JP5256089B2 - Parts demand forecasting method, parts demand forecasting system - Google Patents

Description

  The present invention relates to a demand prediction method for parts constituting a device. More specifically, in the event that a malfunction occurs in the equipment used by the user, the repairer may replace the equipment parts in order to correct this malfunction. It relates to a demand forecast method.

  In the repair work of electrical equipment and gas equipment, it is important to predict the demand for parts required for repair.

  If there is a shortage of parts in stock, the repairer must wait for the parts to arrive and go back to repairs, leading to increased visits to the customer. This is inefficient in terms of business efficiency, and also affects customer satisfaction in terms of longer days to complete repairs.

  On the other hand, for example, by accurately predicting parts demand on a weekly / monthly / periodic basis, it is possible to reduce the occurrence of out of stock and solve the above-mentioned problems.

  In addition, accurately predicting parts demand leads to avoiding the situation of purchasing and storing parts that are rarely used (stock parts).

  It is customary for repairers of electrical equipment and gas equipment to purchase parts of equipment to be repaired by themselves and bring the purchased parts to customers for repair. At this time, the parts of the device are stored by the repairer himself. Therefore, in order to store parts in a limited installation space, there is a desire to avoid unnecessary stock as much as possible. There is also a desire to reduce costs by refraining from purchasing unnecessary parts.

  In this regard, if the demand for parts can be accurately predicted, the number of stock parts can be reduced, so that the problem of installation space and the problem of parts purchase cost can be solved.

  As a conventional example of a demand prediction method in parts inventory management, a technique for predicting a demand at the end of a part group whose frequency is particularly low is provided in Patent Document 1 below.

JP 2002-73746 A

  The method described in Patent Document 1 is to determine the parameters of the probability distribution model from the past component usage results and predict future component demand based on the parameters.

  In the case of this method, the demand is simply predicted based on the past results of component use. However, when a certain period of time has elapsed, the customer may replace the device itself instead of replacing the parts. In other words, there are devices (so-called old-type devices) in which the number actually used decreases with time.

  Some parts are common to old and new equipment, while others differ from equipment to equipment (parts specific to equipment). Therefore, in the case of a part peculiar to an old style device, the number of use of the device itself decreases with the lapse of time, and accordingly, the number of use of the component also decreases. As a result, the conventional method inevitably reduces the demand for parts.

  However, when a malfunction occurs in a device, just because a customer owns an old-style device, it is not always necessary to replace the device. In particular, if a certain period of time has passed since the installation of an old-type device, it may be considered that a malfunction of a part is likely to occur.

  Considering such points, especially in the case of equipment such as gas equipment and power equipment that are regularly used every day, the method described in Patent Document 1 can always perform demand prediction with high accuracy. I can't say that.

  In view of the above problems, an object of the present invention is to increase the prediction accuracy of a method for predicting the demand for parts to be replaced in equipment repairs.

  In order to achieve the above object, a parts demand prediction system according to the present invention is configured with the following features.

That is, the present invention is a parts demand prediction system that predicts demand for parts required for equipment repair, and includes a customer-owned equipment data storage unit, a repair work report data storage unit, a component usage record data storage unit, and a component use record A data creation unit, a component usage rate calculation unit, an existing number calculation unit by device type, and a component demand prediction unit by device type,
The customer-owned equipment data storage unit is
Data relating to owned equipment that is currently owned by the owning customer or previously owned, including at least information on the owning customer, information on the type of the owned equipment, and information on when the equipment was first used. If the device has been owned in the past, it records customer-owned device data that further includes information on the end of use,
The repair work report data storage unit is
Data on the contents of repairs that have been performed in the past, at least information on the repair customer who performed the repair, information on the model of the repair device that performed the repair, information on the parts used in the repair, and the repair Repair work report data including information on repair timing
The component usage record data creation unit
By associating the information of the owned customer indicated in the customer owned equipment data with the information on the type of the owned equipment, and the information on the repair customer indicated in the repair work report data with the information on the type of the repaired equipment, Create part usage record data including information on a specific usage period that defines the model, the used part used for repair, and the period from the use start time to the repair time as a multiple of a predetermined unit period. Record it in the usage data storage unit,
The component usage rate calculation unit
The number of actual use of all the owned devices registered in the customer-owned device data storage unit is the period in which the user has experienced use and is defined by a multiple of the unit period, and for each device type To calculate the number of devices by device type and experience period,
Based on the component usage record data, the number of the repair devices in which the used components are actually used is counted for each specific usage period and for each device type, and for each used component, By dividing by the number of devices by experience type by device type, the ratio of the number of used parts required for all the owned devices was defined for each specific use period and for each device type and for each used component. Calculate the component usage rate that is the value,
The existing number calculation unit by the device type is:
Based on the customer-owned device data, the number of devices that are currently owned by the owned customer among the owned devices is defined as a multiple of the unit period from the start of use to the present. By counting for each actual use period and for each device type, the number of existing devices by the actual use period by the device type of the owned equipment is calculated,
The part demand forecasting unit by device type is
With respect to the same used parts, the calculation result obtained by multiplying the corresponding specific usage period and the actual usage period and the corresponding device type by the number of parts used and the existing number by the actual usage period by the equipment type is the same. The number of predicted demands for each used part over the unit period is derived by adding up each used part.

  Hereinafter, the configuration of the system having the above features will be described in more detail.

  The part usage record data is intermediate data in the present system, which is created for associating and integrating customer-owned equipment data and repair work report data. This data was created especially when the repair contents shown in the repair work report data were completed after the repair equipment was actually used. Done to figure out. What defines the period from the use start time to the repair time as a multiple of the unit period corresponds to the “specific use period”.

  In addition, here, “specify in multiple of unit period” is defined as an integer value based on the value obtained by dividing the period from the start of use to the repair period by the unit period when the unit period is one year. As an example, it can be defined by an integer value obtained by rounding down the decimal point of the division value.

  The “experience period” used when calculating the number of units by experience type by device type is, for example, a device that has been used for 3.5 years since the start of use. Experience the use of period. Therefore, if the experience period is defined by an integer value obtained by dividing 3.5 by the unit period and rounded down to the nearest decimal point, one unit will be provided for each year of the experience period of 0 years, 1 year, 2 years, and 3 years. It will be counted one by one. In this example, the experience period of 0 years means that the device has actually been used for 0 to 1 year. Similarly, the experience period of 1 year is 1 to 2 years of use, and the experience period of 2 It refers to use over a period of 2 to 3 years and use over a period of 3 to 4 years.

  Depending on the number of devices by device type by experience period, all owned devices that have been used in the past or are still used are counted for each period of actual use. That is, among these devices, how many devices have been used over a period of 0 to 1 year, how many devices have been used over a period of 1 to 2 years, and so on. Can be grasped.

  For example, paying attention to “Equipment A” that is 2 years or more and less than 3 years from the start of use, this device has also experienced a period of 0 years or more and less than 1 year from the start of use. I have also experienced each season from 2 years to less than 3 years. That is, for one device A, one device is counted for each of “0 years”, “1 year”, and “2 years”.

  Focusing on “Device B”, which is 4 years or more and less than 5 years from the start of use, and thinking in the same way as in the case of Device A, this device naturally experiences a period of 0 years or more and less than 1 year. They have also experienced 1 to 2 years, 2 to 3 years, 3 to 4 years, 4 to 5 years. That is, for one device B, one device is counted for each of “0 years”, “1 year”, “2 years”, “3 years”, and “4 years”.

  That is, for each device with a specific usage period of C years (C is an integer of 0 or more), the number of units by experience period is counted in all integer years of 0 or more and C or less. Accordingly, the number of vehicles by experience period naturally increases as the experience period is shorter, and conversely indicates data that decreases as the experience time increases.

  For example, even if it is a device that has been erased immediately after the start of use, or a device that has been in use for a long time, it will experience a situation where the period of use is "0 years or more and less than 1 year". It is counted in the period “0 years”. On the other hand, if the period of experience is “5 years”, that is, the period of use is “5 years or more and less than 6 years”, it will not be counted unless it has been continuously used for at least 5 years from the start of use. In other words, devices deleted within 5 years from the start of use are not counted. Therefore, naturally, the number is smaller than the number of vehicles in the experience period “0 year”.

  The “experience period” is a different concept from the “actual use period” used when calculating the “existing number of machines by actual use period by device type”.

  The “component usage rate” is an index indicating how long and how long a device is used and what proportion of components are required.

  In deriving this part usage rate, for each part for which a part has been replaced, the number of parts replaced for each period that has elapsed since the start of use of the equipment at the time of replacement (i.e., Deriving the number of devices that have been replaced. More specifically, the number of parts exchanged for each “specific use period”, in which the period elapsed from the start of use is defined as a multiple of the unit period, is derived. This derivation is performed for each component in the device, and the same is performed for other devices. As a result, the number of devices for which parts have been newly replaced during past repairs is calculated for each used component by device type and by specific usage period.

  Assuming that the unit period is one year, specifically, for the part i of the type X equipment, the number of equipment of the type X in which the part i is exchanged during the specific usage period of 0 to 1 year is N (X , I, 1) units, the number of devices of type X in which the same part i is replaced in one to two years is calculated as N (X, i, 2) units, and so on. This is calculated for all parts i in the device of type X and for all devices of type X.

