calculated by the optimization algorithm execution unit are provided. There is known a dispatch support device characterized by this (see Patent Document 1).

Prior Art Documents

Patent Documents

[0005]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0006] In the conventional seconded support device disclosed in the above Patent Document 1, the work schedule of the workers on duty is optimized. Therefore, when the number of interruption cases (also referred to as emergency cases) urgently requested on the day increases, it may be difficult to handle them only with the workers on duty, or there may be a risk of imposing an excessive burden on the workers on duty. In particular, since the number of interruption cases can also depend on external factors such as weather and temperature, it is not easy to predict the required number of personnel.

[0007] Therefore, a main object of the present disclosure is to provide a demand prediction device and a demand prediction method that can appropriately predict the number of requests and the content of requests in a maintenance service that assigns work to workers in response to requests from customers.

Means for Solving the Problems

[0008] The demand prediction device of the present disclosure is a demand prediction device for predicting the demand for maintenance work on a prediction date, and includes a storage device that stores, as a request history, request information related to each emergency request for the past maintenance work, and a processor that predicts the number of requests and the content of requests for the maintenance work on the prediction date as the demand for the maintenance work on the prediction date based on the request history stored in the storage device. The request information includes the date attribute and external factor information of the day when the request was made, and the content of the request. The processor executes steps of acquiring the date attribute and the external factor information of the prediction date, and predicting the number of requests for each time period of the prediction date and the corresponding content of the request based on the acquired date attribute and external factor information and the request history.

[0009] The demand prediction method of the present disclosure is a demand prediction method for predicting the demand for maintenance work on the prediction date, and includes a storage device that stores request information related to each of the past requests for the maintenance work as a request history, and a processor that predicts the number of requests for the maintenance work and the content of the request as the demand for the maintenance work on the prediction date based on the request history stored in the storage device. The request information includes the date attribute and external factor information of the date when the request was made, and the content of the request. The processor is configured to execute steps of acquiring the date attribute and the external factor information on the prediction date, and predicting the number of requests for each time period on the prediction date and the corresponding content of the request based on the acquired date attribute and external factor information and the request history.

Effect of the Invention

[0010] According to the present disclosure, it is possible to provide a demand prediction device and a demand prediction method that can appropriately predict the number of requests and the content of the requests.

Brief Description of the Drawings

[0011]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Best Mode for Carrying Out the Invention

[0012] A first invention made to solve the above problems is a demand prediction device for predicting the demand for maintenance work on a prediction date, comprising: a storage device that stores request information related to each of the past requests for the maintenance work as a request history; and a processor that predicts the number of requests for the maintenance work and the content of the requests as the demand for the maintenance work on the prediction date based on the request history stored in the storage device. The request information includes the date attribute and external factor information of the date when the request was made, and the content of the request. The processor executes steps of: acquiring the date attribute and the external factor information of the prediction date; and predicting the number of requests for each time period of the prediction date and the corresponding content of the requests based on the acquired date attribute and external factor information and the request history.

[0013] According to this, based on the date attribute and the external factor information, the number of requests and the content of the requests on the prediction date are predicted. Therefore, it is possible to provide a demand prediction device that can appropriately predict the number of requests and the content of the requests.

[0014] Further, a second invention relates to the first invention, and the processor executes steps of: acquiring the maintenance resources pre-allocated for the prediction date; and determining whether the pre-allocated maintenance resources can cope with the number of requests and the content of the requests on the prediction date, and outputting an alert signal when it is determined that they cannot cope.

[0015] According to this, when the pre-allocated maintenance resources are insufficient for the predicted number of requests and the content of the requests, an alert signal is output. Therefore, by giving a warning using the alert signal, it is possible to notify the user that the pre-allocated maintenance resources are insufficient.

[0016] Further, the third invention relates to the first invention, and when the processor determines that it is impossible to cope with the number of requests and the content of the requests on the predicted date with the maintenance resources allocated in advance, the processor is configured to output the alert signal including the shortage information of the maintenance resources.

[0017] According to this, since an alert signal including the shortage information of the maintenance resources is output, the degree of shortage of the maintenance resources can be notified to the user by using the alert signal.

[0018] Further, the fourth invention relates to the first invention, and includes a step of acquiring the maintenance resources allocated in advance on the predicted date, and determining whether it is possible to cope with the number of requests and the content of the requests on the predicted date with the maintenance resources allocated in advance. When it is determined that it is impossible to cope, a step of redistributing the maintenance resources is executed to cope with the number of requests and the content of the requests on the predicted date.

