Remaining money management system with cash demand prediction

Abstract

The invention provides a remaining money management system which can predict and manage a demanded cash amount for an ATM. The remaining money management system includes one or more automatic cash transaction apparatus, and a remaining money management apparatus which manages remaining money amounts based on transaction data from the automatic cash transaction apparatus. The remaining money management apparatus includes a storage section for cumulatively storing transaction data from the automatic cash transaction apparatus, and a prediction section for calculating and predicting demanded cash amounts within a designated time for the automatic cash transaction apparatus based on transaction data in the past accumulated in the storage section. The remaining money amount management system can be applied to an apparatus for managing ATMs installed in monetary facilities such as banks.

Claims

What is claimed is:

1. A remaining money management system, comprising:

one or more automatic cash transaction apparatus; and

a remaining money management apparatus for managing remaining money amounts corresponding to cash accommodated in said automatic cash transaction apparatus based on transaction data from said automatic cash transaction apparatus;

said remaining money management apparatus including a storage section for cumulatively storing transaction data from said automatic cash transaction apparatus, and a prediction section for calculating and predicting demanded cash amounts within a designated period of time for said automatic cash transaction apparatus based on the transaction data in the past stored in said storage section.

2. A remaining money management system as claimed in claim 1, wherein said prediction section

calculates a mean value of the demanded cash amounts in the past based on the transaction data in the past stored in said storage section and converts, for each of qualitative factors which may have an influence on the demanded cash amount, a degree of influence of the qualitative factor on the demanded cash amount into a numerical value, and

calculates a sum value of the mean value and the numerical data of the degrees of influence of the quality factors as a predicted value of the demanded cash amount.

3. A remaining money management system as claimed in claim 2, wherein said prediction section converts the degree of influence of each of the quality factors for each of said automatic cash transaction apparatus and produces and holds a prediction table of the numerical data for each of said automatic cash transaction apparatus.

4. A remaining money management system as claimed in claim 3, wherein said prediction table is produced for each of different currencies and after each fixed time interval.

5. A remaining money management system as claimed in claim 2, wherein said remaining money management apparatus calculates an error between an actual result value of the demanded cash amount and the predicted value of the demanded cash amount obtained by said prediction section and automatically selects, based on the error, a qualitative factor to be used for calculation of the predicted value of the demanded cash amount.

6. A remaining money management system as claimed in claim 1, wherein said prediction section predicts, before money is supplemented to said automatic cash transaction apparatus, demanded cash amounts for a period of time after the current money supplementing time until a next money supplementing time and outputs the predicted values as initial load money amounts to said automatic cash transaction apparatus.

7. A remaining money management system as claimed in claim 1, wherein said prediction section predicts demanded cash amounts per unit time of said automatic cash transaction apparatus and predicts demanded cash amounts on the real time basis till a predetermined point of time, normally supervises errors between actual remaining money amounts and the real time predicted values obtained by said prediction section, calculates amounts of cash to be supplemented or extracted with which the errors are canceled, and indicates the calculated amounts of cash to be supplemented or extracted.

8. A remaining money management system as claimed in claim 7, wherein said remaining money management apparatus

calculates, for each of said automatic cash transaction apparatus, errors between actual remaining cash amounts and the real time predicted values obtained by said prediction section,

discriminates, based on the errors calculated for each of said automatic cash transaction apparatus, one of said automatic cash transaction apparatus which requires cash to be supplemented thereto and another one of said automatic cash transaction apparatus which holds surplus cash, and

develops an instruction to extract cash from the automatic cash transaction apparatus which holds surplus cash and supplement the extracted cash to the automatic cash transaction apparatus which requires cash to be supplemented thereto.

9. A remaining money management system as claimed in claim 8, wherein said remaining money management apparatus displays the instruction regarding the extraction and supplementing of cash so that a staff member referring to the displayed instruction may perform extraction and supplementing of cash from and to the relevant automatic cash transaction apparatus.

10. A remaining money management system as claimed in claim 8, further comprising a self-traveling robot for automatically performing extraction and supplementing of cash from and to the relevant automatic cash transaction apparatus, said remaining money management apparatus including a robot control section for controlling cash extraction and supplementing operations of said self-traveling robot in response to the instruction regarding the extraction and supplement of cash.

11. A remaining money management system as claimed in claim 10, wherein said remaining money management apparatus displays the instruction regarding the extraction and supplementing of cash so that, when said self-traveling robot is out of order or the extraction or supplementing of cash by said self-traveling robot is impossible, a staff member referring to the displayed instruction may perform extraction and supplementing of cash from and to the relevant automatic cash transaction apparatus.

12. A remaining money management system as claimed in claim 1, wherein said prediction section

produces and holds, for each of said automatic cash transaction apparatus, a predicted error table of errors between actual result values of the demanded cash amount and predicted values of the demanded cash amount obtained by said prediction section is stored,

calculates a standard deviation of the errors of the predictive error table as a standard error, and

calculates a value obtained by multiplying the calculated standard error by a safety coefficient as a safety cash amount to be added to the predicted value of the demanded cash amount obtained by said prediction section.

13. A remaining money management system as claimed in claim 1, wherein said prediction section

predicts, where a plurality of automatic cash transaction apparatus are involved, a total demanded cash amount for said plurality of automatic cash transaction apparatus,

produces and holds a predicted error table of errors between sums of actual result values of the demanded cash amount of said plurality of automatic cash transaction apparatus and predicted values of the total demanded cash amounts obtained for said plurality of automatic cash transaction apparatus by said prediction section,

calculates a standard deviation of the errors of said predicted error table as a standard error,

multiplies the calculated standard error by a safety coefficient to obtain a safety cash amount to be added to the predicted value of the total demanded cash amount obtained for said plurality of automatic cash transaction apparatus by said prediction section, and

distributes the calculated safety cash amount at a distribution ratio to said automatic cash transaction apparatus.

14. A remaining money management system as claimed in claim 13, wherein the safety cash amount is distributed equally to said plurality of automatic cash transaction apparatus.

15. A remaining money management system as claimed in claim 13, wherein the safety cash amount is distributed to said automatic cash transaction apparatus at a ratio of safety cash amounts individually calculated for said automatic cash transaction apparatus.

16. A remaining money management system as claimed in claim 13, wherein the safety cash amount is distributed to said automatic cash transaction apparatus at a ratio of squares of safety cash amounts individually calculated for said automatic cash transaction apparatus.

17. A remaining money management system as claimed in claim 1, wherein said prediction section predicts a predicted value of the demanded cash amount for each of said automatic cash transaction apparatus after the current point of time after each time interval and determines a point of time at which the predicted value coincides with a current remaining cash amount in the automatic cash transaction apparatus as an operable time.

18. A remaining money management system as claimed in claim 1, wherein, where said remaining money management apparatus is provided for each of a plurality of business offices, transaction data in the past of said plurality of business offices are analyzed to obtain a plurality of transaction form patterns to which said business offices individually belong, and in a first one of said business offices with which transaction data in the past to be used for prediction of demanded cash amounts by said prediction section are not cumulatively stored in said storage section, transaction data in the past of a second one of the other business offices which belongs to the same transaction form pattern as that of the first business office are stored into said storage section, said prediction section of the first business office predicting demanded cash amounts using the transaction data in the past of the second business office until new sufficient transaction data of the first business office are accumulated in said storage section.

19. A remaining money management system as claimed in claim 1, wherein, when an installation environment condition of said automatic cash transaction apparatus changes, a predicted value of a sum of demanded cash amounts obtained by said prediction section for said automatic cash transaction apparatus before the change of the installation environment condition is distributed at a distribution ratio to said automatic cash transaction apparatus after the change of the installation environment condition, and the thus distributed values are used as predicted values for said automatic cash transaction apparatus after the change of the installation environment condition.

20. A remaining money management system as claimed in claim 19, wherein the predicted value of the sum is distributed equally to said automatic cash transaction apparatus after the change of the installation environment condition.

21. A remaining money management system as claimed in claim 19, wherein the predicted value of the sum is distributed at a distribution ratio set by a staff member to said automatic cash transaction apparatus after the change of the installation environment condition.

22. A remaining money management system as claimed in claim 19, wherein the predicted value of the sum is distributed to said automatic cash transaction apparatus after the change of the installation environment condition at a distribution ratio calculated based on actual result data of said automatic cash transaction apparatus after the change of the installation environment condition.

23. A remaining money management system as claimed in claim 1, wherein an accounting system host computer connected by means of dedicated lines to said automatic cash transaction apparatus in a plurality of business offices functions as said remaining money management apparatus.

24. A remaining money management system as claimed in claim 1, wherein a staff member terminal provided at a window in a business office functions as said remaining money management apparatus.

25. A remaining money management system as claimed in claim 1, wherein transaction data of a staff member terminal provided at a window in a business office are cumulatively stored into said storage section, and said prediction section calculates and predicts demanded cash amounts within a designated time for said staff member terminal based on the transaction data in the past of said staff member terminal accumulated in said storage section.

26. A remaining money management system as claimed in claim 1, wherein a remote supervision apparatus connected by means of a communication line to said automatic cash transaction apparatus in a plurality of business offices functions as said remaining money management apparatus.

27. A remaining money management system as claimed in claim 21, wherein an instruction for said self-traveling robot to move to an object one of said automatic cash transaction apparatus and another instruction for said self-traveling robot to extract or supplement cash from or to the object automatic cash transaction apparatus are conveyed from said robot control section of said remaining money management apparatus to said self-traveling robot by a radio communication channel via a radio communication base station.

28. A remaining money management system as claimed in claim 21, wherein a free track for said self-traveling robot is formed from a magnetic tape applied to a floor, and said self-traveling robot includes a self-traveling truck for traveling along said free track while detecting said magnetic tape by means of a guiding magnetic sensor, and a robot mechanism mounted on said self-traveling truck for removing or mounting a cash cassette from or onto one of said automatic cash transaction apparatus from the rear side of the automatic cash transaction apparatus to perform extraction or supplementing of cash.

29. A remaining money management system as claimed in claim 28, wherein marking magnetic tapes for individually specifying positions of said automatic cash transaction apparatus are applied along said magnetic tape which forms said free track, and said self-traveling robot detects any of said marking magnetic tapes by means of a marking magnetic sensor to specify one of said automatic cash transaction apparatus designated by said robot control section.

30. A remaining money management system as claimed in claim 29, wherein a positioning apparatus for guiding a guide roller mounted on said self-traveling robot to control a stopping position of said self-traveling robot with respect to one of said automatic cash transaction apparatus is provided on a rear face side of each of said automatic cash transaction apparatus.

31. A remaining money management system as claimed in claim 30, wherein said positioning apparatus includes an arrival sensor for detecting that said self-traveling robot arrives and stops at one of said automatic cash transaction apparatus in order to extract or supplement cash from or to the automatic cash transaction apparatus.

32. A remaining money management system as claimed in claim 31, wherein said positioning apparatus includes a pair of feed terminals for supplying power for operating said robot mechanism of said self-traveling robot when said self-traveling robot stops while said self-traveling robot includes a pair of collection terminals for contacting with said feed terminals, and when said arrival sensor detects arrival of said self-traveling robot, power is supplied from said feed terminals to said collection terminals.