  Then, by dividing the number of units by experience period by device type by the number of units by specific use period for each corresponding device type and period, the component usage rate by part by device type for each specific use period is obtained. To derive. The component usage rate r (X, i, y) obtained in this way is a function of X, i, y as a probability that a device of type X whose specific usage period y has passed will replace the component i. Represented.

  If a particular device α (device type is X) is in the fourth year of use, this device has actually experienced (used) all years less than four years. For this reason, it counts as a number for every specific use period with respect to each of a specific use period 0 years, 1 year, 2 years, 3 years, and 4 years. On the other hand, when a certain part i constituting this type X device is replaced in the fourth year, the presence of this device is counted as the number of type X devices that have been replaced in the fourth year. .

  In other words, the fact that this device α has replaced the component i in the fourth year is counted as the number of use of the component i in the fourth year of the model X device. On the other hand, this apparatus (alpha) is counted every year of 0 to 4 years as the number of apparatus of type X for every specific use period. That is, the “component usage rate” obtained by dividing the number of parts i used for each specific usage period of the type X equipment obtained in this way by the number of type X equipment for each experience period is This means that the existence of a certain device α that is replaced in the fourth year without being replaced until the third to third years among the components i, which is one component of the device of type X, is taken into consideration. To do.

  According to the above explanation, this “part usage rate” indicates that each device that has been present or has been replaced in the past has been replaced every time it has been used, or has not been replaced. It is understood that the ratio of using a part for each device type is a value calculated for each actually used period in consideration of the actual results.

  Then, by counting the number of devices currently owned by the owning customer for each device type for each actual usage period from the start of use to the present, it is a device of type X that has passed the period y from the start of use. The number W (X, y) currently used is determined, and the existing number W (X, y) and the component utilization rate r (X, i, y) according to the actual usage period according to the device type are expressed as X, By multiplying by i and y, the number of parts i used for each device type (X) actual usage period (y) is calculated. Then, this calculation is performed for each actual use period for each part, and is added up between the same parts. Thereby, the predicted demand number in the unit period for each part is obtained.

  The value obtained in this way counts the number of devices that have been replaced in the past, for each length of period that the device was actually used, and for each component that was replaced at that time. Based on the calculated values. Even in the same equipment model, there may be a tendency that the parts that cause the failure differ depending on the period of use. This depends on the period that has elapsed since the start of use of the equipment currently in use. This means that the parts that are expected to be defective, in other words, the types of parts that need to be replaced are different.

  The predicted demand calculated by the system of the present invention is based on the past repair record, taking into account the specific usage period in which the device was actually used by the time of repair and the tendency of the parts replaced at that time. In addition, this is a value calculated by reflecting the number of units actually used, which indicates how much time has elapsed since the start of use of the devices currently used. Therefore, the system can be a useful system for predicting the parts demand for devices that tend to produce differences in the types of parts that cause problems, particularly depending on the period of use.

  In addition to the above features, the present invention preferably further includes the following points.

That is, when the unit period is one year,
The part demand forecasting unit by device type is
Based on the part usage record data, the number of used parts for each used part is totaled for each month indicated by the repair period, and further, the ratio of the total value for each month is obtained to obtain the monthly usage ratio for each used part. To calculate
After deriving the predicted demand number for each used part over one year, the predicted demand number for each used part by month is derived by proportionally allocating according to the monthly usage ratio.

  According to this configuration, it is possible to predict the demand for monthly used parts. Here, when calculating the monthly predicted demand, the monthly usage ratio is calculated based on the number of monthly used parts based on the past repair record, and the annual predicted demand is proportionally distributed based on this value. doing.

  Among gas and power equipment, there are equipment whose operating conditions are affected by the outside temperature, such as air-conditioning equipment and heating equipment, which are considered to be relatively likely to be repaired by gas companies and power companies. However, in the case of these devices, the usage mode changes greatly according to the four seasons. For this reason, it is also conceivable that a malfunction of the device is induced due to the change in the usage mode. According to this feature configuration, by performing monthly demand prediction based on past results, it is possible to perform detailed demand prediction in consideration of the influence of outside air temperature by a simple method.

  In addition to the above features, the present invention preferably further includes the following points.

That is, the device type-specific parts demand forecasting unit
Based on the part usage record data, the number of used parts for each used part is tabulated according to weekdays and holidays indicated by the repair time, and the tabulated value for each weekday / holiday is further divided by the number of weekdays / holidays when the tabulation is performed. Later, by calculating these ratios, the usage ratio by weekdays and holidays is calculated,
After calculating the forecasted demand per day by dividing the forecasted demand for each used part by month by the number of days organized by month, the day-by-weekday / holiday It is characterized by deriving the number of forecast demand per hit.

  With this configuration, it is possible to calculate the predicted number of demands for each day while taking into account the difference in parts replacement frequency due to the difference between weekdays and holidays. Of course, according to the system of the present invention, the demand for parts can be predicted on a daily basis. However, when actually operating the system of the present invention, for example, the predicted demand is performed on a weekly or biweekly basis. It can be used in some cases. In this case, the predicted demand numbers for each weekday and each holiday may be added according to the number of weekdays and the number of holidays existing within the target period for which the demand prediction is to be performed. In addition, when the target period spans different months, the predicted demand number for each weekday / holiday may be added for each month according to the number of weekdays and the number of holidays in the month.

  In addition to the above features, the present invention preferably further includes the following points.

That is, in addition to the above configuration, the configuration further includes a weather data storage unit that records weather data including at least past temperature data over a plurality of years from the latest,
The part demand forecasting unit by device type is
While reading out the weather data from the weather data storage unit, by collating with the repair time of the parts usage record data, the number of used parts for each used part for each temperature range constituted by a predetermined temperature range, Furthermore, by deriving these ratios, the component usage ratio by temperature range is calculated,
Further, an average temperature over the unit period is derived based on the weather data, a reference temperature zone to which the average temperature belongs is recognized, and the predicted number of demand for each used part over the unit period is divided by the number of days in the unit period. By calculating the number of predicted demand for each used part on the day that belongs to the reference temperature range, and to this value, the component usage ratio for each temperature range in the reference temperature range and the prediction target date for performing the demand prediction The predicted number of demand for each used part on the forecast target day is calculated by multiplying the ratio with the part usage ratio for each temperature range in the temperature range to which the predicted temperature belongs.

  According to this structure, based on the past temperature data, a demand prediction can be performed in consideration of the tendency of the number of used parts for each used part. For this reason, it is possible to make a more accurate prediction than when simply making a monthly demand prediction.

  As an aspect of the above configuration, for example, the temperature range may be 1 ° C., and a value obtained by rounding off the decimal point of the temperature indicated by the temperature data may be set as the “temperature zone”. Another example is to change the temperature range according to the temperature value, narrowing the temperature range in the vicinity where the temperature sensitivity is higher, and conversely reducing the temperature range in the vicinity where the temperature sensitivity is lower. It may be widened. In this way, it is possible to accurately evaluate the temperature sensitivity while increasing the number of data belonging to the same temperature range.

  In addition, when deriving the component usage ratio by temperature range, in addition to the method of deriving based on the temperature information on the day of the repair period, the temperature information from the date of repair to the predetermined date or the average over several days from the predetermined date A method using temperature can also be used. At this time, regarding how many days before the date of repair, the component usage ratio by temperature range is derived as a reference, the device type component demand forecasting unit stores in advance as the method for deriving the component usage ratio by temperature range. What should be. Further, a different value may be used for the point of determination based on the temperature before the repair date based on the date (month, season) of the repair time.

  In addition to the above features, the present invention preferably further includes the following points.

That is, the device type-specific parts demand forecasting unit
Based on the part usage record data, the number of used parts for each used part is tabulated by weekday / holiday indicated by the repair period, and further, the number of used parts is divided by the number of weekdays / holiday that has been tabulated. By deriving the ratio, the usage ratio by weekday holiday is calculated,
A correction is performed on the forecast demand number for each used part for the forecast target day in consideration of the weekday / holiday use ratio according to whether the forecast target day is a weekday or a holiday.

  According to the above configuration, in addition to the influence of the temperature, the demand for parts can be predicted in consideration of the difference in parts replacement frequency according to weekdays and holidays.

  In addition to the above features, the present invention preferably includes the following points.

That is, the customer-owned equipment data includes information about a customer area to which the owned customer belongs,
The repair work report data includes information on a customer area to which the repair customer belongs,
The existing number calculation unit by the device type is:
Within the target customer area, which is the target customer area for which demand is to be predicted, the number of devices currently owned by the owned customer among the owned devices is the period from the start of use to the present. By counting for each device type and the actual usage period defined by a multiple of the unit period, the number of existing devices by the actual usage period by the device type of the owned device is calculated,
The part demand forecasting unit by device type is
For the same used parts, multiply the corresponding specific usage period and the actual usage period and the corresponding equipment type by the part usage rate and the number of existing equipment by the actual usage period in the target customer area. The calculated demand for each of the used parts in the target customer area over the unit period is derived by adding the calculated results for each of the same used parts.