[0019] According to this, when the pre-allocated maintenance resources are insufficient, the maintenance resources are redistributed by the processor, so that the burden on the person in charge of managing the distribution of the maintenance resources can be reduced.

[0020] Further, the fifth invention relates to the first to fourth inventions, and the date attribute includes weekdays and holidays.

[0021] According to this, the date attribute can be obtained simply and appropriately.

[0022] Further, the sixth invention relates to the first to fourth inventions, and the date attribute includes peak seasons and slack seasons.

[0023] According to this, the date attribute can be obtained simply and appropriately.

[0024] Further, the seventh invention relates to the first to fourth inventions, and the date attribute includes the day after a holiday.

[0025] According to this, the date attribute can be obtained simply and appropriately.

[0026] Further, the eighth invention is related to the first to fourth inventions, and the external factor information is configured to include weather information.

[0027] According to this, the external factor information can be obtained simply and appropriately.

[0028] Further, the ninth invention is related to the first to fourth inventions, and is configured to include an input device for obtaining the external factor information.

[0029] According to this, the external factor information can be obtained simply.

[0030] Further, the tenth invention is a demand prediction method for predicting the demand for maintenance work on a prediction date, including a storage device that stores request information related to each of the past requests for the maintenance work as a request history, and based on the request history stored in the storage device, a processor that predicts the number of requests and the content of the requests for the maintenance work as the demand for the maintenance work on the prediction date. The demand prediction method is executed by a demand prediction device, the request information includes the date attribute and external factor information of the day when the request was made, and the content of the request, and the processor includes a step of obtaining the date attribute and the external factor information of the prediction date, and based on the obtained date attribute and external factor information, and the request history, predicting the number of requests for each time zone on the prediction date and the corresponding content of the request.

[0031] According to this, based on the date attribute and the external factor information, the number of requests and the content of the requests on the prediction date are predicted. Therefore, it is possible to provide a demand prediction method capable of appropriately predicting the number of requests and the content of the requests.

[0032] Hereinafter, the demand prediction device and the demand prediction method of the present disclosure will be described with reference to the drawings.

[0033] FIG. 1 shows that a demand prediction device 1 according to an embodiment of the present disclosure is provided in a maintenance system 2 for performing a maintenance service that, in response to a request from a customer, sends a worker staying at a base to a work site by a vehicle V or the like and has the worker perform maintenance work.

[0034] The maintenance system 2 is used by a person who operates a maintenance service in which a worker goes to a work site by a vehicle V or the like, for example, a business operator providing a load service, an electric power company, a water and gas utility, etc. Hereinafter, mainly, an example in which the maintenance system 2 is used by an electric power company will be described.

[0035] The maintenance system 2 includes an administrator terminal 3, a worker terminal 4, and a demand prediction device 1. The administrator terminal 3 is used by an administrator who operates the maintenance system 2, and the worker terminal 4 is used by a worker who performs maintenance work. The worker terminal 4, the administrator terminal 3, and the demand prediction device 1 can be operated by a business operator that provides the maintenance service.

[0036] The administrator terminal 3, the worker terminal 4, and the demand prediction device 1 can communicate with each other via a known communication network 5 such as the Internet or an intranet, respectively.

[0037] The demand prediction device 1 is configured to be communicable with an information providing server 6 that provides external factor information via a network, in addition to the administrator terminal 3 and the worker terminal 4. The external factor information is information related to a factor for which a customer has urgently requested maintenance (hereinafter, factor information), and includes, for example, meteorological information such as weather (sunny, heavy rain, snow, etc.) and temperature. When the information providing server 6 receives a request signal for obtaining external factor information for a predetermined date from the demand prediction device 1, it transmits the predicted external factor information (for example, heavy rain forecast, etc.) for the corresponding date to the demand prediction device 1. In FIG. 1, an example in which one information providing server 6 is connected to the demand prediction device 1 is shown, but a configuration in which two or more information providing servers 6 are connected to the demand prediction device 1 may also be possible.

[0038] The administrator terminal 3 is a computer that receives inputs from the administrator and performs processes for presenting information to the administrator. The administrator terminal 3 includes an information device having a communication function such as a smartphone, a tablet terminal, or a notebook PC. The administrator terminal 3 may include known hardware such as a processor (CPU, MPU, etc.), a memory (RAM, ROM, etc.), a display, an input device, a communication interface, and a storage device (HDD, SSD, etc.).