33. A remaining money management system as claimed in claim 32, wherein said self-traveling robot includes a switching circuit for automatically switching power for said robot mechanism supplied to said collection terminals so that the power may be used to charge a self-traveling car power supply while said self-traveling robot stops and said robot mechanism of said self-traveling robot is inoperative.

34. A remaining money management system as claimed in claim 32, wherein a power supply circuit for supplying power for said robot mechanism to said feed terminals is provided separately from a power supply circuit for supplying power to said automatic cash transaction apparatus.

35. A remaining money management system as claimed in claim 34, further comprising a station controller including an arrival sensor supervision section for supervising detection conditions of said automatic cash transaction apparatus by the arrival sensors in a concentrated manner and notifying the detection conditions by the arrival sensors as an operation condition of said self-traveling robot to said remaining money management apparatus, an operation panel for inputting an instruction to control the operation condition of said self-traveling robot, a power supply section for supplying power for said robot mechanism, and a door opening/closing supervision section for supervising an opening/closing condition of a rear side door of one of said automatic cash transaction apparatus which is opened when a cash extracting or supplementing operation is to be performed for the automatic cash transaction apparatus by said self-traveling robot.

36. A remaining money management system as claimed in claim 35, wherein said rear side door of any of said automatic cash transaction apparatus is automatically driven to open or close by power supplied from said power supply section of said station controller, and each of said automatic cash transaction apparatus includes a door control section for driving, when a cash extracting or supplementing operation by said self-traveling robot is to be performed for the automatic cash transaction apparatus, said rear side door to open to provide a space sufficient to allow said cash cassette to be removed or mounted from or onto the automatic cash transaction apparatus therethrough in response to an instruction from said self-traveling robot and for driving, when the removal or mounting of said cash cassette comes to an end, said rear side door to be closed.

37. A remaining money management system as claimed in claim 36, wherein said self-traveling robot includes a light emitting element for emitting an optical signal to open or close said rear side door of any of said automatic cash transaction apparatus while each of said automatic cash transaction apparatus includes a light receiving element for receiving the optical signal from said light emitting element, and, when to open or close said rear side door of one of said automatic cash transaction apparatus in order to perform a cash extraction or supplementing operation by said self-traveling robot, said light emitting element of said self-traveling robot emits an optical signal for the automatic cash transaction apparatus toward said light receiving element of the automatic cash transaction apparatus and, only when said light receiving element of the automatic cash transaction apparatus receives an optical signal specified for the automatic cash transaction apparatus, said door control section of the automatic cash transaction apparatus drives said rear side door to open or close.

38. A remaining money management system as claimed in claim 36, wherein each of said automatic cash transaction apparatus includes a door open/close detection sensor for detecting and notifying an open or closed condition of said rear side door of any of said automatic cash transaction apparatus to said door opening/closing supervision section, which supervises a detection condition by said door open/close detection sensor, and said station controller includes a service interruption detection circuit for detecting a service interruption of said power supply section, and a control section for causing, if the detection condition by said door open/close detection sensor is an open condition in any of said automatic cash transaction apparatus when a service interruption is detected by said service interruption detection circuit, power of a backup power supply to be supplied until said door open/close detection sensor detects that said rear side door of the automatic cash transaction apparatus enters a closing condition.

39. A remaining money management system as claimed in claim 38, wherein said station controller includes a time supervision section for supervising an elapsed time after detection of a service interruption by said service interruption detection circuit, and said control section stops the backup power supply and performs trouble ending processing if the detection condition of said door open/close detection sensor still is the open condition in the automatic cash transaction apparatus at a point of time when it is detected by said time supervision section that a fixed time elapses.

40. A remaining money management system as claimed in claim 28, wherein said robot mechanism includes a receipt paper supplementing apparatus for supplementing receipt paper to any of said automatic cash transaction apparatus.

41. A remaining money management system as claimed in claim 28, further comprising a cassette station for defining a location at which said cash cassette is removed from or mounted onto said self-traveling robot.

42. A remaining money management system as claimed in claim 41, wherein said cassette station includes a locking mechanism for restricting removal of said cash cassette from said cassette station.

43. A remaining money management system as claimed in claim 41, wherein said cassette station includes a counting apparatus for counting an amount of cash in said cash cassette, and an automatic cash loading and extracting apparatus for automatically loading or extracting cash into or from said cash cassette.

44. A remaining money management system as claimed in claim 28, wherein one of said automatic cash transaction apparatus is used as a location where said cash cassette is removed from or mounted onto said self-traveling robot.

45. A remaining money management system as claimed in claim 44, wherein an amount of cash in said cash cassette is counted making use of a counting function of the automatic cash transaction apparatus which is used as the location where said cash cassette is removed from or mounted onto said self-traveling robot, and a result of the counting is notified to said remaining money management apparatus.

Description

BACKGROUND OF THE INVENTION

1. Field of the Invention

This invention relates to a remaining money management system for managing remaining money (remaining amounts of different currencies such as paper money or bills and coins) in an automatic teller machine installed in a bank or the like.

2. Description of the Related Art

Generally, supplement or extraction of cash into or from an automatic teller machine (hereinafter referred to ATM or automatic machine) installed in an automatic machine corner of a bank is performed by a staff member.

In such supplement or extraction, the numbers of different currencies to be supplemented or extracted and the time for supplement or extraction are decided frequently relying upon the experience or the sixth sense of the staff member, and usually, when shortage of the cash occurs with an ATM, a call is developed from the ATM and a staff member supplementally loads the ATM with cash in response to the call.

Since the timing at which cash is to be supplemented to or extracted from an ATM or the amount of money to be supplemented or extracted is decided relying upon the experience or the sixth sense of a staff member and cash is supplemented or extracted by hand of the staff member in this manner, the burden to the staff member is heavy. The burden is very heavy particularly where the staff member takes charge of a large number of ATMs.

Further, although the amount of money to be loaded into an ATM is desired to be as small as possible, where supplement and extraction of cash into and from the ATM is performed by a staff member as described above, the reduction of the amount of money to be loaded into the ATM depends upon the ability of the staff member.

SUMMARY OF THE INVENTION

It is an object of the present invention to provide a remaining money management system wherein demanded numbers of different currencies for an ATM can be predicted and managed to achieve reduction of the amount of money to be loaded into the ATM and reduction of the burden to a staff member.

In order to attain the object described above, according to the present invention, there is provided a remaining money management system, which comprises one or more automatic cash transaction apparatus, and a remaining money management apparatus for managing remaining money amounts corresponding to cash accommodated in the automatic cash transaction apparatus based on transaction data from the automatic cash transaction apparatus, the remaining money management apparatus including a storage section for cumulatively storing transaction data from the automatic cash transaction apparatus, and a prediction section for calculating and predicting demanded cash amounts within a designated period of time for the automatic cash transaction apparatus based on the transaction data in the past stored in the storage section.

In the remaining money management system, the prediction section of the remaining money management apparatus calculates and predicts demanded cash amounts within a designated period of time for the automatic cash transaction apparatus based on transaction data in the past stored in the storage section, and the remaining money amounts in the automatic cash transaction apparatus can be managed based on the predicted values. Consequently, the amounts of cash to be loaded into the automatic cash transaction apparatus can be reduced without relying upon the staff member, and the burden to the staff member can be reduced remarkably.

Further objects, features and advantages of the present invention will become apparent from the following detailed description when read in conjunction with the accompanying drawings in which like parts or elements are denoted by like reference characters.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a block diagram illustrating an aspect of the present invention;

FIG. 2 is a block diagram showing a general construction of a remaining money management system to which the present invention is applied;

FIG. 3 is a block diagram showing a detailed functional construction of a remaining money management terminal in the remaining money management system;

FIG. 4 is a diagrammatic view showing a directory structure of transaction data stored on a hard disk in the remaining money management system;

FIG. 5 is a diagrammatic view illustrating a concept of a prediction method for a demanded cash amount employed in the remaining money management system;

FIG. 6 is a diagrammatic view illustrating details of the concept of FIG. 5 using exemplary numerical values;

FIG. 7 is a flow chart illustrating data collection processing of the remaining money management system;

FIG. 8 is a diagrammatic view illustrating examples of the qualitative factor in the remaining money management system;

FIG. 9 is a diagrammatic view illustrating an example of acquisition of data for each qualitative factor upon prediction of a demanded cash amount in the remaining money management system;

FIG. 10 is a diagrammatic view illustrating a cross tabulation table of qualitative factors produced upon conversion of qualitative data into quantitative data in prediction of a demanded cash amount in the remaining money management system;

FIG. 11 is a diagrammatic view illustrating values of factors defined upon conversion of qualitative data into quantitative data in prediction of a demanded cash amount in the remaining money management system;

FIG. 12 is a graph illustrating significance of a value of a factor obtained upon prediction of a demanded cash amount in the remaining money management system;

FIG. 13 is a diagrammatic view illustrating prediction tables for individual ATMs in the remaining money management system;

FIG. 14 is a similar view but illustrating an example of details of the forecast tables illustrated in FIG. 13;

FIG. 15 is a flow chart illustrating demanded cash amount prediction processing of the remaining money management system;

FIG. 16 is a diagrammatic view illustrating a technique of calculating predicted values for a designated period of time based on the forecast tables of FIGS. 13 and 14 in the remaining money management system;

FIG. 17 is a diagrammatic view illustrating an automatic selection function for a qualitative factor in the remaining money management system;

FIG. 18 is a graph illustrating an initial prediction function of the remaining money management system;

FIG. 19 is a similar view but illustrating a remaining money variation prediction function of the remaining money management system;

FIGS. 20 and 21 are schematic views illustrating different examples of a display of a result of demanded cash amount prediction of the remaining money management system;

FIG. 22 is a diagrammatic view illustrating a safe money amount determination method employed in the remaining money management system;

FIGS. 23(a) and 23(b) are a diagrammatic view and a graph, respectively, illustrating processing when group control is performed upon determination of a safe money amount in the remaining money management system;

FIGS. 24(a) and 24(b) are a diagrammatic view and a graph, respectively, illustrating processing when group control is not performed upon determination of a safe money amount in the remaining money management system;

FIG. 25 is a graph illustrating an effect which is achieved when group control is performed upon determination of safe currency numbers in the remaining money management system;

FIG. 26 is a graph illustrating an operable time prediction function of the remaining money management system;

FIG. 27 is a flow chart illustrating the operable time forecasting function of FIG. 26;

FIG. 28 is a flow chart illustrating an initial transaction data setting method employed in the remaining money management system;

FIG. 29 is a flow chart illustrating a demanded cash amount predicted value setting method when an ATM environment varies in the remaining money management system;

FIG. 30 is a diagrammatic view illustrating initial cash loading of the remaining money management system;

FIG. 31 is a diagrammatic view illustrating a money circulation method in an ATM group by the remaining money management system;

FIG. 32 is a graph illustrating an extraction money amount determination method upon money circulation in an ATM group by the remaining money management system;

FIG. 33 is a block diagram schematically showing an internal structure of an ATM employed in the remaining money management system;