  According to the system having the above characteristics, it is possible to predict the demand for parts for each customer area. For this reason, in the case where there are a plurality of repair shops having different areas in charge of repair, a system useful for each repair shop to properly manage the inventory of parts can be realized.

Furthermore, in addition to this configuration, it further includes a weather data storage unit that records weather data including at least past temperature data over a plurality of years,
The part demand forecasting unit by device type is
While reading out the weather data from the weather data storage unit and comparing it with the repair time and the customer area of the part usage record data, the number of used parts for each used part for each temperature range constituted by a predetermined temperature range To calculate the component usage ratio by temperature range, and to derive these ratios,
Further, an average temperature over the unit period in the target customer area is derived based on the weather data, a reference temperature range to which the average temperature belongs is recognized, and the used parts in the target customer area over the unit period Is divided by the number of days in the unit period to calculate the number of predicted demand for each used part on the day belonging to the reference temperature range, and this value is the temperature-specific part in the reference temperature range. Multiplication by the ratio between the usage ratio and the component usage ratio for each temperature range in the temperature range to which the predicted temperature of the prediction target day for which demand prediction is to be performed, makes a prediction for each used component for the prediction target date in the target customer area It is characterized by calculating the number of demands.

  According to the system having the above characteristics, it is possible to perform demand prediction for each customer area after taking into account the tendency of the number of used parts for each used part based on past temperature data.

  In addition to the above features, the present invention preferably includes the following points.

That is, the customer-owned device data storage unit, the repair work report data storage unit, the component usage record data storage unit, the component usage record data creation unit, the component usage rate calculation unit, the existing number calculation unit by device type, And a server computer comprising the device type-specific parts demand forecasting unit,
A plurality of operation terminals electrically connected to the server computer,
The operation terminal belongs to a repair company in charge of the different customer area, and periodically transmits the repair work report data including information on the customer area in charge to the server computer, and the data is transmitted to the server computer. It is configured to record in the repair work report data storage unit,
When a demand prediction instruction is given from the operation terminal to the server computer, the predicted demand number in the customer area in charge of the repair shop to which the operation terminal belongs is calculated, and the result is transmitted to the operation terminal. It is characterized by being.

  According to the system having the above-described feature, each repairer transmits repair work direction data to the server computer via the electric communication line, so that the repair work report data of a plurality of repair shops is centrally managed by the server computer. be able to. As a result, since the number of registered data is dramatically increased in the server computer as compared with information management in a single repair company, many past repair records are recorded. The component usage rate is derived based on this data. The component usage rate derived in this way more reflects the relationship between the period during which the device is used and the replaced component (here, in the form of the relationship between the specific usage period and the replaced component).

  Therefore, by multiplying this part usage rate by the number of owned equipment that currently exists in each customer area by the actual usage period, and adding up each same used part, the relationship between the actual usage period of the equipment and parts replacement Thus, the demand for parts at each repair shop can be predicted.

Further, the present invention is a parts demand forecasting system for forecasting demands for parts required for equipment repair,
A customer-owned equipment data storage unit, a repair work report data storage unit, a component usage rate calculation unit, an existing number calculation unit by device type, and a part demand prediction unit by device type,
The customer-owned equipment data storage unit is
Data relating to owned equipment that is currently owned by the owning customer or previously owned, including at least information on the owning customer, information on the type of the owned equipment, and information on when the equipment was first used. If you have owned the device in the past, we record customer-owned device data including information on when to end use,
The repair work report data storage unit is
Data on the contents of repairs that have been performed in the past, at least information on the repair customer who performed the repair, information on the model of the repair device that performed the repair, information on the parts used in the repair, and the repair Repair work report data including information on the repair time and information on the start of use of the equipment.
The component usage rate calculation unit
The number of actual use of all the owned devices registered in the customer-owned device data storage unit is the period in which the user has experienced use and is defined by a multiple of the unit period, and for each device type To calculate the number of devices by device type and experience period,
Based on the part usage record data, the number of the repair devices in which the used parts are actually used, for each specific usage period in which the period from the use start time to the repair time is defined as a multiple of a predetermined unit period And for each device type and counting by each used component, by dividing this by the number of corresponding device type by experience period, the ratio of the number of used devices required among all the owned devices A component usage rate which is a value defined for each specific usage period and for each device type and for each used component,
The existing number calculation unit by the device type is:
Based on the customer-owned device data, the number of devices that are currently owned by the owned customer among the owned devices is defined as a multiple of the unit period from the start of use to the present. By counting for each actual use period and for each device type, the number of existing devices by the actual use period for each device type is calculated,
The device-specific parts demand forecasting unit
With respect to the same used parts, the calculation result obtained by multiplying the corresponding specific usage period and the actual usage period and the corresponding device type by the number of parts used and the existing number by the actual usage period by the equipment type is the same. The number of predicted demands for each used part over the unit period is derived by adding up each used part.

  According to the above feature of the present system, the specific use period, which is the period from the use start time to the repair time, can be grasped only from the repair work report data. For this reason, in order to recognize the specific usage period, it is possible to omit the process of creating the part usage record data by associating the repair work report data with the customer-owned device data.

Further, the present invention is a parts demand forecasting method for forecasting demands for parts required for repair of equipment by computer data processing,
Data relating to owned equipment that is currently owned by the owning customer or previously owned, including at least information on the owning customer, information on the type of the owned equipment, and information on when the equipment was first used. If the equipment has been owned in the past, record customer-owned equipment data that further includes information on the end of use, and data on the details of repairs that were performed in the past, including at least information on repair customers who performed repairs. Record repair work report data including information on the type of repair equipment that performed the repair, information on the parts used in the repair, and information on the timing of the repair.
Information of the owned customer and the type of the owned equipment indicated in the recorded customer owned equipment data, and information of the repair customer and the type of the repair equipment indicated in the recorded repair work report data By associating the above information, the component usage record includes information on the device type, the used parts used for repair, and the specific use period that specifies the period from the use start time to the repair time as a multiple of a predetermined unit period. Create and record data,
Counts the number of devices actually used by all the owned devices in which the customer-owned device data is recorded, for each experience period that is a period of experience of use and defined by a multiple of the unit period, and for each device type. To calculate the number of devices by device type and experience period,
Based on the component usage record data, the number of the repair devices in which the used components are actually used is counted for each specific usage period and for each device type, and for each used component, By dividing by the number of devices by experience type by device type, the ratio of the number of used parts required for all the owned devices was defined for each specific use period and for each device type and for each used component. Calculate the component usage rate that is the value,
Based on the customer-owned device data, the number of devices that are currently owned by the owned customer among the owned devices is defined as a multiple of the unit period from the start of use to the present. By counting for each actual use period and for each device type, the number of existing devices by the actual use period by the device type of the owned equipment is calculated,
With respect to the same used parts, the calculation result obtained by multiplying the corresponding specific usage period and the actual usage period and the corresponding device type by the number of parts used and the existing number by the actual usage period by the equipment type is the same. The number of predicted demands for each used part over the unit period is derived by adding up each used part.

  In addition to the above features, the present invention preferably further comprises the following points.

That is, the customer-owned equipment data includes information on a customer area to which the owned customer belongs, and the repair work report data includes information on a customer area to which the repair customer belongs,
Within the target customer area, which is the target customer area for which demand is to be predicted, the number of devices currently owned by the owned customer among the owned devices is the period from the start of use to the present. By counting for each device type and the actual usage period defined by a multiple of the unit period, the number of existing devices by the actual usage period by the device type of the owned device is calculated,
For the same used parts, multiply the corresponding specific usage period and the actual usage period and the corresponding equipment type by the part usage rate and the number of existing equipment by the actual usage period in the target customer area. The calculated demand for each of the used parts in the target customer area over the unit period is derived by adding the calculated results for each of the same used parts.

  In addition to the above features, the present invention preferably further comprises the following points.

That is, the weather data including at least past temperature data over a plurality of years is read out, and the use is performed for each temperature range constituted by a predetermined temperature range by collating with the repair time and the customer area of the part use record data. By calculating the number of parts used for each part and deriving these ratios, the part usage ratio by temperature range is calculated.
Deriving an average temperature over the unit period in the target customer area based on the weather data, recognizing a reference temperature zone to which the average temperature belongs, and prediction for each used part in the target customer area over the unit period By dividing the number of demands by the number of days in the unit period, the predicted demand number for each used part on the day belonging to the reference temperature range is calculated, and this value is used as the component usage ratio for each temperature range in the reference temperature range. And the ratio of the component usage ratio for each temperature range in the temperature range to which the predicted temperature of the prediction target day for which the demand prediction is performed is multiplied by the number of predicted demands for each used component in the target target area in the target customer area Is calculated.

  According to the configuration of the present invention, the demand for parts can be predicted based on past repair results. More specifically, taking into account the period of actual use of the equipment until the replacement of the part (specific use period) and the tendency of the replaced part, reflect the period of actual use of the existing equipment, Demand for parts can be predicted. This can be a useful system for forecasting parts demand for devices that tend to produce differences in the types of parts that cause problems, particularly depending on the period of use.