[0039] The worker terminal 4 is a computer that receives inputs from the worker and performs processes for presenting information to the worker. The worker terminal 4 also includes an information device having a communication function such as a smartphone, a tablet terminal, or a notebook PC. The administrator terminal 3 may include known hardware such as a processor (CPU, MPU, etc.), a memory (RAM, ROM, etc.), a display, an input device, a communication interface (also referred to as a network interface), and a storage device (HDD, SSD, etc.).

[0040] The administrator terminal 3 and / or the worker terminal 4 acquires information related to a maintenance request for maintenance work requested by a customer. The information related to the maintenance request may include the work date and work content for which maintenance is desired. The received maintenance requests include requests that do not require urgency (hereinafter, normal requests) and requests that require maintenance to be performed on the same day as the day the request was made (hereinafter, urgent requests, also referred to as interrupt requests). An operator for acquiring information related to a maintenance request for maintenance work requested by a customer may be configured to use one of the administrator terminal 3 and / or the worker terminal 4.

[0041] The management terminal 3 and / or the worker terminal 4 obtains information related to maintenance resources by receiving inputs from managers and workers. The maintenance resources refer to the personnel, vehicle V, and equipment required for performing maintenance work. The management terminal 3 and / or the worker terminal 4 respectively obtains information related to the attendance date and attendance time of workers and information related to the vehicle V and equipment used for maintenance work. Information for obtaining maintenance resources may also be obtained by various sensors that detect the position of the vehicle V, etc.

[0042] The demand prediction device 1 includes a computer that implements a demand prediction method by executing a predetermined program and predicts the demand for maintenance work within a predetermined area on a predetermined date (hereinafter referred to as the prediction date) after the current date (the current day). In the present embodiment, the demand prediction device 1 obtains external factor information from the information providing server 6 and predicts the demand for urgent requests to be made on the prediction date. The computer constituting the demand prediction device 1 may include known hardware such as a processor 1A (CPU, MPU, etc.), a memory 1B (RAM, ROM, etc.), a display 1C, an input device 1D, a communication interface 1E, and a storage device 1F (HDD, SSD, etc., also referred to as a storage device), as shown in FIG. 2.

[0043] As shown in FIG. 3, the demand prediction device 1 includes, as functional units, a communication unit 11, a storage unit 12, and a control unit 13.

[0044] The communication unit 11 may be constituted by the communication interface 1E. The communication unit 11 performs wireless communication or wired communication with other devices (here, the management terminal 3, the worker terminal 4, the information providing server 6, etc.) via the communication network 5 in accordance with a known communication protocol.

[0045] The storage unit 12 can be composed of a memory 1B and a storage device 1F. The storage unit 12 stores various types of information required for the processing performed by the control unit 13. The information stored in the storage unit 12 may include information related to each urgent request for maintenance work (hereinafter referred to as urgent request information). The storage unit 12 stores past urgent request information as an urgent request history in a database (hereinafter referred to as the urgent request database 12A).

[0046] FIG. 4 shows an example of the urgent request database 12A stored in the storage unit 12. In the urgent request database 12A, a request history is recorded for each urgent request. The request history includes the identification number (ID) of each urgent request and the corresponding request information.

[0047] The request information includes the date (i.e., the date when the request was made), the date attribute, and the time zone of the day when the maintenance work was performed based on the urgent request, the external factor information of that day, and the information related to the factor that led to the request for the maintenance work (i.e., the factor information).

[0048] The date attribute mentioned here means an attribute related to the date, and may include information such as weekdays and holidays, peak seasons and off-peak seasons, and the day after a long holiday (the day after a holiday). In the present embodiment, for each urgent request, information indicating whether the day when the maintenance work was performed is a weekday or a holiday is recorded in the urgent request database 12A as the date attribute.

[0049] The external factor information included in the request information may be meteorological information such as weather and temperature. In the present embodiment, as the external factor information included in the request information, the weather (sunny, rainy, snowy, etc.) on the day when the urgent request occurred (i.e., the day when the maintenance work was performed) is recorded in the urgent request database 12A.

[0050] For example, when heavy snow falls, due to the weight of the snow, wire repair may be required, and an urgent request may be made. Thus, it is considered that there may be a causal relationship between an urgent request and external factors including meteorological conditions such as the weather and temperature on the day when the maintenance work was performed based on the urgent request.