FIG. 34 is a flow chart illustrating operation of a remaining money management terminal when money circulation processing is performed in the remaining money management system;

FIG. 35 is a flow chart illustrating operation in the inside of an ATM when money circulator processing is performed in the remaining money arrangement system;

FIG. 36 is a flow chart illustrating cash circulation processing of the remaining money arrangement system where it includes a self-traveling car;

FIG. 37 is a block diagram of the remaining money arrangement system wherein a host computer or a remote control apparatus has a function as a remaining money management terminal;

FIG. 38 is a block diagram showing a general construction of the remaining money management system where it includes a self-traveling car;

FIG. 39 is a block diagram showing a function construction of the remaining money management system where it includes a self-traveling car;

FIG. 40 is a schematic perspective view, partly broken, showing a self-traveling car (self-traveling robot) employed in the remaining money management system;

FIG. 41 is a perspective view showing a robot mechanism of the self-traveling car of FIG. 40;

FIG. 42 is a bottom plan view of the self-traveling car of FIG. 40;

FIG. 43 is a schematic plan view showing a marking magnetic tape for cooperating with the self-traveling car of FIG. 40;

FIG. 44 is a schematic plan view showing a positioning mechanism for the self-traveling car of FIG. 40;

FIG. 45 is a plan view showing details of the positioning mechanism of FIG. 44;

FIG. 46 is a plan view showing a guide rail of the positioning mechanism shown in FIG. 44;

FIG. 47 is a schematic plan view of the remaining money management system illustrating an effect achieved by employment of a free track in the form of a magnetic tape;

FIG. 48 is a perspective view of an ATM employed in the remaining money management system as viewed from the rear side;

FIG. 49 side elevational view of the ATM shown in FIG. 48;

FIG. 50 is a schematic plan view showing a mechanism employed in the remaining money management system for supplying power for the robot mechanism of FIG. 41;

FIG. 51 is a block diagram showing a construction in the remaining money arrangement system for achieving a switching function for power for the robot mechanism of FIG. 41;

FIG. 52 is a block diagram showing a construction in the remaining money arrangement system for achieving an automatic door compulsory closing function by a station controller upon service interruption;

FIG. 53 is a schematic view showing the rear face of the ATM of FIG. 48 where a light reception element is provided;

FIG. 54 is a schematic side elevational view illustrating an optical signal communication condition when the door of the ATM of FIG. 48 is automatically opened or closed;

FIG. 55 is a block diagram showing a construction of an optical signal communication system of the ATM of FIG. 48 when the door of the ATM is automatically opened or closed;

FIGS. 56(a) and 56(b) are waveform diagrams illustrating driving conditions of a light emitting element of the optical signal communication system of FIG. 55 when the door of the ATM is automatically opened or closed;

FIG. 57 is a flow chart illustrating a cash cassette removing operation of the robot mechanism of FIG. 41;

FIG. 58 is a flow chart illustrating a money cassette loading operation of the robot mechanism of FIG. 41;

FIG. 59 is a perspective view showing a receipt roll supplement unit employed in the remaining money management system;

FIG. 60 is a side elevational view schematically showing a receipt roll supplementing mechanism employed in the remaining money management system;

FIG. 61 is a perspective view schematically showing a cassette station employed in the remaining money management system;

FIG. 62 is a side elevational view schematically showing a locking mechanism of the cassette station of FIG. 61; and

FIG. 63 is a perspective view schematically showing the locking mechanism of FIG. 62.

DESCRIPTION OF THE PREFERRED EMBODIMENT

A. Aspect of the Invention

FIG. 1 is a block diagram illustrating an aspect of the present invention. Referring to FIG. 1, a remaining money management system includes one or more ATMs 1, and a remaining money management apparatus 2 for managing remaining money amounts corresponding to cash accommodated in the ATMs 1 based on transaction data from the ATMs 1.

The remaining money management apparatus 2 includes a storage section 2A for cumulatively storing transaction data from the ATMs 1, and a prediction section 2B for calculating and predicting demanded cash amounts within a designated period of time for the ATMs 1 based on the transaction data in the past stored in the storage section 2A.

The remaining money management system of the present invention may have the following various additional features.

(1) The prediction section 2B calculates a mean value of the demanded cash amounts in the past based on the transaction data in the past stored in the storage section 2A and converts, for each of qualitative factors which may have an influence on the demanded cash amount, a degree of influence of the qualitative factor on the demanded cash amount into a numerical value, and calculates a sum value of the mean value and the numerical data of the degrees of influence of the quality factors as a predicted value of the demanded cash amount. In this instance, the prediction section 2B converts the degree of influence of each of the quality factors for each of the ATMs 1 and produces and holds a prediction table of the numerical data for each of the ATMs 1. Further, the prediction table is produced for each of different currencies and after each fixed time interval.

(2) The remaining money management apparatus 2 calculates an error between an actual result value of the demanded cash amount and the predicted value of the demanded cash amount obtained by the prediction section 2B and automatically selects, based on the error, a qualitative factor to be used for calculation of the predicted value of the demanded cash amount.

(3) The prediction section 2B predicts, before money is supplemented to the ATMs 1, demanded cash amounts for a period of time after the current money supplementing time until a next money supplementing time and outputs the predicted values as initial load money amounts to the ATMs 1.

(4) The prediction section 2B predicts demanded cash amounts per unit time of the ATMs 1 and predicts demanded cash amounts on the real time basis till a predetermined point of time, normally supervises errors between actual remaining money amounts and the real time predicted values obtained by the prediction section 2B, calculates amounts of cash to be supplemented or extracted with which the errors are canceled, and indicates the calculated amounts of cash to be supplemented or extracted.

(5) The prediction section 2B produces and holds, for each of the ATMs 1, a predicted error table of errors between actual result values of the demanded cash amount and predicted values of the demanded cash amount obtained by the prediction section 2B is stored, calculates a standard deviation of the errors of the predictive error table as a standard error, and calculates a value obtained by multiplying the calculated standard error by a safety coefficient as a safety cash amount to be added to the predicted value of the demanded cash amount obtained by the prediction section 2B.

(6) The prediction section 2B predicts, where a plurality of ATMs 1 are involved, a total demanded cash amount for the plurality of ATMs 1, produces and holds a predicted error table of errors between sums of actual result values of the demanded cash amount of the plurality of ATMs 1 and predicted values of the total demanded cash amounts obtained for the plurality of ATMs 1 by the prediction section 2B, calculates a standard deviation of the errors of the predicted error table as a standard error, multiplies the calculated standard error by a safety coefficient to obtain a safety cash amount to be added to the predicted value of the total demanded cash amount obtained for the plurality of ATMs 1 by the prediction section 2B, and distributes the calculated safety cash amount at a distribution ratio to the ATMs 1.

In this instance, the safety cash amount may be distributed equally to the plurality of ATMs 1, or may be distributed to the ATMs 1 at a ratio of safety cash amounts individually calculated for the ATMs 1, or else may be distributed to the ATMs 1 at a ratio of squares of safety cash amounts individually calculated for the ATMs 1.

(7) The prediction section 2B predicts a predicted value of the demanded cash amount for each of the ATMs 1 after the current point of time after each time interval and determines a point of time at which the predicted value coincides with a current remaining cash amount in the ATM 1 as an operable time. (8) Where the remaining money management apparatus 2 is provided for each of a plurality of business offices, transaction data in the past of the plurality of business offices are analyzed to obtain a plurality of transaction form patterns to which the business offices individually belong, and in a first one of the business offices with which transaction data in the past to be used for prediction of demanded cash amounts by the prediction section 2B are not cumulatively stored in the storage section 2A, transaction data in the past of a second one of the other business offices which belongs to the same transaction form pattern as that of the first business office are stored into the storage section 2A, and then the prediction section 2B of the first business office predicts demanded cash amounts using the transaction data in the past of the second business office until new sufficient transaction data of the first business office are accumulated in the storage section 2A.

(9) When an installation environment condition of the ATMs 1 changes, a predicted value of a sum of demanded cash amounts obtained by the prediction section 2B for the ATMs 1 before the change of the installation environment condition is distributed at a distribution ratio to the ATMs 1 after the change of the installation environment condition, and the thus distributed values are used as predicted values for the ATMs 1 after the change of the installation environment condition.

In this instance, the predicted value of the sum may be distributed equally to the ATMs 1 after the change of the installation environment condition, or may be distributed at a distribution ratio set by a staff member to the ATMs 1 after the change of the installation environment condition, or else may be distributed to the ATMs 1 after the change of the installation environment condition at a distribution ratio calculated based on actual result data of the ATMs 1 after the change of the installation environment condition.

(10) The remaining money management apparatus 2 calculates, for each of the ATMs 1, errors between actual remaining cash amounts and the real time predicted values obtained by the prediction section 2B, discriminates, based on the errors calculated for each of the ATMs 1, one of the ATMs 1 which requires cash to be supplemented thereto and another one of the ATMs 1 which holds surplus cash, and develops an instruction to extract cash from the ATM 1 which holds surplus cash and supplement the extracted cash to the ATM 1 which requires cash to be supplemented thereto.

In this instance, the remaining money management apparatus 2 may display the instruction regarding the extraction and supplementing of cash so that a staff member referring to the displayed instruction may perform extraction and supplementing of cash from and to the relevant ATM 1.

(11) The remaining money management system further comprises a self-traveling robot for automatically performing extraction and supplementing of cash from and to the relevant ATM 1, and the remaining money management apparatus 2 includes a robot control section for controlling cash extraction and supplementing operations of the self-traveling robot in response to the instruction regarding the extraction and supplement of cash. Further, the remaining money management apparatus 2 displays the instruction regarding the extraction and supplementing of cash so that, when the self-traveling robot is out of order or the extraction or supplementing of cash by the self-traveling robot is impossible, a staff member referring to the displayed instruction may perform extraction and supplementing of cash from and to the relevant ATM 1.

(12) An accounting system host computer connected by means of dedicated lines to the ATMs 1 in a plurality of business offices may function as the remaining money management apparatus 2, or a staff member terminal provided at a window in a business office may function as the remaining money management apparatus 2, or else a remote supervision apparatus connected by means of a communication line to the ATMs 1 in a plurality of business offices may function as the remaining money management apparatus 2.

Transaction data of a staff member terminal provided at a window in a business office may be cumulatively stored into the storage section 2A, and the prediction section 2B may calculate and predict demanded cash amounts within a designated time for the staff member terminal based on the transaction data in the past of the staff member terminal accumulated in the storage section 2A.

(13) Various instructions (an instruction for the self-traveling robot to move to an object one of the ATMs 1 and another instruction for the self-traveling robot to extract or supplement cash from or to the object ATM 1) may be conveyed from the robot control section of the remaining money management apparatus 2 to the self-traveling robot by a radio communication channel via a radio communication base station.

(14) A free track for the self-traveling robot is formed from a magnetic tape applied to a floor, and the self-traveling robot includes a self-traveling truck for traveling along the free track while detecting the magnetic tape by means of a guiding magnetic sensor, and a robot mechanism mounted on the self-traveling truck for removing or mounting a cash cassette from or onto one of the ATMs 1 from the rear side of the ATMs 1 to perform extraction or supplementing of cash.