Network configuration diagram for explaining a parts demand forecasting system according to the present invention Example of repair work report data Conceptual block diagram showing the configuration of a server computer provided in the data center Example of customer-owned equipment data Example of parts usage record data Graph for explaining the component usage rate The conceptual block diagram which shows the structure of the server computer in 3rd embodiment.

  Embodiments of a component demand prediction method and a component demand prediction system according to the present invention will be described.

  As an application example of this system, it is assumed that a plurality of repair shops make a repair visit of equipment to a customer in a customer area in which each repair shop is responsible. In the following, a case where an employee of a gas company service shop makes a repair visit to each customer's gas equipment will be described as an example. Of course, the present system is not specific to a repair visit of a gas appliance, and can be applied to a similar configuration.

[First embodiment]
FIG. 1 is a network configuration diagram for explaining the present system. A plurality of repair companies (M0, M1,..., MN) that perform on-site repair of equipment are in charge of visiting repairs in their respective customer areas (A0, A1,..., AN).

  First, the process until the repair shop makes a repair visit to the customer will be described.

  When the customer recognizes the malfunction of the gas equipment, the customer contacts the reception of the gas company or the repair company in charge of his area. The contact information of the gas company reception or the repairer in charge of the area is usually attached to the gas equipment and its surroundings, and can be recognized by the customer.

  When the gas company receptionist or the repairer in charge of his / her area receives the contact from the customer, he / she asks for information such as which device is causing what kind of symptoms and when it has occurred. .

  When the contact from the customer is made at the reception of the gas company, the gas company searches for a repair agent in charge of the customer's residence area, and contacts the supplier with information on defects obtained from the customer. Further, when the customer directly contacts the repair company, the repair company can naturally know the information on the occurrence of the malfunction.

  When a customer in his / her area in charge recognizes information on a malfunction of a device, the repairer carries parts expected to be replaced from the symptoms of the device and visits the customer. In some cases, after the worker directly confirms the symptom of the equipment, he / she goes back to pick up the necessary parts and visits the repair again.

  Parts necessary for equipment repair are purchased from parts manufacturers by each repair company and stored in each repair shop. When the inventory runs short, the repair shop places an order with the parts manufacturer as needed. In the case of gas appliances, since there are tens of thousands of household gas appliances sold by gas companies so far, there are a considerable number of types of parts attached to the appliances.

  Of course, there are many devices that have been discontinued at present, but as there are still customers who use such devices, the inventory of parts is limited for a certain period after production is discontinued. Needless to say, it is necessary to secure it.

  For this reason, each repair company does not store all parts to a certain extent, but strives to secure only the necessary amount according to the demand for parts. At that time, the demand for parts is predicted using the system described below, and inventory parts are managed based on the result.

  When the visitor repair such as parts replacement is completed at the customer site, the worker registers the work content using the operation terminal. Each repairer has an operation terminal, and this operation terminal is connected to a server computer of the data center DCR via a telecommunication line NW. As the operation terminal, a portable handy terminal, a mobile phone or the like may be used in addition to a normal personal computer.

  Information on the work content input by the worker and other input information are sent to a server computer in the data center DCR, and the information is accumulated. Conversely, information stored in the data center DCR can be read by the operation terminal of each repair company.

  The operator operates the operation terminal and visits customer information, the model of the defective device, symptoms, actions taken, parts used, repair costs, etc. (hereinafter referred to as “repair work report data”) ). The information to be input is coded in advance, and the worker selects and inputs the corresponding code. The inputted information is transmitted via the telecommunication line NW and stored in the data center DCR.

  FIG. 2 is an example of repair work report data. In FIG. 2, in addition to the items described above, the reception No. It also includes information about the reception date and reception requirements. In addition, these information is good also as a structure which diverts the data numbered in the stage of repair reception as it is.

  Each repair company (M0, M1,..., MN) registers similar information using the operation terminal after the repair is completed. The data center DCR side can naturally recognize from which repairer the information is registered.

  FIG. 3 is a conceptual block diagram illustrating a configuration of a server computer included in the data center DCR. The server computer includes a data input / output unit 11, a component usage result data creation unit 13, a component usage rate calculation unit 14, a device type / area-based existing number calculation unit 16, a device type / area-specific component prediction unit 17, a repair work report data. A storage unit 21, a customer-owned device data storage unit 22, a component usage record data storage unit 23, a component usage rate storage unit 24, and a control unit 31 for controlling the whole are provided.

  The data input / output unit 11 is an interface that performs transmission / reception of data and necessary processing associated therewith. The component usage record data creation unit 13, the component usage rate calculation unit 14, the existing number calculation unit 16 for each device type / area, and the component prediction unit 17 for each device type / area include software means and hardware for performing predetermined calculation processing, respectively. And a storage means for temporarily storing data as needed. The repair work report data storage unit 21, customer-owned equipment data storage unit 22, and component usage record data storage unit 23 are storage units for recording various types of information, such as writing and reading based on instructions from the control unit 31. Processing is performed.

  The repair work report data transmitted from the repair company (M0, M1,..., MN) through the telecommunication line NW is stored in the repair work report data storage unit 21 after the data input / output unit 11 acquires information. The repair work report data storage unit 21 stores past repair work report data of each repair shop.

  Registered in the customer-owned device data storage unit 22 is data of owned devices currently owned by each customer or previously owned (hereinafter referred to as “customer-owned device data”). FIG. 4 shows an example of customer-owned device data. Each table has a customer area, a customer No. , Owned equipment model No. In addition, a start date when each device is used and an erasure date when it is already removed are registered.

  For example, according to the data table in the first row of FIG. The customer of “A112553” has the owned equipment model No. It has been registered that a device of the type corresponding to “10010595” has been used since February 10, 1990, and that the device has been deleted on January 15, 2005. Further, according to the data table in the second row of the figure, the customer has the owned equipment model No. from January 15, 2005 to the present. It is registered that a device of a type corresponding to “12400210” is used. The item “Erase Date” indicates “E”, which indicates that the item has not been erased at this time.

  The component usage record data creation unit 13 reads out the repair work report data from the repair work report data storage unit 21 and the customer owned device data from the customer owned device data storage unit 22, respectively. Is associated with the aging information of the device. Data created by this association (hereinafter referred to as “part use record data”) is recorded in the part use record data storage unit 23.

  FIG. 5 is an example of part usage record data. In each table, in addition to each item of repair work report data, the number of days and years elapsed from the start of use of the device to the date of repair are shown. It is registered.

  The component usage record data in the data table on the first line in FIG. 5 will be described as an example. Customer No. shown in the first line of FIG. The customer of “A112553” received the model No. Processing corresponding to the action code “001” is performed on the type of equipment corresponding to “12400210” using the part “YA021302”.

  Here, referring to the data table in the second row in FIG. 4 (customer-owned device data), the customer has started using the device from January 15, 2005. Based on this use start date, the number of days until the part replacement (December 15, 2008) is 1430 days, which is 3 years (rounded down to the nearest decimal place). In the data table in the first row of FIG. 5, “1430” is added to the “Elapsed Days” item and “03” is added to the “Elapsed Years” item.

  This period (1430 days) is a period elapsed from the start of use until the part is replaced. In FIG. 5, in this embodiment, the unit period is defined as “one year”, and the period elapsed from the start of use to the replacement of the part is defined as “elapsed years” in a multiple of one year. This “elapsed years” corresponds to a “specific use period” defined by a multiple of the unit period.

  The component usage record data creation unit 13 reads the repair work report data newly registered in the repair work report data storage unit 21 every day, and associates it with the customer owned device data in the customer owned device data storage unit 22. Component usage record data as shown in FIG. For example, the creation unit 13 may create new component usage record data based on the repair work report data registered in the repair work report data storage unit 21 on the previous day at a time determined every day. .

  Thus, the component usage record data storage unit 23 stores the component usage record data up to the latest. Therefore, it is possible to recognize information indicating which component is used for each specific usage period and for each device type with respect to all component usage record data.

  The component usage rate calculation unit 14 calculates the component usage rate for each device type according to the following procedure. Here, the component usage rate refers to a certain device type, and out of the number of devices that this type of device has actually been used by customers so far, the proportion of devices in which new parts were used at the time of repair Is a value calculated for each specific use period and for each used component. This point will be further described with reference to the graph of FIG.

  6 shows changes in the component usage rate of a certain device (model No. 12400210) in time series with the specific usage period (year) as the horizontal axis. As an example, this graph shows model No. This is a graph based on all the past results of replacement of parts constituting the device of the type 12400210, focusing on the device of the type 12400210 (hereinafter referred to as “model 12400210”).

  Hereinafter, the method of calculating the component usage rate will be described by explaining the graphing method shown in FIG.

  First, the component usage rate calculation unit 14 stores the owned device type No. from the customer-owned device data storage unit 22. As a result, all customer-owned device data with “12400210” is read out, and the number of devices of type 12400210 that have been used so far and are currently used is ascertained for each period of use. To do.

  Here, the “period of experience of use” will be described. For example, in the case of a device in which 3.5 years have passed since the start of use, naturally, the device has experienced all use within a period of 3.5 years.