[0051] The information stored in the memory unit 12 may include information related to normal requests for maintenance work before the predicted date (for example, date, time zone, work content, etc.).

[0052] The memory unit 12 includes information related to maintenance resources that can respond to emergency requests (hereinafter referred to as emergency maintenance resource information). FIG. 9 shows an example of emergency maintenance resource information related to personnel.

[0053] The emergency maintenance resource information includes the time zones in which each worker on the predicted date can respond to emergency requests. The memory unit 12 holds a database (maintenance resource database 12B) in which information related to the emergency maintenance resource information is recorded. FIG. 5 schematically shows an example of the emergency maintenance resource information recorded in the maintenance resource database 12B. In FIG. 5, the working hours of each worker are shown as thick-bordered frames, and the working hours for normal requests and work that does not require emergency (hereinafter collectively referred to as normal work) are shown as shaded portions.

[0054] That is, in FIG. 5, the portion surrounded by the thick line and not shaded is the time zone in which emergency requests can be responded to, that is, it corresponds to the emergency maintenance resources. That is, the emergency maintenance resources correspond to the portion of the maintenance resources excluding the maintenance resources allocated to respond to normal requests (hereinafter referred to as normal maintenance resources). That is, the emergency maintenance resources can be said to be the surplus of the maintenance resources.

[0055] Although not shown in FIG. 5, for each vehicle V, the available time zone and the scheduled usage time zone may similarly be recorded in the maintenance resource database 12B as emergency maintenance resource information related to the vehicle V.

[0056] In the maintenance services provided by power companies, routine operations typically include, for example, regular inspections of electric wires and inspections of facilities. In addition, as maintenance services, there are operations in response to urgent requests (hereinafter referred to as emergency operations), such as repairs of electric wires and utility poles and responses to power leakage. In road services, routine operations include repairs and inspections of vehicle V, and emergency operations include towing, repairing, wheel dropping, and responses to fuel exhaustion of vehicle V. In the maintenance services of water and gas utilities, routine operations include pipe maintenance and operations for deterioration inspection, and emergency operations include operations for responding to gas leakage, water leakage, and damage to pipes. Note that emergency operations may also include operations other than other routine operations (for example, newly generated additional operations and additional operations due to communication omissions or registration omissions).

[0057] The allocation of regular maintenance resources may be performed by displaying an input screen on the administrator terminal 3 and receiving input from the administrator. At this time, the administrator allocates regular maintenance resources by setting the working hours for each worker so that all routine operations can be performed without problems.

[0058] At this time, if levels are assigned to each worker and the types of operations that each worker can perform are different, the administrator may allocate the working hours considering the levels and the types of operations that each worker can perform.

[0059] In addition, the maintenance resources may include resources related to vehicle V in addition to personnel. For vehicles V (see FIG. 1) having special functions such as an aerial work vehicle V1 and a high-voltage generator vehicle V2, the administrator may allocate vehicle V to routine operations considering its functions.

[0060] Also, the allocation of normal maintenance resources may be configured to be automatically performed by the demand prediction device 1. Note that, whether it is performed by the administrator or by the demand prediction device 1, the allocation to normal maintenance resources shall be made before the day on which maintenance related to normal requests is performed and on the day before the prediction date.

[0061] The storage unit 12 may store a skill table indicating the level for each worker and a vehicle table in which information such as the functions for each vehicle V is recorded. In addition, the storage unit 12 may store, as information for deriving the maintenance resource database 12B, for example, a table indicating the working hours for each worker, a pre-reserved case table in which information on cases with pre-reserved work is recorded, and the like.

[0062] The control unit 13 may be configured by the processor 1A executing a program stored in the memory 1B or the storage device 1F.

[0063] The control unit 13 includes, as functional units, a demand prediction unit 21 and an allocation update unit 22. Note that the allocation update unit 22 is not essential and is provided when the allocation of normal maintenance resources is automatically performed by the demand prediction device 1.

[0064] When there is a predetermined input on the administrator terminal 3, the demand prediction unit 21 (processor 1A) performs a demand prediction method and predicts the number of requests for maintenance work and the content of the requests as the demand for maintenance work on the prediction date based on the request history stored in the emergency request database 12A of the storage unit 12. Hereinafter, the details of the demand prediction process performed by the demand prediction unit 21 (processor 1A) will be described with reference to the flowchart shown in FIG. 6.