Where the self-traveling robot is provided in the remaining money management system of the present invention, the following various additional features can be provided.

(15) Marking magnetic tapes for individually specifying positions of the ATMs 1 are applied along the magnetic tape which forms the free track, and the self-traveling robot detects any of the marking magnetic tapes by means of a marking magnetic sensor to specify one of the ATMs 1 designated by the robot control section.

(16) A positioning apparatus for guiding a guide roller mounted on the self-traveling robot to control a stopping position of the self-traveling robot with respect to one of the ATMs 1 is provided on a rear face side of each of the ATMs 1. The positioning apparatus includes an arrival sensor for detecting that the self-traveling robot arrives and stops at one of the ATMs 1 in order to extract or supplement cash from or to the ATMs 1.

(17) The positioning apparatus includes a pair of feed terminals for supplying power for operating the robot mechanism of the self-traveling robot when the self-traveling robot stops while the self-traveling robot includes a pair of collection terminals for contacting with the feed terminals, and when the arrival sensor detects arrival of the self-traveling robot, power is supplied from the feed terminals to the collection terminals. A power supply circuit for supplying power for the robot mechanism to the feed terminals is provided separately from a power supply circuit for supplying power to the ATMs 1. The self-traveling robot includes a switching circuit for automatically switching power for the robot mechanism supplied to the collection terminals so that the power may be used to charge a self-traveling car power supply while the self-traveling robot stops and the robot mechanism of the self-traveling robot is inoperative.

(18) The remaining money management system further includes a station controller including an arrival sensor supervision section for supervising detection conditions of the ATMs 1 by the arrival sensors in a concentrated manner and notifying the detection conditions by the arrival sensors as an operation condition of the self-traveling robot to the remaining money management apparatus 2, an operation panel for inputting an instruction to control the operation condition of the self-traveling robot, a power supply section for supplying power for the robot mechanism, and a door opening/closing supervision section for supervising an opening/closing condition of a rear side door of one of the ATMs 1 which is opened when a cash extracting or supplementing operation is to be performed for the ATM 1 by the self-traveling robot.

(19) The rear side door of any of the ATMs 1 is automatically driven to open or close by power supplied from the power supply section of the station controller, and each of the ATMs 1 includes a door control section for driving, when a cash extracting or supplementing operation by the self-traveling robot is to be performed for the ATM 1, the rear side door to open to provide a space sufficient to allow the cash cassette to be removed or mounted from or onto the ATM 1 therethrough in response to an instruction from the self-traveling robot and for driving, when the removal or mounting of the cash cassette comes to an end, the rear side door to be closed.

(20) The self-traveling robot includes a light emitting element for emitting an optical signal to open or close the rear side door of any of the ATMs 1 while each of the ATMs 1 includes a light receiving element for receiving the optical signal from the light-emitting element, and, when to open or close the rear side door of one of the ATMs 1 in order to perform a cash extraction or supplementing operation by the self-traveling robot, the light emitting element of the self-traveling robot emits an optical signal for the ATM 1 toward the light receiving element of the ATM 1 and, only when the light receiving element of the ATM 1 receives an optical signal specified for the ATM 1, the door control section of the ATM 1 drives the rear side door to open or close.

(21) Each of the ATMs 1 includes a door open/close detection sensor for detecting and notifying an open or closed condition of the rear side door of any of the ATMs 1 to the door opening/closing supervision section, which supervises a detection condition by the door open/close detection sensor, and the station controller includes a service interruption detection circuit for detecting a service interruption of the power supply section, and a control section for causing, if the detection condition by the door open/close detection sensor is an open condition in any of the ATMs 1 when a service interruption is detected by the service interruption detection circuit, power of a backup power supply to be supplied until the door open/close detection sensor detects that the rear side door of the ATM 1 enters a closing condition.

In this instance, the station controller includes a time supervision section for supervising an elapsed time after detection of a service interruption by the service interruption detection circuit, and the control section stops the backup power supply and performs trouble ending processing if the detection condition of the door open/close detection sensor still is the open condition in the ATM 1 at a point of time when it is detected by the time supervision section that a fixed time elapses.

(22) The robot mechanism includes a receipt paper supplementing apparatus for supplementing receipt paper to any of the ATMs 1.

(23) The remaining money management system may further include a cassette station for defining a location at which the cash cassette is removed from or mounted onto the self-traveling robot. In this instance, the cassette station includes a locking mechanism for restricting removal of the cash cassette from the cassette station. Or, the cassette station may include a counting apparatus for counting an amount of cash in the cash cassette, and an automatic cash loading and extracting apparatus for automatically loading or extracting cash into or from the cash cassette.

(24) One of the ATMs 1 may be used as a location where the cash cassette is removed from or mounted onto the self-traveling robot. In this instance, an amount of cash in the cash cassette is counted making use of a counting function of the ATM 1 which is used as the location where the cash cassette is removed from or mounted onto the self-traveling robot, and a result of the counting is notified to the remaining money management apparatus 2.

With the remaining money management system described above with reference to FIG. 1, the following advantages can be anticipated.

(1) Since demanded cash amounts within a designated period of time for the ATMs 1 are calculated and predicted based on the transaction data in the past stored in the storage section 2A by the prediction section 2B and the demanded cash amounts for the ATMs 1 can be managed based on the predicted values, the remaining money management system is advantageous in that the amounts of money to be loaded can be compressed without depending upon a staff member and the burden to the staff member can be reduced significantly.

(2) Where the prediction section 2B converts a degree of influence of each of qualitative factors such as the day of the week or the date into a numerical value, demanded cash amounts can be predicted taking various qualitative factors into consideration, and consequently, prediction of a high degree of accuracy can be achieved. In this instance, where a prediction table is produced for each of the ATMs 1, a demanded cash amount can be predicted for each of the ATMs 1. Further, where the prediction table is produced for each of different currencies and after each fixed time interval, a demanded cash amount can be predicted for a currency or for a time interval designated by the staff member.

(3) Where the remaining money management apparatus 2 calculates an error between an actual result value of the demanded cash amount and the predicted value of the demanded cash amount obtained by the prediction section 2B and automatically selects, based on the error, a qualitative factor which has a significant influence on the demanded cash amount, the demanded cash amount can be predicted with a higher degree of accuracy.

(4) Where the prediction section 2B predicts demanded cash amounts for a period of time after the current money supplementing time until a next money supplementing time and outputs the predicted values as initial load money amounts to the ATMs 1, the staff member can determine an amount of money to be loaded initially upon starting of operation of the system without depending upon the experience or the sixth sense, but is required merely to supplementally load cash corresponding to the money amount into the ATMs 1. Consequently, the burden to the staff member can be reduced.

(5) Where the prediction section 2B normally supervises errors between actual remaining money amounts and the real time predicted values of demanded cash amounts for the ATMs 1 and calculates and indicates amounts of cash to be supplemented or extracted with which the errors are canceled, the staff member can cope with variations of remaining money amounts only by supplementing or extracting indicated amounts of cash without depending upon the experience or the sixth sense. Consequently, the burden to the staff member can be further reduced.

(6) A standard deviation of errors of the predictive error table (errors between actual result values of the demanded cash amount and predicted values of the demanded cash amount obtained by the prediction section 2B) is calculated as a standard error and a value calculated by multiplying the calculated standard error by a safety coefficient is added as a safety cash amount to the predicted value of the demanded cash amount. Accordingly, a safe money amount with which occurrence of stoppage of any ATM 1 arising from consumption of cash can be prevented with certainty can be determined without depending upon the experience or the sixth sense of the staff member. Consequently, the burden to the staff member can be further reduced.

(7) Where the prediction section 2B predicts, where a plurality of ATMs 1 are involved, a total demanded cash amount for the plurality of ATMs 1, calculates a sum total of actual result values of the demanded cash amount for the plurality of ATMs 1, calculates a safety cash amount for the entire plurality of ATMs 1 based on the predicted value and the actual result value thus calculated, and distributes the calculated safety cash amount at a distribution ratio to the ATMs 1, a safe money amount with which occurrence of stoppage of an ATM 1 arising from consumption of cash can be prevented with certainty can be determined without depending upon the experience or the sixth sense of the staff member. Further, since the plurality of ATMs 1 are treated as if they were a single ATM, a necessary minimum safety money amount can be determined, and the amount of money for the ATMs can be reduced with a higher degree of certainty.

(8) Where the prediction section 2B predicts an operable time based on a current remaining money amount, the staff member can refer to a result of the prediction to grasp till when each of the ATMs 1 can operate with the current remaining money amount, and consequently, the burden to the staff member can be reduced.

(9) Even if transaction data in the past are not cumulatively stored in the storage section 2A because the business office in which the system is incorporated is opened newly or a like case, where transaction data in the past of another business office which belongs to the same transaction form pattern are utilized as transaction data in the past of the business office, a demanded cash amount can be predicted with some degree of accuracy before transaction data for the business office are accumulated.

(10) Where, when an installation environment condition of the ATMs 1 changes, a predicted value of a sum of demanded cash amounts obtained by the prediction section 2B for the ATMs 1 before the change of the installation environment condition is distributed at a distribution ratio to the ATMs 1 after the change of the installation environment condition, demanded cash amounts for the ATMs after the change of the installation environment condition can be predicted with a high degree of accuracy using a data base produced from data accumulated for the individual ATMs 1.

(11) Where the remaining money management apparatus 2 discriminates one of the ATMs 1 which requires cash to be supplemented thereto and another one of the ATMs 1 which holds surplus cash and develops an instruction to extract cash from the ATM 1 which holds surplus cash and supplement the extracted cash to the ATM 1 which requires cash to be supplemented thereto, the staff member is required only to extract or supplement cash in accordance with the instruction, which further reduces the burden to the staff member. Further, after necessary amounts of cash are initially loaded into the plurality of ATMs 1, the cash in the ATMs can be circulated in the ATM group except when shortage of cash in the entire ATM group occurs. Consequently, loading of surplus cash can be eliminated, and accordingly, the amount of money for the ATM group can be compressed with a higher degree of certainty.

(12) Where the remaining money management system includes a self-traveling robot which automatically performs extraction and supplementing of cash from and to a relevant ATM 1, extraction or supplementing of cash from or to an ATM 1 need not be performed manually by the staff member at all, which can reduce the burden to the staff member remarkably. However, when the self-traveling robot is out of order or the extraction or supplementing of cash by the self-traveling robot is impossible, the staff member can refer to a displayed instruction to perform extraction or supplementing of cash can perform extraction and supplementing of cash from and to the ATM 1.

(13) Where an accounting system host computer or a remote supervision apparatus functions as the remaining money management apparatus 2 according to the present invention, the money amount can be managed for a plurality of business offices. Consequently, a demanded cash amount with regard to the entire money of the bank can be predicted to perform the remaining money amount management in units of a bank, and the money compression effect can be raised remarkably.

(14) Where the staff member terminal functions as the remaining money management apparatus 2 according to the present invention, there is no need of providing a remaining money management apparatus separately. Since the staff member terminal is normally used at a window by the staff member referring to a display thereon, when some instruction to any ATM 1 is developed, the instruction can be notified immediately to the staff member.