  The “experience period” defines the “period of experience of use” as a multiple of the unit period, and the number of the devices is counted for each experience period.

  In this embodiment, since the unit period is set to “1 year”, it is assumed that the experience period is defined by an integer value obtained by dividing 3.5 years by the unit period “1 year” and truncating after the decimal point. One unit will be counted for each year of 0 years, 1 year, 2 years, and 3 years of experience. In this example, needless to say, the experience period of 0 years means that the device was actually used for 0 to 1 year, and similarly, the experience period of 1 year is used for 1 to 2 years. The experience period of 2 years indicates use over 2 to 3 years, and the experience period of 3 years indicates use over 3 to 4 years.

  In this way, for all the devices of type 12400210 registered in the customer-owned device data storage unit 22, the number is counted for each experience period of each owned device and calculated. Hereinafter, the number calculated in this way is referred to as “number of devices by experience type by device type”.

  Next, from the component usage record data storage unit 23, the model No. All component usage record data with “12400210” is read out. Then, the read component usage record data of the device 12400210 is counted for each specific usage period and for each used component. Thereby, the past number of parts replacement of the device of model 12400210 can be calculated for each specific use period and for each used part.

  Then, the number of component replacements calculated for each specific use period and for each used component is divided by the number of devices by experience type in the type 12400210 that has already been calculated. Accordingly, the component usage rate is calculated for each specific usage period and for each used component.

  FIG. 6 shows the component usage rate calculated in this manner for each specific usage period and for each used component, with the horizontal axis indicating the specific usage period (year) and the values for each used component being stacked on the vertical axis. It is a thing. In FIG. 6, the value on the vertical axis is displayed as a ratio of a constant a, thereby providing a relative graph display. In addition, descriptions such as “FE066536” and “FC022355” are reference numerals of component parts of the model 12400210.

  When the system is actually used for forecasting parts demand, the part usage rate calculation unit 14 may calculate the part usage rate for each specific usage period of the target device and for each used part. There is no need to create such a graph for display. FIG. 6 is displayed for explaining the component usage rate. Of course, you may have the function to create such a graph.

  When the component usage rate calculation unit 14 calculates the component usage rate for each specific usage period and for each used component for each device type, the component usage rate calculation unit 14 records the data in the component usage rate storage unit 24. This information is updated as needed every predetermined period.

  The component usage rate calculated by such a method is determined by three items: the type of the device to be calculated, the type of the component constituting the device, and the specific usage period of the device. In other words, the component usage rate r calculated by the component usage rate calculation unit 14 is expressed as a function of the device type X, the component i, and the specific usage period y. The component usage rate r can be defined as the probability that the device of type X in the state where the period y has elapsed since the start of use will replace the component i. Hereinafter, the component usage rate of the component i of the device of the type X whose year (specific use period) from the start of use is y years is described as r (X, i, y).

  Assuming that the component usage rate is calculated for each device type (X) for each specific usage period (y) and for each used component (i) by the above method, In contrast, the number of operations required for calculation is large. Therefore, the component usage rate storage unit 24 does not need to be updated daily, unlike the component usage record data storage unit 23, and has a larger span than the component use record data storage unit 23 (for example, in units of one month, three months). The data may be updated on a unit basis.

  In the following, a case where a repairer operates his / her own operation terminal to predict parts demand under such a configuration will be described. Here, a case where the repairer M0 predicts parts demand will be described as an example.

  When the repairer M0 operates the operation terminal and inputs information instructing prediction of parts demand, this information is sent to the server computer in the data center DCR. The server computer DCR acquires this information from the data input / output unit 11 and recognizes that an instruction for parts demand prediction has been given from the repairer M0.

  Next, the device type / area-existing number-of-units calculation unit 16 reads out customer-owned device data corresponding to the customer area A0 that the repairer M0 is in charge of from the customer-owned device data storage unit 22, and all currently used The number of owned devices is extracted by device type.

  Further, the number of the extracted device types is counted for each period from the use start time to the present time. At this time, this period is counted for each “actual use period” defined by a multiple of the unit period. In this embodiment, since the unit period is one year, the actual use period corresponds to the period from the use start time to the present expressed in years.

  As a result, for all devices existing in the customer area A0, the number of existing devices by device type and actual usage period is calculated. The number calculated in this way is referred to as “existing number according to actual use period by device type” in the customer area A0.

  Here, as a method of calculating the actual usage period, simply count the number of days from the use start date attached to the customer-owned equipment data, and divide by 365. Then, it is possible to adopt a value derived (truncated after the decimal point).

  For example, when the extraction process is performed on February 13, 2009, a device of the type 12400210 described in the data table on the second row in FIG. 4 is counted as four years. As a result, one device is counted as a device of type 12400210 with an actual usage period of “4 years”. In the same manner, the number of all devices existing in the customer area A0 is measured for each device type / actual use period, and the number of existing devices for each device type in the customer area A0 is calculated.

  Note that the device of type 10010595 described in the data table in the first row in the figure is currently deleted, so it is not counted in the number of existing devices by device type and actual usage period.

  In this way, when the number of existing devices by device type and actual use period is calculated for all devices X existing in the customer area A0, the device type / area-specific component prediction unit 17 reads from the component usage rate storage unit 24. The corresponding component usage rate is calculated and multiplied.

  More specifically, it is as follows.

  The existing number for each actual usage period calculated by the device type / area-based existing number calculation unit 16 is determined by three items: the customer area to be calculated, the device type, and the actual usage period of the device. In other words, the number of existing machines W by use period by device type is represented by a function of customer area A, device type X, and actual use period (years) y. In the following, the existing number of the device type X in the customer area A with the actual usage year y is described as W (A, X, y).

  On the other hand, as described above, the component usage rate storage unit 24 stores, in the component usage rate calculation unit 14, the component i of the device type X that has passed the number of years (specific usage period) y years elapsed from the start of use. The component usage rate r (X, i, y) is calculated for each X, each i, and each y.

  For example, when A = A0, X = 12400210, the existing number of devices of type 12400210 in the customer area A0 is determined by the device type / area existing number calculation unit 16 for each specific usage period (and as described above). (Equivalent to the number of years that have passed since the start of use). That is, in the same area, the device of type 12400210 has a specific use period of 0 years, W (A0,12400210,0) units, a specific use period of 1 year has W (A0,12400210,1) units,. Those with a specific usage period of y years are extracted as W (A0,12400210, y) units.

  Further, when X = 12400210 and i = YA021302, the component usage rate calculation unit 14 extracts the component usage rate of the part YA021302 of the device of the model 12400210 for each specific usage period. That is, among the devices of type 12400210 with a specified usage period of 0 years, the component usage rate that newly uses the component YA021302 is r (12400210, YA021302,0), and the same device with a specified usage period of 1 year newly uses the same component. The component usage rate to be used is r (12400210, YA021302,1), ..., and the component usage rate at which the device with the specified usage period y years newly uses the same component is extracted as r (12400210, YA021302, y).

  Therefore, by multiplying W (A0,12400210,0) and r (12400210, YA021302,0), the number of parts YA021302 that are newly used among the devices of the model 12400210 in the customer area A0 with the specified usage period of 0 years. Is calculated. If this value is described as V (A0,12400210, YA021302,0), similarly, it is specified in customer area A0 by multiplying W (A0,12400210,1) and r (12400210, YA021302,1). The number V (A0,12400210, YA021302,1) that is expected to newly use the part YA021302 among the devices of the model 12400210 with a usage period of one year is calculated.

  In this way, V (A0,12400210, YA021302, y) is calculated by multiplying W (A0,12400210, y) and r (12400210, YA021302, y) for all specified usage periods y years. By adding up, it is possible to calculate the number V (A0,12400210, YA021302) that is expected to newly use the part YA021302 among the devices of the model 12400210 in the customer area A0. This V (A0,12400210, YA021302) indicates the expected number of equipment 12400210 that will newly use the part YA021302 in the customer area A0 in this year, in other words, the demand for the part YA021302 for equipment of the model 12400210. It is.

  Thereafter, by performing the same calculation while changing the component i in the same device, the number of units V (A0, 12400210,...) That are expected to newly use the component i among the devices of the model 12400210 in the customer area A0. i) is calculated. Further, the same calculation is performed while changing the model X. By performing such calculation, the number V (A0, X, i) of devices of type X that are predicted to newly use the component i in the customer area A0 is calculated. Hereinafter, this value is referred to as “the predicted number of demand by component by device type”.

  Some components are used across different types of equipment. For this reason, the predicted number of demands for each device type calculated as described above is further aggregated for each part to derive the predicted number of demand for each part. For example, it is assumed that the part YA021302 is used in common for the devices of the type 12400210 and the type 12400211, and is not used for other types of devices. According to the above method, V (A0,12400210, YA021302) and V (A0,12400211, YA021302) are calculated separately. By adding these, the prediction of the part YA021302 in the customer area A0 is performed. V (A0, YA021302) is derived as the demand number. Hereinafter, similarly, the predicted demand number V (A0, i) for each part in the customer area A0 is derived by aggregating the predicted demand number for each part.