[0065] When it is necessary to start the prediction process, the demand prediction unit 21 first acquires external factor information for the prediction date from the information providing server 6 (S101). The external factor information may include the weather, temperature, etc. on the prediction date. In this embodiment, the demand prediction unit 21 acquires the weather forecast (sunny, rainy, heavy rain, snowy, heavy snow, etc.) on the prediction date as the external factor information from the information providing server 6 (weather server).

[0066] Next, the demand prediction unit 21 acquires the date attribute of the prediction date (S102). In this embodiment, the demand prediction unit 21 acquires information (for example, a flag, etc.) indicating whether the prediction date is a weekday or a holiday as the date attribute.

[0067] Subsequently, based on the acquired date attribute and external factor information of the prediction date, and the request history of the emergency request database 12A stored in the storage unit 12, the demand prediction unit 21 predicts the number of requests and the request content on the prediction date. In this embodiment, the demand prediction unit 21 extracts a request history including external factor information that is substantially the same (or similar) to the external factor information on the prediction date and a date attribute that is substantially the same (or similar) to the date attribute of the prediction date from the emergency request database 12A. The demand prediction unit 21 performs demand prediction of the emergency request on the prediction date based on the extracted request history (S103).

[0068] In this embodiment, the demand prediction unit 21 outputs the number of requests for the emergency request and the corresponding request content for each time period (i.e., in time units) predicted on the prediction date based on the extracted request history.

[0069] At this time (in S103 of FIG. 6), the demand prediction unit 21 may count each emergency request included in the extracted request history for each request time and each request content, and calculate the average (overall average or moving average) to predict the number of requests for the emergency request for each request content.

[0070] In addition, the demand prediction unit 21 may classify (cluster) the request information recorded in the emergency request database 12A in advance based on patterns such as "December, holiday, heavy snow", and record it in the storage unit 12. In that case, the demand prediction unit 21 may select the class corresponding to the prediction date, and using the clustered result, obtain the number of emergency requests and the corresponding request content for each time period of the prediction date.

[0071] When the demand prediction of the emergency request for the prediction date is completed, the demand prediction unit 21 acquires information related to the maintenance resources pre-allocated from the storage unit 12, and determines whether it is possible to handle the number of requests and the request content on the prediction date with the pre-allocated maintenance resources without changing the pre-allocated maintenance resources.

[0072] In order to handle interrupt cases (emergency cases), the pre-allocated maintenance resources may be allocated so that surplus maintenance resources are generated with respect to all the maintenance resources. As the maintenance resources, for example, KPIs (Key Performance Indicators) may be set so that surpluses are generated for the operation rate, travel time, and working time, respectively. Specifically, if the operation rate is set to 80%, the surplus operation rate will be about 20%, making it possible to handle emergency requests.

[0073] In the present embodiment, first, the demand prediction unit 21 acquires emergency maintenance resource information as information related to the maintenance resources pre-allocated from the storage unit 12. Then, the demand prediction unit 21 calculates the resources such as the number of personnel and vehicle V required to handle the number of requests and the request content on the prediction date. Further, the demand prediction unit 21 determines whether it is possible to allocate emergency maintenance resources to all the resources required to handle the number of requests and the request content on the prediction date, that is, whether it is possible to handle them without changing the pre-allocated maintenance resources (S104).

[0074] At this time, the demand prediction unit 21 may calculate the total maintenance resources required after adding the work corresponding to the predicted urgent request to the normal work, and determine that it is unable to respond when the difference from the normally allocated maintenance resources is equal to or greater than the threshold (i.e., when there is a shortage).

[0075] When it is determined that it is possible to respond with the pre-allocated maintenance resources, the demand prediction unit 21 ends the demand prediction process.

[0076] When it is determined that it is possible to respond with the pre-allocated maintenance resources, the demand prediction unit 21 transmits a first alert signal for causing the administrator terminal 3 to display an alert screen (first warning screen 30).

[0077] FIG. 7 shows an example of a warning screen displayed on the administrator terminal 3 when the first alert signal is transmitted. The first warning screen 30 may include text 30A indicating that the pre-allocated maintenance resources are insufficient, an icon, and the like. The first warning screen 30 may also include a display column 30B for displaying the predicted number and content of urgent requests, and a display column 30C for displaying information related to the operator on the predicted date (for example, name, vehicle V scheduled to be used, scheduled working time, travel time, and level (skill)).