(15) Where transaction data of the staff member terminal are cumulatively stored into the storage section 2A and the prediction section 2B calculates and predicts demanded cash amounts within a designated time for the staff member terminal based on the transaction data, remaining money management for the entire money in the business office including the ATMs 1 and the staff member terminal can be realized.

(16) Where various instructions (an instruction for the self-traveling robot to move to an object one of the ATMs 1 and another instruction for the self-traveling robot to extract or supplement cash from or to the object ATM 1 (including an instruction to open or close an automatic door)) are conveyed from the robot control section of the remaining money management apparatus 2 to the self-traveling robot by a radio communication channel via a radio communication base station, an instruction can be sent with certainty to the self-traveling robot which moves around in the area in which the ATMs are arranged.

(17) Where a free track for the self-traveling robot is formed from a magnetic tape applied to a floor, there is no need of installing a rail or a like element, and even when a variation in layout is required upon removal of the ATMs 1 or installation of additional ATMs or where an obstacle such as a pillar is present between the ATMs 1, the travel path of the self-traveling robot can be arranged or varied very readily.

(18) Where the self-traveling robot detects any of the marking magnetic tapes applied along the free track, the self-traveling robot can specify the ATM instructed from the robot control section readily.

(19) Where the positioning apparatus compulsorily controls the stopping position of the self-traveling robot with respect to one of the ATMs 1, the distance between the ATM 1 and the self-traveling robot can be controlled to a fixed value, and consequently, automatic extraction or supplementing of cash from or to the ATM 1 can be performed with certainty by the robot mechanism. Further, by means of the arrival sensor provided for the positioning mechanism, it can be detected at which one of the ATMs the self-traveling robot arrives and stops.

(20) Where power is supplied from the feed terminals of the positioning mechanism to the collection terminals of the self-traveling robot only when the self-traveling robot stops and the robot mechanism must operate, there is no need of providing the power supply for the robot mechanism on the self-traveling robot. Consequently, the self-traveling robot can be formed light and compact, and power can be supplied very efficiently. In this instance, since power for the robot mechanism is supplied by a power supply circuit separate from the power supply circuit for supplying power to the ATMs 1, even if power supply interruption occurs and the ATMs 1 stop, power can be supplied separately from the power for the ATMs 1. Further, since the power for the robot mechanism is used to charge the self-traveling car power supply while the robot mechanism of the self-traveling robot is inoperative (in a waiting condition), the peak of current consumed by the self-traveling robot can be moderated.

(21) Where the arrival sensor supervision section of the station controller supervises detection conditions of the ATMs 1 by the arrival sensors in a concentrated manner and notifies the detection conditions by the arrival sensors as an operation condition of the self-traveling robot to the remaining money management apparatus 2, the operation condition of the self-traveling robot can usually be grasped by the remaining money management apparatus 2 side. Further, the staff member can control the operation condition of the self-traveling robot by manually operating the operation panel of the station controller. Furthermore, since an opening/closing condition of the rear side door of any of the ATMs 1 is supervised by the door opening/closing supervision section, the opening/closing condition of the rear side door of one of the ATMs 1 which is opened when a cash extracting or supplementing operation is to be performed for the ATM 1 by the self-traveling robot can always be grasped by the station controller side.

(22) The opening/closing condition of the rear side door of any of the ATMs 1 is controlled by the door control section so that, when a cash extracting or supplementing operation by the self-traveling robot is to be performed for the ATM 1, the rear side door is opened to provide a space sufficient to allow such cash extracting or supplementing operation to be performed therethrough.

(23) Where, when to open or close the rear side door of one of the ATMs 1 by the self-traveling robot, an optical signal for specifying the ATM 1 is emitted from the light emitting element toward the light receiving element of the ATM 1 and only when the light receiving element of the ATM 1 receives an optical signal specified for the ATM 1, the rear side door is opened by the door control section, the security against robbery required for an apparatus which handles cash is assured.

(24) Where, if a service interruption occurs while the rear side door of an ATM 1 is open and a cash extraction or supplementing operation by the self-traveling robot is proceeding, the control section of the station controller continues supply of power until the cash extraction or supplementing operation by the self-traveling robot comes to an end and the rear side door of the ATM 1 is closed, it can be prevented that the rear side door of the ATM 1 is left open, and the security against robbery required for an apparatus which handles cash is assured.

In this instance, the time supervision section supervises an elapsed time after detection of a service interruption by the service interruption detection circuit, and if the rear side door still remains in the open condition after the fixed time elapses, it is determined that some trouble has occurred, and the control section stops the backup power supply and performs trouble ending processing. Consequently, a countermeasure against the trouble can be taken.

(25) Where the robot mechanism includes a receipt paper supplementing apparatus, when a cash extraction or supplementing operation is performed by the self-traveling robot, supplementing of receipt paper to the ATM 1 can be performed at a same time, which contributes very much to reduction of the burden to the staff member.

(26) Where the remaining money management system includes a cassette station for removing or mounting the cash cassette from or onto the self-traveling robot, the staff member is required only to place the cash cassette, in which cash to be supplemented is loaded, in position into the cassette station or receive the cash cassette, in which extracted cash is held, in the cassette station.

In this instance, where removal of the cash cassette from the cassette station is restricted by the locking mechanism, the cash cassette cannot be pulled out inadvertently. Consequently, the safety against robbery is assured.

Further, where the counting apparatus of the cassette station counts an amount of cash in the cash cassette, the remaining money in the entire ATM group can be settled taking also the amount of cash in the cash cassette in the cassette station into consideration. Further, loading or extraction of cash into or from the ATM group can be performed automatically by the automatic cash loading and extracting apparatus of the cassette station.

(27) Where one of the ATMs 1 is used as a location where the cash cassette is removed from or mounted onto the self-traveling robot, there is no need of providing such a separate location as a cassette station as such location for removal or mounting of the cash cassette. In this instance, an amount of cash in the cash cassette can be counted making use of an original counting function of the ATM 1 which is used as the location for removal or mounting of the cash cassette, and where a result of the counting is notified to the remaining money management apparatus 2, the remaining money amount of the entire ATM group can be settled readily.

B. Embodiment of the Invention

An embodiment of the present invention will be described in detail below.

B-1. General Construction of the Remaining Money

Management System

Referring to FIG. 2, there is shown in block diagram a general construction of a remaining money management system to which the present invention is applied. The remaining money management system shown is constructed as a remaining money management system, for example, for a branch (business office) of a bank and includes a plurality of ATMs 11 installed in an automatic machine corner 10 and an additional ATM (lobby installed machine) 12 installed in a lobby of the business office.

Further, a universal banking terminal (UBT; staff member terminal) 19 which is handled by a clerk at the window as well as a safe 21 and a remaining money management terminal 18 are installed on the inner side of a counter 20 in the business office. The remaining money management terminal 18 includes a display unit (display screen) 18L, and a printer 23 is connected to the remaining money management terminal 18.

The ATMs 11 and 12, universal banking terminal 19 and remaining money management terminal 18 are connected for communication with each other by means of a local area network (LAN) 17.

The LAN 17 is connected to a public network (for example, an ISDN) 14 via an adapter (I-ADP) 15. The LAN 17 is further connected for communication with ATMs 13, which belong to the business office but are installed outside the business office, via the public network 14 and another adapter (I-ADP) 16.

The remaining money management terminal 18 collects transaction data from the ATMs 11 to 13 via the LAN 17 or/and public network 14 and manages remaining cash amounts (here, the numbers of currencies) corresponding to amounts of cash accommodated in the ATMs 11 to 13 based on the transaction data.

Each of the ATMs 11 and 12 includes a cash cassette 22 for accommodating cash therein. Though not shown in FIG. 2, a similar cash cassette is provided also in each of the ATMs 13. Further, each of the ATMs 11 to 13 is connected to an accounting system host computer (refer to FIG. 37) in a computer center via a terminal controller (HOST/TC) and a dedicated line not shown.

B-2. Construction of the Remaining Money Management Terminal

Referring to FIG. 3, there is shown in block diagram a detailed functional construction (software construction) of the remaining money management terminal employed in the present embodiment. The remaining money management terminal 18 includes a pair of LLC drivers 18A, a LAN controller 18B, a data communication controller 18C, a remaining money management controller 18D, a remaining money management library 18E, a cash amount management section 18F, a hard disk 18G, a self-traveling car controller 18H, an electronic journal controller 18I, a printer controller 18J and a system library 18K.

Though not shown in FIG. 3, the remaining money management terminal 18 includes a display unit 18L (refer to FIG. 2 or 39), a keyboard 18N (refer to FIG. 39) and a LAN adapter 18P (refer to FIG. 39), and the printer 23 is connected to the remaining money management terminal 18. Further, the body of the remaining money management terminal 18 is actually formed from, for example, a personal computer 18M (PC) as seen in FIG. 39. It is to be noted that, in the present embodiment described below, unless otherwise specified, the cash (or money) amount is treated as a number or quantity of a representative (highest) currency (such as a 10,000 yen bill in Japan).

The LLC drivers 18A are connected to the LAN 17 so as to transmit or receive a signal to and from the LAN 17. The LAN controller 18B controls driving of the LLC drivers 18A to perform connection control or transmission or reception control of a message to or from the LAN 17. The data communication controller 18C controls the LAN controller 18B to control transmission or reception of data.

The remaining money management controller 18D controls the entire remaining money management terminal 18 in the present embodiment. To this end, the remaining money management controller 18D includes a main controller 18a for controlling the other components described below, a supplement/extraction section 18b for delivering an instruction for supplement of cash or for extraction of cash to a staff member or the self-traveling car controller 18H in order to realize circulation of cash in an ATM group which will be hereinafter described, a staff member management section 18c for delivering an instruction for careful examination to a staff member in the bank or setting a schedule peculiar to the business office, a remaining money management section 18d for delivering an instruction for prediction to the remaining money management library 18E in response to screen inputting processing (refer to FIGS. 20 and 21) by a staff member and controlling the screen display of a result of prediction processing (display control of the display unit 18L), and a maintenance management section 18e for performing ATM setting processing (setting of a machine type, inside/outside the business office, a machine number and so forth) which is processing corresponding to maintenance by a customer engineer (CE) or dumping processing.

The remaining money management library (prediction section) 18E performs calculation of remaining money in the ATMs 11 to 13 or prediction of cash amounts demanded by the ATMs 11 to 13 in response to an instruction from the remaining money management section 18d. The remaining money management library 18E has a function of writing, after transaction data are collected from the ATMs 11 to 13, the transaction data into a predetermined area of the hard disk 18G as hereinafter described with reference to FIG. 4 and another function of calculating and predicting, after a prediction period is designated by screen inputting processing by a staff member, demanded cash amounts within the prediction period based on transaction data of the ATMs 11 to 13 in the past stored in the hard disk 18G. The remaining money management library 18E further has various functions which will be hereinafter described with reference to FIGS. 5 to 21.