  The predicted demand number V (A0, i) derived for every part i in this way is transmitted from the data input / output unit 11 to the operation terminal of the repairer M0 via the telecommunication line NW. The repairer M0 can recognize the predicted demand quantity for each part by using its own operation terminal, and use it to determine whether or not to order a new part by comparing with the existing inventory quantity for each part. Can do.

  According to the configuration of the present embodiment, each repair shop can recognize the predicted demand number of parts calculated according to the years of use of the equipment existing in the customer area in charge of itself, so that it is more than conventional. Highly accurate demand prediction is possible. In this regard, the accuracy of demand prediction is greatly increased when compared with a method of calculating the component usage rate by dividing the number of devices whose components have been replaced by the number of devices sold.

  In particular, in equipment that is constantly used every day, such as gas equipment and power equipment, it may be assumed that some parts of the parts that cause problems with the passage of time will exhibit some characteristics. This system, which can derive the expected number of demanded parts, is extremely useful in this respect.

  This system exhibits high utility value in an environment where a considerable amount of various data such as the start date of use of each device and the date of replacement of parts is stored. From this point of view, in the field of gas equipment and electric power equipment, where gas equipment and electric power companies collectively manage equipment, there are operations that constantly store the various data. It is a system with extremely high utility value where data is accumulated.

  In the above description, the repairer M0 in charge of the customer area A0 has been described as an example, but the other repairers M1, M2,... Another predicted demand number V (A, i) can be recognized. This also applies to the following embodiments.

[Second Embodiment]
In addition to the first embodiment, this embodiment has a function of performing demand prediction for each more detailed period. Since the system configuration itself is the same as that of the first embodiment, the description thereof is omitted.

  In the first embodiment, the predicted demand number V (A, i) for each part i in the customer area A calculated by the equipment type / area-specific part demand forecasting unit 17 is calculated for one year including the day of calculation. Corresponds to the number of parts demand over time. The present embodiment is configured to have a function of finely disassembling this parts demand by month or by weekdays and holidays.

  When the component usage record data creation unit 13 creates the component usage record data, in addition to the items shown in FIG. 5, the repair date and the weekday holiday indicated in the repair work report data are added. Then, component usage record data is created and registered in the component usage record data storage unit 23.

  When the device type / area-specific component demand forecasting unit 17 calculates the predicted demand number V (A, i) for each component i in the customer area A, as in the first embodiment, the component usage result data storage unit next. 23, all the component usage record data with the component i added as an item related to the used component are read, and the number of uses is disassembled and grasped by month. Then, the monthly usage ratio of the part i from January to December is calculated.

  Then, the device type / area-specific component demand forecasting unit 17 proportionally distributes the predicted demand number V (A, i) at the monthly usage ratio calculated earlier. Thereby, the predicted demand number V (A, i, M) of the part i in the customer area A every month (M) can be calculated (M = 1 to 12).

  The monthly predicted demand number V (A, i, M) derived for every part i in this way is transmitted from the data input / output unit 11 to the operation terminal of the repairer via the telecommunication line NW. . As a result, the repair shop can recognize the predicted number of demands for each part for each part at its own operation terminal. Thereby, it is possible to recognize more detailed demands for parts than in the first embodiment.

  In particular, among gas equipment and power equipment, equipment whose operating conditions are affected by the outside temperature, such as air-conditioning equipment and heating equipment, which are considered relatively likely to be repaired by gas companies and power companies. In the case of these devices, the mode of use varies greatly depending on the seasons. For this reason, it is also conceivable that a malfunction of the device is induced due to the change in the usage mode. According to the configuration of the present embodiment, by performing monthly demand prediction based on past results, it is possible to perform detailed demand prediction in consideration of the influence of outside air temperature by a simple method.

  Moreover, it can also be set as the structure provided with the function which decomposes | disassembles the said monthly forecast demand number V (A, i, M) according to a weekday and a holiday further.

  The device type / area-specific component demand forecasting unit 17 reads all the component usage record data to which the component i is added as an item related to the used component from the component usage record data storage unit 23, and calculates the number of uses by weekdays and holidays. Disassemble and grasp. Then, by calculating the ratio between the average number of uses on weekdays (a) and the average number of uses on holidays (b), the ratio of the number of uses on weekdays and holidays (εia: εib) is calculated for the part i. Note that (εia + εib) / 2 = 1. Thus, εia and εib correspond to the weekday usage rate ratio and holiday usage rate ratio, respectively, with respect to the average value.

  Then, the device type / area-specific parts demand forecasting unit 17 calculates the forecast demand number V (A, i, M), and then the number of weekdays per month (M) × da weekday usage ratio (εia). And the ratio of the number of holidays (db) × the number of holidays used (εib). Thereby, the predicted demand number V (A, i, M) is decomposed into the total predicted demand number on weekdays (a) of the month (M) and the total predicted demand number on holidays (b).

  Then, by dividing the calculated total number of demands on weekdays by the number of weekdays (da) in the month (M), the number of forecast demands V (A, i, M, a) is calculated. Similarly, the predicted demand number V (A, i, M, b) per day on weekdays is calculated by dividing the total predicted demand number for holidays by the number of weekdays (db) in the month (M). The

  By using the predicted demand numbers V (A, i, M, a) and V (A, i, M, b) calculated in this way, the presence / absence of holidays is taken into account when making a demand forecast for a certain week. Predicted demand can be made. For example, assuming that the number of weekdays is pa and the number of holidays including holidays is pb in a certain week, the predicted demand number of part i in area A in this week is pa × V (A, i, M, a ) + Pb × V (A, i, M, b). As a result, it is possible to predict demand for parts with a finer span than monthly.

[Third embodiment]
As in the second embodiment, this embodiment has a function of performing demand prediction for each more detailed period in addition to the first embodiment, and in particular, has a function of calculating a predicted demand number based on temperature information. is there.

  FIG. 7 is a conceptual block diagram showing the configuration of the server computer of this embodiment. In the present embodiment, a weather data storage unit 26 is provided in addition to the configurations of the first and second embodiments (FIG. 3). The weather data storage unit 26 is a storage unit that records various types of information, and performs processing such as writing and reading based on instructions from the control unit 31.

  In the weather data storage unit 26, weather data from the date closest to the oldest date to the present is registered among the use start dates registered in the customer-owned device data storage unit 22. As registered data, at least the temperature data of the day is registered for each day / customer area. Other than that, various weather data such as precipitation, wind speed, wind power, humidity, and solar radiation are included as appropriate. May be.

  Furthermore, information on which temperature range the temperature data belongs to is also registered in the weather data. For example, the temperature range is defined by an integer value obtained by rounding off the value indicated by the temperature data. That is, in the weather data storage unit 26, which temperature zone each day belongs to is registered in advance for each customer area.

  When the device / area-specific component demand prediction unit 17 calculates the predicted demand number V (A, i) for each component i in the customer area A, as in the first embodiment, next, the component usage record data storage unit 23. As a result, all the component usage record data to which the component i is attached as an item related to the used component is read out, and the weather data is read out from the weather data storage unit 26 to recognize the temperature data on the “acceptance date” of the component usage record data. . Then, by counting the number of parts usage record data for each temperature zone to which the temperature data belongs, the number of parts used is decomposed and grasped for each temperature zone. Furthermore, the ratio of the number of parts used for each temperature range is calculated.

  Further, the device type / area-specific part demand forecasting unit 17 reads the weather data from the weather data storage unit 26, derives the average temperature in the customer area A for a predetermined period (for example, the most recent year), and determines the temperature range to which the average temperature belongs. To grasp.

  As a result, the component type / area-specific component demand prediction unit 17 uses the component usage ratio s for each temperature zone (T) when the component i in the customer area A is based on the temperature zone Ta at the average temperature in the area. (A, i, T) can be calculated.

  Next, the device type / area-specific part demand forecasting unit 17 divides the forecast demand number V (A, i) by the total number of days in a year to obtain the average forecast demand number Va (A, i) for each day. Is calculated. Then, this Va (A, i) is recognized as the predicted demand number of the part i for one day under the temperature zone Ta to which the average temperature in the area belongs.

  Thus, by obtaining the average predicted demand number Va (A, i) for each day and the component usage ratio s (A, i, T) for each temperature range, the temperature data for the day on which the demand is to be predicted can be obtained. The predicted demand number V on the same day can be calculated as V = Va (A, i) · s (A, i, tc) using the indicated temperature zone tc. By performing such calculation for each component i, it is possible to perform a predicted demand for all the components.

  For example, when data on the predicted temperature from the calculation date to one week later is recorded in the weather data storage unit 26, the device type / area-specific component demand prediction unit 17 stores the predicted temperature data for each day. The predicted demand number V is calculated for each day using the indicated temperature zone tc, and is transmitted from the data input / output unit 11 to the operation terminal of the repairer M0 via the telecommunication line NW. The repairer M0 can recognize the predicted number of demand for each part at his / her operation terminal, and use it to determine whether or not to order a new part by comparing with the existing number of parts in stock by part. Can do.

  In addition, when the data regarding the predicted temperature is not recorded in the meteorological data storage unit 26, information regarding the predicted temperature on the day on which the demand prediction is desired can be input from the operation terminal of the repairer. Is preferred.