[0078] When the transmission of the first alert signal (or the display of the first warning screen 30) is completed, the demand prediction unit 21 ends the demand prediction process.

[0079] However, when the demand prediction unit 21 determines that it is impossible to respond to the number of requests and the content of requests on the predicted date with the pre-allocated maintenance resources, it may generate an alert signal including information related to the shortage part of the maintenance resources (shortage information). For example, when there is a time period in which the number of personnel corresponding to the number of requests and the content of requests on the predicted date is insufficient, the demand prediction unit 21 may generate an alert signal based on the number of such time periods.

[0080] In that case, as shown in FIG. 7, the administrator terminal 3 may display a first warning screen 30 including time zones with insufficient staff. Additionally, the administrator terminal 3 may be configured to display, on the first warning screen 30, the degree of shortage (such as the length of the time zone with insufficient staff and the number of personnel) based on the shortage information, using colors, icons, etc.

[0081] In this way, since an alert screen is displayed when there is a shortage of maintenance resources, the administrator can easily recognize that there is a shortage of maintenance resources and can review the allocation thereof. Also, by considering the number of requests and the content of the requests, it is possible to predict the maintenance resources required on the predicted date, thereby preventing an excessive burden on the workers working on the predicted date.

[0082] There may be cases where the number of emergency requests increases due to bad weather such as heavy snow. Since the demand prediction unit 21 performs prediction using the date attribute and the external factor information, the number of requests and the content of the requests on the predicted date are appropriately predicted.

[0083] The allocation update unit 22 (processor 1A) performs an allocation update process when there is an input (for example, an input to the reallocation button shown in FIG. 7) to reallocate the previously allocated maintenance resources again on the first warning screen 30. The allocation update process is a process in which the allocation update unit 22 reallocates the maintenance resources to correspond to the number of requests and the content of the requests on the predicted date. Hereinafter, the details of the allocation update process performed by the allocation update unit 22 (processor 1A) will be described with reference to the flowchart shown in FIG. 8.

[0084] In the distribution update process, the distribution update unit 22 first determines whether it is possible to handle including the emergency requests predicted on the prediction date by redistributing maintenance resources using a known method (S201). That is, the distribution update unit 22 determines whether it is possible to execute the work for handling the emergency requests predicted on the prediction date and the normal work by redistributing the maintenance resources. Examples of redistributing the maintenance resources include increasing the number of standby personnel in the time period when the maintenance resources are insufficient by shifting the time period of the normal work or changing the break time, etc.

[0085] When it is possible to execute the emergency requests on the prediction date and the normal work by redistributing the maintenance resources, based on the redistributed maintenance resources, the distribution update unit 22 updates the emergency maintenance resource information stored in the storage unit 12 (S201). When the update is completed, the distribution update unit 22 transmits a change notification signal notifying that the distribution of the maintenance resources has been changed to the administrator terminal 3.

[0086] When the administrator terminal 3 receives the change notification signal, it displays a distribution change notification screen 32 notifying that the distribution of the maintenance resources has been changed. FIG. 9 shows an example of the distribution change notification screen 32 displayed by the administrator terminal 3 when the change notification signal is transmitted. The distribution change notification screen 32 includes a text 32A indicating that the distribution of the maintenance resources has been changed, an icon, etc. The distribution change notification screen 32 preferably also includes a display column 32B for displaying the number and content of the predicted emergency requests, and a display column 32C for displaying information related to the workers on the prediction date.

[0087] When the transmission of the change notification signal (or the display of the distribution change notification screen 32) is completed, the distribution update unit 22 ends the distribution update process.

[0088] When it is difficult to execute the emergency requests on the prediction date and the normal work by redistributing the maintenance resources, the distribution update unit 22 transmits a second alert signal for causing the administrator terminal 3 to display a second alert screen (warning screen) (S203).

[0089] FIG. 10 shows an example of a warning screen displayed on the administrator terminal 3 when the second alert signal is received. The second warning screen 34 may include text 34A, an icon, etc. indicating that there may be a shortage of maintenance resources that can be used on the day. The second warning screen 34 may also include a display column 34B for displaying the predicted number and content of urgent requests, and a display column 34C for displaying information related to the operator on the predicted date.

[0090] When the transmission of the second alert signal (or the display of the second warning screen 34) is completed, the allocation update unit 22 ends the allocation update process.