The cash amount management section 18F manages the amounts of cash for different currencies currently present in the ATMs 11 to 13 and manages, where a cassette station (refer to reference numeral 53 of FIG. 38) is provided, the amounts of cash for different currencies currently present in the cassette station. Further, when necessary, the cash amount management section 18F writes the thus managed amounts of cash into an apparatus information file in the hard disk 18G.

Into the hard disk (apparatus information file, storage section) 18G, transaction data of the ATMs 11 to 13 are written by the remaining money management library 18E as described above, and also the amounts of cash in the ATMs 11 to 13 and the cassette station are written by the cash amount management section 18F.

As transaction data, for example, such data 1 to 6 as listed below are periodically collected from control sections 24 of the ATMs 11 to 13 by the data communication controller 18C and stored into the hard disk 18G from the remaining money management library 18E or the cash amount management section 18F. It is to be noted that a detailed processing procedure of data collection will be hereinafter described with reference to FIG. 7.

1 Apparatus specifying data: bank number/branch number/apparatus number

2 Transaction date/time: year/month/day/transaction start time/transaction end time

3 Transaction type: delivery/acceptance/transfer/postal transfer/inquiry about the balance

4 Delivered transaction amounts: ten thousand yen bill/five thousand yen bill/one thousand yen bill/coins

5 Accepted transaction amounts: ten thousand yen bill/five thousand yen bill/one thousand yen bill/coins

6 Remaining amounts in the apparatus: ten thousand yen bill/five thousand yen bill/one thousand yen bill/coins

Those transaction data are stored, for example, in such a directory structure as shown in FIG. 4 in the hard disk 18G. In particular, in a portion of the hard disk 18G specified by the bank name/branch name/apparatus name (apparatus number), a day demanded money amount storage area, a past demanded money amount storage area and an in-apparatus remaining money information storage area are set.

Since data of the day are not used for prediction, contents of the data of 2 to 6 collected on the day are stored as they are into the day demanded currency number storage area, and then, after the date changes, they are saved into the past demanded currency number storage area. When the data are saved in this manner, they are converted into data of a form in which demanded money amount prediction processing which will be hereinafter described may be performed readily using the data. Thus, in the past demanded current number storage area, for example, past data of 2 to 5 for one year (365 days) in the past taken after each 10 minutes are stored in a classified condition as seen in FIG. 4. Further, in the in-apparatus remaining money information storage area, remaining money for the individual currencies held in the ATM at present are written. It is to be noted that illustration of data for the five thousand yen bill is omitted in FIG. 4.

The self-traveling car controller 18H functions only where the remaining money management system includes a self-traveling car system (refer to FIG. 38). Although details are hereinafter described, the self-traveling car controller 18H has an automatic machine controlling function of controlling opening and closing movements of automatic doors (rear doors) of the ATMs 11 to 13, a self-traveling car controlling function of controlling operation of a self-traveling car (refer to reference numeral 50 in FIG. 38), and, where the remaining money management system includes a cassette station (refer to reference numeral 53 of FIG. 38), a cassette ST controlling function of controlling operation of the cassette station (cassette ST).

The electronic journal controller 18I manages the changed amounts of money for different currencies. The electronic journal controller 18I keeps the changed amounts of money for different currencies as an electronic journal on a hard disk or the like and delivers an instruction to output the electronic journal by printing.

The printer controller 18J actually controls, when a printing outputting instruction from the remaining money management controller 18D or an instruction to output an electronic journal by printing from the electronic journal controller 18I is received, operation of the printer 23 using the system library 18K to perform printing outputting in accordance with the printing outputting instruction.

It is to be noted that, in FIG. 3, reference numeral 24 denotes a control section provided in each of the ATMs 11 and 12, and the ATMs 11 and 12 are connected to the LAN 17 via the control sections 24. Also the ATMs 13 outside the business office are connected to a communication network via a similar control section 24.

B-3. Prediction Method for Demanded Amounts of Money

A method of predicting demanded amounts of money executed by the remaining money management library 18E will be described with reference to FIGS. 5 to 21.

The remaining money management library 18E in the present embodiment predicts the demanded amounts of bills (demanded amounts of money) which are predicted to be consumed by customers using a prediction method called Hayashi's quantization theory.

The concept of the prediction method will be described with reference to FIGS. 5 and 6. Basically, the prediction method predicts quantitative data (demanded amounts of money) from qualitative data (qualitative factors) such as a day of the week, the date or the like. In particular, as seen from FIG. 5, a total average value of demanded amounts of money in the past are calculated based on transaction data in the past stored on the hard disk 18G, and for each qualitative factor (day of the week, date, a specific day or the like) which may have an influence on the demanded amounts of money, a degree of influence of the qualitative factor on the demanded amounts of money is converted into a numerical value in advance. Then, a sum value of the average value and the numerical value data (a detailed calculation process will be hereinafter described) of the degrees in influence of the qualitative factors are calculated as a predicted value of the demanded amounts of money.

An example of exemplary values is illustrated in FIG. 6. Referring to FIG. 6, as items of the qualitative factors, a day of the week, the date and a specific date (payday, bonus day and so forth) are set. Then, for the day of the week of the item 1, numerical value data of the degrees in influence of the individual days of the week from Monday to Sunday are calculated (500, 600, -300, 400, 800, -600, and -700), and for the date of the item 2, numerical value data of the degrees in influence, for example, of different weeks (first to sixth weeks in a month) are calculated (500, -300, -450, 550, 1,100, 450). Further, for the specific date of the item 3, numerical value data of the degrees in influence of the payday, a day or days following the payday (the next day or days to the payday) and the other days are calculated (2,300, 1,550, -850). Then, when it is intended to predict the demanded amounts of money in a day or days following the payday of Monday in the six week, the numerical value data 500 of Monday, the numerical value data 450 of the sixth week, the numerical value data 1,550 of a day or days after the payday and the average value are added as seen in FIG. 6. Consequently, the predicted value of 3,000 for the demanded number of currencies for Monday in the sixth week after the payday is obtained.

The prediction processing for the demanded amounts of money will be described in more detail with reference to FIGS. 7 to 16.

A procedure of collection processing of transaction data prior to prediction will first be described with reference to the flow chart (steps S1 to S13) shown in FIG. 7.

The data communication controller 18C of the remaining money management terminal 18 normally operates in a timer monitor processing loop (step S1) and discriminates whether or not a data collection time comes (step S2). If a data collection time which comes, for example, after each 10 minutes comes, the data communication controller 18C transmits a data collection inquiry to the ATMs 11 to 13 (step S3). The data collection inquiry is transmitted via the LAN controller 18B and one of the LLC drivers 18A to the ATM control section 24 of each of the ATMs 11 to 13. Then, the data communication controller 18C enters into a data reception waiting loop operating condition (step S4). After the data communication controller 18C receives data from the ATMs 11 to 13 (YES at step S5), it passes over the data to the main controller 18a.

The main controller 18a analyzes the data received from each of the ATMs 11 to 13 (step S6) and discriminates whether or not the received data is transaction data (step S7). When the received data is not transaction data (NO at step S7), the main controller 18a passes over the data to the supplement/extraction section 18b, the staff member management section 18c or the maintenance management section 18e in accordance with the contents of the received data (step S8). On the contrary when the received data is transaction data (YES at step S7), the main controller 18a passes over the transaction data to the remaining money management library 18E via the remaining money management section 18d.

The remaining money management library 18E analyzes the received transaction data and distributes the transaction data in accordance with a result of the analysis (step S9). When the received transaction data is demanded money amount information (step S10), the transaction data is written into the day data storage area (day demanded money amount storage area in FIG. 4) of the hard disk 18G from the remaining money management library 18E, and after the date changes, the transaction data is saved into the past data storage area (past demanded money amount storage area of FIG. 4) (step S11). On the other hand, when the received transaction data is apparatus remaining money information (step S12), the transaction data is passed over to the cash amount management section 18F. Consequently, the transaction data is written into the in-apparatus remaining money information storage area (refer to FIG. 4) of the hard disk 18G by the cash amount management section 18F (step S13).

Based on the transaction data in the past collected by such data collection processing as described above, in the present embodiment, a demanded amount of money within a designated period in each of the ATMs 11 to 13 is predicted basically by the following procedures 1 to 3.

›Procedure 1!

First, the transaction data in the past stored on the hard disk 18G are first tabulated for the individual qualitative factors, and an average value of the demanded money amounts in the past is calculated. Here, as items of the qualitative factors, for example, as seen in FIG. 8, 1 a day of the week (Monday to Sunday), 2 the date (first to sixth week in a month), 3 a particular date (a day at the end or the beginning of a month/fifth/10th day/payday/day or days before and after the payday/a bonus day or the like), 4 a date (first to 30th day/ever 3rd day/every fifth day), 5 the end or the beginning of a year/summer vacation/Christmas holidays and so forth, 6 the season (spring, summer, autumn, winter), 7 the weather (fine/rain/snow or the like) and so forth may be used. It is to be noted that the following description relates to a case wherein the factors 1 to 3 are selected as items of the qualitative factors. Further, for the particular date 3 the payday/a day or days following the payday (after the payday)/any other day are used as the qualitative factors.

Then, as seen from FIG. 9, data of those items corresponding to the individual qualitative factors are collected from among the transaction data in the past of the hard disk 18G (all data from the starting date of data collection to yesterday), and for each qualitative factor, totals of demanded amounts of money are calculated (if the day is Monday, then totals of demanded amounts of money in Mondays in the past are calculated). Or, simultaneously, an average value (total average value) of the demanded amounts of money in the past is calculated.

›Procedure 2!

Then, such a cross tabulation table of qualitative factors as shown in FIG. 10 is produced. The table shown in FIG. 10 indicates a result of tabulation of how many times two different qualitative factors have been satisfied simultaneously in the past. For example, with regard to Monday, from FIG. 10, it can be seen that the two factors that the day is Monday and that the date belongs to the first week have been satisfied three times in the past, that the two factors that the day is Monday and that the date belongs to the sixth week have been satisfied three times in the past, and that the two factors that the day is Monday and the specific date is the payday have been satisfied once in the past. Also the other numbers of times have similar meanings.

Then, the values (category scores) of the individual factors are defined, for example, in such a manner as illustrated in FIG. 11, and those values are applied to the values tabulated in the cross tabulation table shown in FIG. 10 to produce such simultaneous equations as given below. ##EQU1##

The simultaneous equations above are solved and a weighted mean is calculated for each item (the day of the week, the data, the specific date). The difference between the weighted mean and the corresponding qualitative factor makes numerical value data of the degree of influence of the qualitative factor. As a result, qualitative data is converted into quantitative data for each factor. For example, the numerical value data x.sub.11 of the degree of influence regarding Monday is calculated as +230, and the numerical value data x.sub.17 of the degree of influence regarding Sunday is calculated as -200. The significance of x.sub.11 =+230 calculated in this manner will be considered. From FIG. 12, it can be seen that Monday has a positive action of increasing the demanded amount of money by 230 from an average value of the entire demanded money amount. It is to be noted that, in FIG. 12, an entire distribution of demanded money amounts and a distribution of demanded money amounts of Monday are shown, and the position of a peak of each distribution is an average value of the distribution. Accordingly, the result of calculation of X.sub.17 =-200 indicates that the Sunday has a negative action of decreasing the demanded money amount by 200 from the entire average value of demanded money amount.