  The predicted number of demands obtained in this way has been made in consideration of temperature data. Therefore, particularly when making demand for parts of gas equipment and power equipment that affect the usage status of equipment according to temperature. Is useful.

  Furthermore, it is good also as a structure which correct | amends according to a weekday and a holiday for the estimated demand number calculated | required in this way.

  Specifically, the component type / area-specific component demand forecasting unit 17 adds the component i as an item related to the used component from the component usage record data storage unit 23 so as to calculate the usage ratio of the component i for each temperature range. When all the component usage record data is read out, the average number of weekdays and the average number of holidays used for the component i are derived based on this data. Further, this average usage number (that is, this corresponds to the annual average usage number) is calculated, and the usage ratio of the weekday usage number and the holiday usage number when the annual average usage number is used as a reference is calculated.

  Then, the predicted demand number V predicted on the basis of the predicted temperature data by the above-described method is multiplied by the weekday usage ratio ξd if the same day is a weekday, and by the holiday usage ratio ξh if it is a holiday. As a result, it is possible to make a prediction in consideration of the difference in usage frequency for each weekday and holiday.

  Further, for weather data other than the temperature data registered in the weather data storage unit 26 (for example, precipitation, wind speed / wind direction, humidity, solar radiation, etc.), as with the weekday / holiday correction, the predicted demand number V A configuration used for correction may be used.

[Another embodiment]
Another embodiment will be described below.

  <1> In the above-described embodiment, the component usage rate calculation unit 14 calculates the component usage rate on the assumption that, in principle, the same component is replaced only once for the same device. This is because the actual situation of replacing the same component multiple times for the same device is considered for the same component, considering the period for replacing the component due to a malfunction of the device and the period for replacing the device itself. This takes into account the fact that it is significantly less than the fact that it is exchanged only once, and the amount of information and the amount of computation that are additionally required when considering this.

  In other words, even if there is a situation where the same parts are replaced multiple times for the same equipment, the number is within the error range, and most of them are used for calculating the component usage rate and the predicted demand number. This calculation method is employed from the viewpoint of not affecting the operation.

  However, of course, it is also possible to calculate the component usage rate after strictly considering the case where a part is replaced multiple times for the same device, and calculate the predicted demand number based on this value. Absent.

  In the method described in the first embodiment (the same applies to the second and third embodiments), the number of parts i used (number of replacements) is calculated for each period (specific use period) y that has elapsed from the use start time of the device. did. On the other hand, by calculating the number of parts i used (number of replacements) for each period z that has elapsed since the start of use of the part i of the device, not for each period y that has elapsed since the start of use of the equipment, Find the usage rate. In addition, by calculating the number of existing units by actual usage period for each part in customer area A and multiplying this by the part usage rate for each part and for each period that has elapsed since the start of use of the part, Ask for.

  However, this method requires a huge amount of computation to calculate the component usage rate and the number of existing units by component, and to replace the same component multiple times depending on the nature of the equipment. Since there are very few cases, when it is assumed that repairers use this system for the purpose of forecasting the demand for parts, the period of use (years) after the start of use of the equipment is used. It is preferable that the demand prediction is performed by the method of the first (second, third) embodiment in which the prediction is performed by regarding the period (years) elapsed from the time.

  <2> In each of the above embodiments, if each repairer can grasp information about the start time of use of each owned device, the repairer will preliminarily “elapsed days” and “elapsed years” (that is, It is good also as a structure which transmits as a repair work report data in the form which attached | subjected the information regarding a specific use period. In this case, since the component usage rate can be calculated only from the information attached to the repair work report data, the component usage record data creation unit 13 and the component use record data storage unit 23 are not necessarily required.

  <3> Since the component usage rate is information used when making a demand forecast for any customer area, the component usage rate calculation unit 14 periodically updates the information in the component usage rate storage unit 24. Is preferred. However, this is not indispensable, and the component usage rate may be calculated by the component usage rate calculation unit 14 each time a demand prediction instruction is given from a repairer.

  <4> In the third embodiment described above, the meteorological data is registered in the meteorological data storage unit 26 for each customer area. However, the meteorological data in a region crossing each customer area is registered. It is also good. In this case, it may be used for the demand prediction of each customer area based on the one meteorological data registered across the cross.

  <5> In the above-described embodiment, it is assumed that customer areas in charge are different for each repair company. However, by setting all areas as calculation targets, each dealer accepts repairs for customers in all areas. Even in this case, the present system can be applied.

  In this case, the “existing number calculation unit 16 by device type / area” is changed to the “existing number calculation unit 16 by device type”, and the “part demand prediction unit 17 by device type / area” is “part demand prediction unit 17 by device type”. "May be substituted. Note that, by omitting the contents related to the customer area, it is possible to perform demand prediction by the same method as the above method. .

  <6> The structure and display method of each data described in the above embodiment is merely an example, and any data structure may be used as long as the data shows the same contents.

11: Data input / output unit 13: Part usage record data creation unit 14: Part usage rate calculation unit 16: Existing number calculation unit by device type / area 17: Part prediction unit by device type / area 21: Repair work report data storage unit 22: Customer-owned equipment data storage unit 23: Component usage record data storage unit 24: Component usage rate storage unit 26: Weather data storage unit 31: Control unit A0, A1,..., AN: Customer area DCR: Data center M0, M1 , ..., MN: Repairman

Claims (12)