[0091] In the above embodiment, the demand prediction device 1 is configured to perform demand prediction in a predetermined area, but it may be configured to perform demand prediction in a plurality of areas. In that case, the demand prediction device 1 may cluster the demand history for each area and perform demand prediction.

[0092] In the above embodiment, the demand prediction device 1 has acquired the external factor information from the information providing server 6, but is not limited to this mode. The demand prediction device 1 may acquire the external factor information by input to the input device 1D from an administrator or the like. Thereby, the external factor information can be simply acquired without connecting to the information providing server 6.

[0093] As described above, the embodiments have been described as examples of the technology disclosed in the present application. However, the technology in the present disclosure is not limited to this, and can also be applied to embodiments in which changes, replacements, additions, omissions, etc. are made. It is also possible to combine the respective components described in the above embodiments to form a new embodiment.

Industrial Applicability

[0094] The demand prediction device and the demand prediction method according to the present disclosure are useful as a demand prediction device and a demand prediction method that can appropriately predict the number of requests and the content of requests in a maintenance service that causes an operator to work in response to a request from a customer.

Explanation of Symbols

[0095] 1: Demand Prediction Device 1A: Processor 1B: Memory 1C: Display 1D: Input Device 1E: Communication Interface 1F: Storage Device 2: Maintenance System 3: Administrator Terminal 4: Operator Terminal 5: Communication Network 6: Information Providing Server 11: Communication Unit 12: Storage Unit 12A: Emergency Request Database 12B: Maintenance Resource Database 13: Control Unit 21: Demand Prediction Unit 22: Allocation Update Unit 30: First Warning Screen 30A: Text 30B: Display Column 30C: Display Column 32: Allocation Change Notification Screen 32A: Text 32B: Display Column 32C: Display Column 34: Second Warning Screen 34A: Text 34B: Display Column 34C: Display Column V: Vehicle V1: Aerial Work Platform V2: High-Voltage Generator Vehicle

Claims

1. A demand prediction device for predicting the demand for maintenance work on a prediction date, comprising: a storage device that stores request information related to each of the past emergency requests for the maintenance work as a request history; a processor that predicts the number of requests for the maintenance work and the content of the requests as the demand for the maintenance work on the prediction date based on the request history stored in the storage device; the request information includes the date attribute and external factor information of the date when the request was made, and the content of the request; the processor: acquires the date attribute and the external factor information on the prediction date; A demand prediction device that executes steps of predicting the number of requests for each time period on the prediction date and the corresponding request content based on the acquired date attribute and external factor information and the request history.

2. the processor: acquires information related to maintenance resources pre-allocated on the prediction date; determines whether the pre-allocated maintenance resources can handle the number of requests and the content of the requests on the prediction date, and when it is determined that they cannot handle, outputs an alert signal. The demand prediction device according to claim 1.

3. The demand prediction device according to claim 2, wherein when the processor determines that the pre-allocated maintenance resources cannot handle the number of requests and the content of the requests on the prediction date, the processor outputs the alert signal including the shortage information of the maintenance resources.

4. acquires the maintenance resources pre-allocated on the prediction date; determines whether the pre-allocated maintenance resources can handle the number of requests and the content of the requests on the prediction date, and when it is determined that they cannot handle, reallocates the maintenance resources to handle the number of requests and the content of the requests on the prediction date. The demand prediction device according to claim 1.

5. The demand prediction device according to any one of claims 1 to 4, wherein the date attribute includes weekdays and holidays.

6. The demand prediction device according to any one of claims 1 to 4, wherein the date attribute includes peak seasons and slack seasons.

7. The demand prediction device according to any one of claims 1 to 4, wherein the date attribute includes the day after a holiday.

8. The demand prediction device according to any one of claims 1 to 4, wherein the external factor information includes weather information.

9. The demand prediction device according to any one of claims 1 to 4, further comprising an input device for acquiring the external factor information.

10. A demand prediction method for predicting the demand for maintenance work on a prediction date, comprising: a storage device that stores request information related to each of the past requests for the maintenance work as a request history; a processor that predicts the number of requests for the maintenance work and the content of the requests as the demand for the maintenance work on the prediction date based on the request history stored in the storage device, the demand prediction method being executed by a demand prediction device comprising the storage device and the processor; the request information including the date attribute and external factor information of the date on which the request was made, and the content of the request; the processor: acquiring the date attribute and the external factor information of the prediction date; predicting the number of requests for each time period on the prediction date and the corresponding content of the requests based on the acquired date attribute and external factor information and the request history.

Images