Numerical value data of the degrees of influence for the individual factors (values of the factors, scores of the factors) are calculated for each of the ATMs 11 to 13, and such a prediction table as shown in FIG. 13 is produced. Where n ATMs of the first to nth machines are involved, a prediction table is produced for each ATM, that is, totaling n prediction tables are produced as seen in FIG. 13.

›Procedure 3!

Then, demanded money amount prediction is performed by collecting, from among the scores of the individual factors of such a prediction table as shown in FIG. 13, scores of a factor corresponding to the day or the time which makes an object of prediction. For example, in order to predict the demanded money amount of one thousand yen bills within a period from 9:00 am to 10:00 am on the sixth day of February (Monday, second week, other than the payday) for the No. 1 ATM, such a prediction table for the period of 9:00 am to 10:00 am as shown in FIG. 14 is obtained, and from an area of the prediction table for the one thousand yen bill and the No. 1 machine, the score +230 of Monday, the score -30 of the second week and the score -500 of a day other than the payday (those scores are netted in FIG. 14) are read out. In this instance, if the result of calculation of the average demanded money amount in the past is 2,000, then the demanded money amount predicted value to be obtained is given by the following equation: ##EQU2##

In the following, a flow of the demanded money amount prediction processing will be described with reference to the flow chart (steps S21 to S32) shown in FIG. 15.

After the system of the present embodiment is rendered operative by turning on of power (power on) (step S21), the remaining money management library 18E is started (step S22). Then, accessing of the remaining money management library 18E to the hard disk 18G is started (step S23). After the remaining money management library 18E reads out transaction data in the past up to yesterday from the hard disk 18G, it produces a prediction table through the procedure 1 and the procedure 2 described above (step S24), and then waits for key inputting by a staff member (inputting of an instruction to predict a demanded cash amount) (step S25).

Then, if the remaining money management section 18d discriminates that key inputting has been performed by a staff member (YES at step S26), the remaining money management section 18d discriminates whether or not inputting processing of a prediction period has been completed regularly (step S27). When inputting processing of a prediction period has been completed regularly (YES at step S27), the remaining money management section 18d delivers the prediction period data to the remaining money management library 18E.

Upon reception of the prediction period data, the remaining money management library 18E first calculates a predicted value of the demanded cash amount for the prediction period in accordance with the procedure 3 described hereinabove (step S28) and then acquires current in-apparatus remaining money information from the cash amount management section 18F (step S29), whereafter it executes calculation processing of a supplement or collection money amount (step S30). In particular, at step S30, the value of ›predicted value of the demanded money amount for the prediction period (result of the calculation at step S28)!-›in-apparatus remaining money at present! is calculated. If the value is in the positive (+), then this signifies that extraction is required, but if the value is in the negative (-), then this signifies that supplement is required. After the calculation processing at step S30, the remaining money management library 18E delivers the result of prediction by step S28 and the result of calculation of the supplement or extraction money amount by step S30 to the remaining money management section 18d.

Upon reception of the data from the remaining money management library 18E, the remaining money management section 18d causes the data to be displayed on the screen of the display unit 18L of the remaining money management terminal 18 (step S31) and, when necessary, edits print data (step S32). Then, the remaining money management library 18E sends the print data to the printer controller 18J so as to execute printing processing.

As a result of the screen display processing at step S31, such a screen display as shown, for example, in FIG. 20 or 21 is provided on the display unit 18L (display screen). In particular, on the display unit 18L shown in FIG. 20, information of a particular ATM is displayed, and particularly, for example, remaining money information at present for the individual currencies, operable times for the individual currencies (which will be hereinafter described with reference to FIGS. 26 and 27), a result of prediction from the date and hour of the start of prediction to the date and hour of next supplement/extraction and so forth are displayed. On the other hand, on the display unit 18L shown in FIG. 21, information regarding four ATMs (the No. 1 to No. 4 machines) is displayed. For each of the ATMs, remaining money information at present for the individual currencies, a result of prediction from the date and hour of the start of prediction to the date and hour of next supplement/extract and so forth are displayed, and also a result of prediction for all of the four ATMs is displayed for the individual currencies.

Here, a process of calculating a predicted value within a designated period (designated time) based on a prediction table will be described. FIG. 16 shows a prediction table produced after each predetermined time interval (after each one hour) for each currency, and how to calculate a predicted value within the designated period based on a predicted value of a demanded money amount calculated using the prediction table will be described below.

Here, it is assumed that a linear prediction method is adopted and, if predicted values of individual prediction tables for a certain designated period are totaled, then predicted values of the demanded cash amounts for the entire designated period are obtained.

Accordingly, in the prediction method of the remaining money management system in the present embodiment, prediction of the demanded money amounts within a designated period is performed in accordance with the following procedure.

First, a predicted value for each prediction table corresponding to a prediction factor (determined based on a designated year, month and day) is calculated. In particular, if the time interval between the prediction tables is one hour as seen in FIG. 16, then predicted values at time intervals of one hour are calculated in advance.

Then, when the magnitude of the designated period and the magnitude of the time zone (predetermined time interval) of the prediction tables coincide with each other, predicted values for all of the prediction tables included in the designated period are totaled to calculate a predictive value within the designated period. For example, if the designated period is 9:00 to 11:00, then

›predicted value for 9:00 to 11:00!=›predicted value for 9:00 to 10:00!+›predicted value for 10:00 to 11:00!

On the other hand, when the magnitude of the designated period and the magnitude of the time zone (predetermined time interval) of the prediction tables do not coincide with each other, a ratio at which the designated period occupies in the predetermined time interval between the prediction tables is calculated, and then a predicted value corresponding to the ratio is added to calculate a predicted value within the designated period. For example, if the designated period is 10:45 to 11:30, then

›predicted value for 10:45 to 11:30!=›predicted value for 10:00 to 11:00!/4+›predicted value for 11:00 to 12:00!/2

In the equation above, ›predicted value for 10:00 to 11:00!/4 is a predicted value for the period of 10:45 to 11:00, and ›predicted value for 11:00 to 12:00!/2 is a predicted value for the period of 11:00 to 11:30. By adding them, a predicted value within the designated period of 10:45 to 11:30 can be calculated.

Since the demanded money amounts within a designated period for the ATMs 11 to 13 are calculated and predicted based on the transaction data in the past stored on the hard disk 18G by the remaining money management library 18E and the demanded money amounts for the ATMs 11 to 13 can be managed based on the predicted values as described above, the amounts of money to be loaded can be compressed without depending upon a staff member and the burden to the staff member can be reduced remarkably. In this instance, in the present embodiment, since a qualitative factor such as a day of the week, the date or the like can be converted into a numerical value as a degree of influence, demanded amounts of money can be predicted taking various qualitative factors into consideration, and consequently, prediction with a high degree of accuracy can be achieved.

The remaining money management terminal 18 in the present embodiment is further provided with such various functions as described below.

›Automatic Factor Selection Function!

The automatic factor selection function is a function of automatically selecting the most effective factor or factors and a combination of such factors. Particularly, while various qualitative factors are available as described hereinabove with reference to FIG. 17, the automatic factor selection function automatically selects, as types of qualitative factors to be used for calculation of predicted values of demanded money amounts by the remaining money management library 18E, types of qualitative factors with which errors between actual values of demanded money amounts and predicted values of demanded money amounts obtained by the remaining money management library 18E may be in the minimum (types of qualitative factors which have a significant influence on the demanded cash amount; for example, 1 to 3 and so forth in FIG. 17). Consequently, the errors between actual values of demanded money amounts and predicted values of demanded money amounts obtained by the remaining money management library 18E can be minimized, and prediction of demanded money amounts can be performed with a higher degree of accuracy.

›Initial Value Prediction Function!

The remaining money management terminal 18 has an initial value prediction function of predicting, before money is supplemented to the ATMs 11 to 13, demanded cash amounts for a period from a present money supplementing time (prediction starting time) to a next money supplementing time and outputting the predicted values as initial charge currency amounts to the ATMs 11 to 13.

FIG. 18 is a graph illustrating an example of a variation of the remaining money amount of an ATM. As seen from FIG. 18, after operation of the system is started (after the business of the business office is started), demanded money amounts for a period after a supplementing time by a staff member of the day till a next supplementing time by a staff member (usually the same hour of the next day) are predicted by the remaining money management library 18E. Then, a staff member uses the predicted values by the remaining money management library 18E as aimed values of initial load money amounts to the ATMs 11 to 13 of the day.

Consequently, the staff member can determine an initial load money amount upon starting of operation or a like case without relying upon the experience or the sixth sense of the staff member and may supplementally load cash according to the initial load money amount into the ATMs 11 to 13. Accordingly, the burden to the staff member can be reduced remarkably

›Remaining Money Amount Variation Prediction Function!

The remaining money management terminal 18 has a remaining money amount variation prediction function of predicting demanded amounts of money for the ATMs 11 to 13 per unit time to predict demanded money amounts till a next collective supplementing time on the real time basis, normally supervising errors between actual remaining amounts and real time predicted values, calculating amounts of cash to be supplemented or collected with which the errors are to be minimized and displaying the thus calculated money amounts on the display unit 18L.

For example, each time a periodical supplement/extraction time ta or tb comes as shown in FIG. 19, a supplement/extraction money amount which is to minimize the error between a real time predicted value (predicted demanded money amount till a next collective supplementing time) and an actual remaining money amount by the remaining money management library 18E is calculated by the remaining money management library 18E and displayed on the display unit 18L. At the periodical supplement/extraction time ta, the error has a negative value, and an instruction to supplement the cash of an amount equal to an absolute value of the error (money amount to be supplemented) into the ATM is displayed on the display unit 18L. On the other hand, at the periodical supplement/extraction time tb, the error has a positive value, and an instruction to extract cash of an amount equal to the error (money amount to be extracted) is displayed on the display unit 18L. From the display, a staff member can discriminate any ATM in which the remaining money is short or excessive and can discriminate such short or excessive amount of money at a glance.

Consequently, if the demanded cash amount is varied by the weather or a like factor after the cash of an initial load money amount predicted by the initial prediction function is loaded, a staff member can correct the money amount readily based on the variation. In other words, the staff member can cope with a variation of the remaining money amount only by supplementing or extracting the cash of the instructed amount. Consequently, the burden to the staff member is reduced remarkably.

B-4. Safety Money Amount Determination Method

While prediction of a demanded cash amount is performed in such a manner as described above by the remaining money management system of the present embodiment, when cash is to be supplemented or extracted actually using the predicted value, a safety money amount is added to a predicted value of a demanded cash amount in order to prevent occurrence of inadvertent stoppage of operation of an ATM by consumption of cash which arises from an error of prediction taking the accuracy in prediction into consideration. In other words, a cash amount obtained by adding a safety cash amount to a predicted value of a demanded cash amount calculated by the remaining money management library 18E is used as an appropriate cash amount adapted upon actual supplement or extraction of cash.