A parts demand forecasting system that predicts the demand for parts required for equipment repair, including customer-owned equipment data storage unit, repair work report data storage unit, component usage record data storage unit, component usage record data creation unit, and component use It has a rate calculation unit, an existing number calculation unit by device type, and a parts demand prediction unit by device type,
The customer-owned equipment data storage unit is
Data relating to owned equipment that is currently owned by the owning customer or previously owned, including at least information on the owning customer, information on the type of the owned equipment, and information on when the equipment was first used. If the device has been owned in the past, it records customer-owned device data that further includes information on the end of use,
The repair work report data storage unit is
Data on the contents of repairs that have been performed in the past, at least information on the repair customer who performed the repair, information on the model of the repair device that performed the repair, information on the parts used in the repair, and the repair Repair work report data including information on repair timing
The component usage record data creation unit
By associating the information of the owned customer indicated in the customer owned equipment data with the information on the type of the owned equipment, and the information on the repair customer indicated in the repair work report data with the information on the type of the repaired equipment, Create part usage record data including information on a specific usage period that defines the model, the used part used for repair, and the period from the use start time to the repair time as a multiple of a predetermined unit period. Record it in the usage data storage unit,
The component usage rate calculation unit
The number of actual use of all the owned devices registered in the customer-owned device data storage unit is the period in which the user has experienced use and is defined by a multiple of the unit period, and for each device type To calculate the number of devices by device type and experience period,
Based on the component usage record data, the number of the repair devices in which the used components are actually used is counted for each specific usage period and for each device type, and for each used component, By dividing by the number of devices by experience type by device type, the ratio of the number of used parts required for all the owned devices was defined for each specific use period and for each device type and for each used component. Calculate the component usage rate that is the value,
The existing number calculation unit by the device type is:
Based on the customer-owned device data, the number of devices that are currently owned by the owned customer among the owned devices is defined as a multiple of the unit period from the start of use to the present. By counting for each actual use period and for each device type, the number of existing devices by the actual use period by the device type of the owned equipment is calculated,
The part demand forecasting unit by device type is
With respect to the same used parts, the calculation result obtained by multiplying the corresponding specific usage period and the actual usage period and the corresponding device type by the number of parts used and the existing number by the actual usage period by the equipment type is the same. By summing up each of the used parts, the predicted demand number for each used part over the unit period is derived. When the unit period is one year,
The part demand forecasting unit by device type is
Based on the part usage record data, the number of used parts for each used part is totaled for each month indicated by the repair period, and further, the ratio of the total value for each month is obtained to obtain the monthly usage ratio for each used part. To calculate
The part demand according to claim 1, wherein after deriving the predicted demand number for each used part over one year, the predicted demand number for each used part by month is derived by proportionally allocating according to the monthly usage ratio. Prediction system. The part demand forecasting unit by device type is
Based on the part usage record data, the number of used parts for each used part is tabulated according to weekdays and holidays indicated by the repair time, and the tabulated value for each weekday / holiday is further divided by the number of weekdays / holidays when the tabulation is performed. Later, by calculating these ratios, the usage ratio by weekdays and holidays is calculated,
After calculating the forecasted demand per day by dividing the forecasted demand for each used part by month by the number of days organized by month, the day-by-weekday / holiday The parts demand forecasting system according to claim 2 which derives the number of forecast demand per hit. It is a configuration further comprising a weather data storage unit that records weather data including at least past temperature data over a plurality of years from the latest,
The part demand forecasting unit by device type is
While reading out the weather data from the weather data storage unit, by collating with the repair time of the parts usage record data, the number of used parts for each used part for each temperature range constituted by a predetermined temperature range, Furthermore, by deriving these ratios, the component usage ratio by temperature range is calculated,
Further, an average temperature over the unit period is derived based on the weather data, a reference temperature zone to which the average temperature belongs is recognized, and the predicted number of demand for each used part over the unit period is divided by the number of days in the unit period. By calculating the number of predicted demand for each used part on the day that belongs to the reference temperature range, and to this value, the component usage ratio for each temperature range in the reference temperature range and the prediction target date for performing the demand prediction The parts demand forecasting system according to claim 1 which calculates the number of forecast demand according to said use parts by said forecast object day by multiplying with the ratio with said parts use ratio according to said temperature zone in the temperature zone where forecast temperature belongs. The part demand forecasting unit by device type is
Based on the part usage record data, the number of used parts for each used part is tabulated by weekday / holiday indicated by the repair period, and further, the number of used parts is divided by the number of weekdays / holiday that has been tabulated. By deriving the ratio, the usage ratio by weekday holiday is calculated,
The correction for the forecast demand number for each used part for the forecast target day is performed in consideration of the use ratio by weekday holiday according to whether the forecast target day is a weekday or a holiday. Item 4. A parts demand forecasting system according to item 4. The customer-owned equipment data includes information on a customer area to which the owned customer belongs,
The repair work report data includes information on a customer area to which the repair customer belongs,
The existing number calculation unit by the device type is:
Within the target customer area, which is the target customer area for which demand is to be predicted, the number of devices currently owned by the owned customer among the owned devices is the period from the start of use to the present. By counting for each device type and the actual usage period defined by a multiple of the unit period, the number of existing devices by the actual usage period by the device type of the owned device is calculated,
The part demand forecasting unit by device type is
For the same used parts, multiply the corresponding specific usage period and the actual usage period and the corresponding equipment type by the part usage rate and the number of existing equipment by the actual usage period in the target customer area. 4. The predicted number of demand for each used part in the target customer area over the unit period is derived by adding the calculated results for each of the same used parts. 5. The parts demand forecasting system according to claim 1. A weather data storage unit that records weather data including at least past temperature data over a plurality of years;
The part demand forecasting unit by device type is
While reading out the weather data from the weather data storage unit and comparing it with the repair time and the customer area of the part usage record data, the number of used parts for each used part for each temperature range constituted by a predetermined temperature range To calculate the component usage ratio by temperature range, and to derive these ratios,
Further, an average temperature over the unit period in the target customer area is derived based on the weather data, a reference temperature range to which the average temperature belongs is recognized, and the used parts in the target customer area over the unit period Is divided by the number of days in the unit period to calculate the number of predicted demand for each used part on the day belonging to the reference temperature range, and this value is the temperature-specific part in the reference temperature range. Multiplication by the ratio between the usage ratio and the component usage ratio for each temperature range in the temperature range to which the predicted temperature of the prediction target day for which demand prediction is to be performed, makes a prediction for each used component for the prediction target date in the target customer area The parts demand prediction system according to claim 6, wherein the number of demands is calculated. The customer-owned device data storage unit, the repair work report data storage unit, the component usage result data storage unit, the component usage result data creation unit, the component usage rate calculation unit, the existing number calculation unit by device type, and the A server computer equipped with a part demand forecasting unit by device type;
A plurality of operation terminals electrically connected to the server computer,
The operation terminal belongs to a repair company in charge of the different customer area, and periodically transmits the repair work report data including information on the customer area in charge to the server computer, and the data is transmitted to the server computer. It is configured to record in the repair work report data storage unit,
When a demand prediction instruction is given from the operation terminal to the server computer, the predicted demand number in the customer area in charge of the repair shop to which the operation terminal belongs is calculated, and the result is transmitted to the operation terminal. The parts demand forecasting system according to claim 6 or 7, wherein the parts demand forecasting system is provided. A parts demand forecasting system for forecasting demand for parts required for equipment repair,
A customer-owned equipment data storage unit, a repair work report data storage unit, a component usage rate calculation unit, an existing number calculation unit by device type, and a part demand prediction unit by device type,
The customer-owned equipment data storage unit is
Data relating to owned equipment that is currently owned by the owning customer or previously owned, including at least information on the owning customer, information on the type of the owned equipment, and information on when the equipment was first used. If you have owned the device in the past, we record customer-owned device data including information on when to end use,
The repair work report data storage unit is
Data on the contents of repairs that have been performed in the past, at least information on the repair customer who performed the repair, information on the model of the repair device that performed the repair, information on the parts used in the repair, and the repair Repair work report data including information on the repair time and information on the start of use of the equipment.
The component usage rate calculation unit
The number of actual use of all the owned devices registered in the customer-owned device data storage unit is the period in which the user has experienced use and is defined by a multiple of the unit period, and for each device type To calculate the number of devices by device type and experience period,
Based on the part usage record data, the number of the repair devices in which the used parts are actually used, for each specific usage period in which the period from the use start time to the repair time is defined as a multiple of a predetermined unit period And for each device type and counting by each used component, by dividing this by the number of corresponding device type by experience period, the ratio of the number of used devices required among all the owned devices A component usage rate which is a value defined for each specific usage period and for each device type and for each used component,
The existing number calculation unit by the device type is:
Based on the customer-owned device data, the number of devices that are currently owned by the owned customer among the owned devices is defined as a multiple of the unit period from the start of use to the present. By counting for each actual use period and for each device type, the number of existing devices by the actual use period by the device type of the owned equipment is calculated,
The part demand forecasting unit by device type is
With respect to the same used parts, the calculation result obtained by multiplying the corresponding specific usage period and the actual usage period and the corresponding device type by the number of parts used and the existing number by the actual usage period by the equipment type is the same. By summing up each of the used parts, the predicted demand number for each used part over the unit period is derived. A parts demand prediction method for predicting demand for parts required for equipment repair by computer data processing,
Data relating to owned equipment that is currently owned by the owning customer or previously owned, including at least information on the owning customer, information on the type of the owned equipment, and information on when the equipment was first used. If the equipment has been owned in the past, record customer-owned equipment data that further includes information on the end of use, and data on the details of repairs that were performed in the past, including at least information on repair customers who performed repairs. Record repair work report data including information on the type of repair equipment that performed the repair, information on the parts used in the repair, and information on the timing of the repair.
Information of the owned customer and the type of the owned equipment indicated in the recorded customer owned equipment data, and information of the repair customer and the type of the repair equipment indicated in the recorded repair work report data By associating the above information, the component usage record includes information on the device type, the used parts used for repair, and the specific use period that specifies the period from the use start time to the repair time as a multiple of a predetermined unit period. Create and record data,
Counts the number of devices actually used by all the owned devices in which the customer-owned device data is recorded, for each experience period that is a period of experience of use and defined by a multiple of the unit period, and for each device type. To calculate the number of devices by device type and experience period,
Based on the component usage record data, the number of the repair devices in which the used components are actually used is counted for each specific usage period and for each device type, and for each used component, By dividing by the number of devices by experience type by device type, the ratio of the number of used parts required for all the owned devices was defined for each specific use period and for each device type and for each used component. Calculate the component usage rate that is the value,
Based on the customer-owned device data, the number of devices that are currently owned by the owned customer among the owned devices is defined as a multiple of the unit period from the start of use to the present. By counting for each actual use period and for each device type, the number of existing devices by the actual use period by the device type of the owned equipment is calculated,
With respect to the same used parts, the calculation result obtained by multiplying the corresponding specific usage period and the actual usage period and the corresponding device type by the number of parts used and the existing number by the actual usage period by the equipment type is the same. A part demand forecasting method for deriving a predicted demand number for each used part over the unit period by summing up each used part. The customer-owned equipment data includes information about a customer area to which the owned customer belongs, and the repair work report data includes information about a customer area to which the repair customer belongs,
Within the target customer area, which is the target customer area for which demand is to be predicted, the number of devices currently owned by the owned customer among the owned devices is the period from the start of use to the present. By counting for each device type and the actual usage period defined by a multiple of the unit period, the number of existing devices by the actual usage period by the device type of the owned device is calculated,
For the same used parts, multiply the corresponding specific usage period and the actual usage period and the corresponding equipment type by the part usage rate and the number of existing equipment by the actual usage period in the target customer area. 11. The component according to claim 10, wherein the predicted demand number for each used component in the target customer area over the unit period is derived by adding the calculated results for each of the same used components. Demand forecast method. Read out weather data including at least past temperature data over a plurality of years, and collate with the repair time and the customer area of the part usage record data for each used part for each temperature range configured with a predetermined temperature range. Calculate the component usage ratio for each temperature range by calculating the number of used parts and deriving these ratios.
Deriving an average temperature over the unit period in the target customer area based on the weather data, recognizing a reference temperature zone to which the average temperature belongs, and prediction for each used part in the target customer area over the unit period By dividing the number of demands by the number of days in the unit period, the predicted demand number for each used part on the day belonging to the reference temperature range is calculated, and this value is used as the component usage ratio for each temperature range in the reference temperature range. And the ratio of the component usage ratio for each temperature range in the temperature range to which the predicted temperature of the prediction target day for which the demand prediction is performed is multiplied by the number of predicted demands for each used component in the target target area in the target customer area The component demand prediction method according to claim 11, wherein:

Images