Such a safety cash amount is determined, in ordinary operation of a bank, based on the experience of a staff member in the past. Where the determination relies upon the experience of a staff member in this manner, the value of the safety cash amount exhibits a variation among different staff members, and this makes a problem in accuracy. Besides, since some experience is required for the determination, a heavy burden is applied to the staff member.

Therefore, in the present embodiment, an appropriate safety cash amount can be calculated and determined using the remaining money management library 18E and the hard disk 18G of the remaining money management terminal 18 as described below with reference to FIGS. 22 to 25.

As shown in FIG. 22, a prediction data base (refer to 1 of FIG. 22) is stored on the hard disk 18C, and the remaining money management library 18E reads out a table (refer to 2 of FIG. 22) between a predicted value and an actual value of a certain prediction factor. Here, in the present embodiment, a minimum unit of prediction designated by three items of an ATM, a currency type and a time zone is used as a prediction factor, and a required prediction value is calculated as a sum total of the prediction factors. It is to be noted that FIG. 22 illustrates a case wherein a safety coefficient of a certain prediction factor is calculated and determined. Further, the table 2 has, for example, predicted values for 366 days and actual values corresponding to the predicted values stored in pair therein.

When an actual value is not stored in the table 2 or the storage date of an actual value is later than the storage date of a corresponding predicted value, a new predicted value is calculated and stored into the table 2.

Then, errors between predicted values and actual values stored in pair in the table 2 are calculated, and a prediction error table (refer to 3 in FIG. 22) in which the errors are set is produced. In this instance, flags of the individual prediction errors are written into the prediction error table 3 in accordance with the following rule. In particular, "0" is placed as the value of a flag where an actual value is absent because of a holiday and consequently a prediction error cannot be calculated whereas "1" is placed as the value of a flag where an actual value has been calculated.

A standard deviation of those of prediction errors stored in the prediction error table 3 produced in this manner whose flags have the value of "1" is calculated, and the standard deviation thus calculated is determined as a standard error (refer to 4 in FIG. 22).

It is assumed that the frequency distribution of prediction errors exhibits a normal distribution centered at the position of the "prediction error=0", and a safety cash amount which corresponds to a probability in which an ATM does not suffer from stoppage because of consumption of cash is calculated in accordance with the following equation:

›safety cash amount!=›standard error!.times.›safety coefficient!

For example, in order to suppress the cash consumption stoppage probability to a value lower than 1%, the safety coefficient is set to "3" from a normal distribution-table, and 3.times.›standard error! is added as a safety cash amount to the demanded cash amount.

While a case wherein a safety coefficient is calculated for a certain one prediction factor is described in the foregoing, another case wherein a safety coefficient for two or more prediction factors or for two or more designated ATMs will be described below.

In the present embodiment, a predicted value of a demanded cash amount for a plurality of prediction factors or for a plurality of ATMs is calculated as a sum value of predicted values for the individual prediction factors (›predicted value of demanded cash amount within designated period!=.SIGMA.›predicted values of individual prediction factors within designated period!) or a sum value of predicted values for the individual designated ATMs (›predicted value of total demanded cash amount of two or more designated ATMs!=.SIGMA.›predicted values of individual designated ATMs!).

Then, when X=x.sub.1 +x.sub.2 + . . . +x.sub.n, from a statistical formula of a dispersion V(X)=V(x.sub.1)+V(x.sub.2)+ . . . +V(x.sub.n), also the square of a standard error of each of the individual predicted values can be calculated readily from

›square of standard error of predicted value of each prediction factor within designated period!=.SIGMA.›square of standard error of predicted value of each predicted error within designated period!

or

›square of standard error of predicted value of total demanded cash amount of two or more designated ATMS!=.SIGMA.›square of standard error of predicted error of each designated ATM!

Also a safety cash amount of a predicted value of a demanded cash amount for a plurality of prediction factors or for a plurality of ATMs (=›safety coefficient!.times.›standard error!) can be calculated readily.

In this manner, in the present embodiment, since a standard error for each prediction factor (qualitative factor described hereinabove) which can be designated by a parameter such as, for example, an ATM, a currency type or a time zone is calculated in advance, a safety money amount for a predicted value for an arbitrary period or for a total predicted value for arbitrary ATMs can be calculated readily.

Accordingly, since the safety money amount which is required for operation of the system can be calculated and determined quantitatively and besides such calculation and determination does not rely upon the experience, the sixth sense or the like of a staff member, it can be suppressed to a minimum level, and this contributes to reduction in burden to a staff member and to compression of money. Further, since the safety money amount can be determined based on a probability in accordance with an operation condition, stoppage of an ATM because of consumption of cash can be prevented with certainty. Furthermore, since also a safety cash amount for a total demanded money amount of two or more ATMs can be calculated readily, the safety money amount can be determined quantitatively when a plurality of ATMs are managed as a group using a self traveling car (self traveling robot) or the like which will be hereinafter described. In addition, also a safety cash amount for a time width of a designated prediction period can be calculated and determined readily.

By the way, in order to determine a safety money amount for two or more ATMs, a safety money amount is calculated for each of the ATMs in the process described above. However, from reasons hereinafter described with reference to FIGS. 23(a) to 25, when a safety money amount is determined for two or more ATMs, further compression of required money can be achieved by using the following method.

In particular, in the present embodiment, the remaining money management library 18E predicts, for example, regarding all of the ATMs 11 of the automatic machine corner 10 as a single ATM (group control), a total demanded cash amount, and produces a table 2 shown in FIG. 22 wherein the predicted value and a sum total of actual values of demanded cash amounts of all of the ATMs 11 are included in pair. Then, regarding all of the ATMs 11 of the automatic machine corner 10 as a single ATM, a prediction error table 3 is produced and a standard deviation of the error set in the prediction error table 3 is calculated as a standard error 4 in a similar manner as in the method described hereinabove with reference to FIG. 22. The safety cash amount for all of the ATMs 11 calculated in this manner is distributed to the ATMs 11 at a predetermined distribution ratio.

In this instance, the distribution method can be selected from among three methods including (i) a method wherein the calculated safety money amount is distributed equally to all of the ATMs 11, (ii) another method wherein the safety money amount is distributed to the ATMs 11 at ratios of safety cash amounts calculated for the individual ATMs 11, and (iii) a further method wherein the safety money amount is distributed at ratios of the squares of safety cash amounts calculated for the individual ATMs 11.

Subsequently, a difference between a case wherein a safety coefficient is determined regarding all of the ATMs 11 of the automatic machine corner 10 as a single ATM (that is, a case wherein group control is performed; refer to FIGS. 23(a) and 23(b)) and another case wherein a safety coefficient is determined based on predicted values for the individual ATMs 11 (that is, a case wherein no group control is performed; refer to FIGS. 24(a) and 24(b)) will be described with reference to FIGS. 23(a) to 25.

When such group control as described above is performed for n ATMs 11 as seen in FIG. 23(a) (in FIG. 23(a), n=10), where the standard error of the entire group is represented by .pi..sub.G and the standard errors of the individual ATMs 11 are represented by .sigma..sub.1, .sigma..sub.2, . . . , .sigma..sub.n,

.sigma..sub.G.sup.2 =.sigma..sub.1.sup.2 +.sigma..sub.2.sup.2 + . . . +.sigma..sub.n.sup.2

By the way, if it is assumed that the standard errors .sigma..sub.1, .sigma..sub.2, . . . , .sigma..sub.n of the ATMs 11 are all equal to each other, then

.sigma..sub.G.sup.2 =n.sub..sigma..sup.2

and the standard error .sigma..sub.G of the entire group is given by ##EQU3## If the safety coefficient is set to 3 as seen in FIG. 23(b), then the safety money amount of the entire group is given by ##EQU4##

On the other hand, when no group control is performed for the n ATMs 11 as seen in FIG. 24(a) (in FIG. 24(a), n=10), where the standard errors of the ATMs 11 are represented by .sigma..sub.1, .sigma..sub.2, . . . .sigma..sub.n and the sum total of the standard errors is represented by .sigma..sub.N,

.sigma..sub.N =.sigma..sub.1 +.sigma..sub.2 + . . . +.sigma..sub.n

Then, if it is assumed that all of the standard errors .sigma..sub.1, .sigma..sub.2, . . . , .sigma..sub.n of the ATMs 11 have an equal value .sigma. similarly as described above, then the sum total .sigma..sub.N of the ATMs 11 is given by

.sigma..sub.N =n.times..sigma.

If the safety coefficient of the ATMs 11 is set to 3 as seen in FIG. 24(b), then the sum of the safety cash amounts of the n ATMs 11 is given by

›sum of safety cash amounts of n ATMs!=3n.times..sigma.

Accordingly, comparison between the total safety cash amount 3.sqroot.n.times..sigma. a when group control is performed and the total safety cash amount 3n.times..sigma. when no group control is performed apparently proves that, as seen in FIG. 25, as the number of the ATMs 11 increases, the total safety cash amount 3.sqroot.n.times..sigma. when group control is performed becomes lower and the difference between the compressed cash amounts increases. For example, where n=10 as seen in FIGS. 23(a), 23(b) and 24(a), 24(b), the compression difference is given by the following expression: ##EQU5## In short, where 10 ATMs 11 are provided, save in cash amount by about 2.sub..sigma. can be achieved by group control comparing with the cash amount required when no group control is performed.

In this manner, when a safety coefficient is to be determined for two or more ATMs, a necessary minimum safety money amount can be determined quantitatively by performing such group control as described above and handling a plurality of ATMs regarding them as a single ATM, and the required cash amount for the ATM group can be compressed with a higher degree of certainty.

B-5. Operable Time Prediction Function

In the remaining money management system of the present embodiment, the remaining money management library 18E has also such an operable time prediction function as described below. The operable time prediction function will be described below with reference to FIGS. 26 and 27.

The operable time prediction function of the remaining money management library 18E predicts, as seen in FIG. 26, a predicted value of a demanded cash amount of each of the ATMs 11 to 13 after each predetermined time interval Tu from the present point of time and determines a time, at which the predicted value and a present remaining cash amount of the ATM 11, 12 or 13 coincide with each other, as an operable time t0.

Prediction of an operable time of a certain ATM 11 will be described in detail with reference to a flow chart (steps S41 to S47) shown in FIG. 27.

First, the value n for setting the number of predetermined time intervals Tu mentioned hereinabove (prediction factors, blocks; a unit of prediction time, for example, 10 minutes) is set to 0 (step S41), and the value n is incremented by one at next step S42. Thereafter, the following processing is performed for n=1.

It is discriminated whether the nth (n=1 at first) block is within an operation time of the system (step S43), and if the nth block is not within an operation time, the block setting value n is further incremented by one at step S42. On the contrary, if the nth block is within an operation time, then a demanded cash amount after lapse of time of n.times.Tu after the present point of time is predicted by the remaining money management library 18E (step S44).

Then, the value predicted at step S44 and a remaining cash amount at present in the ATM 11 are compared with each other to discriminate whether or not ›remaining money amount!<›predicted value! (step S45). If the remaining cash amount is equal to or higher than the predicted value (NO at step S45), then the control sequence returns to step S42. But if the remaining cash amount is lower than the predicted value (YES at step S45), then a point at which ›predicted value!=›remaining money amount at present! stands (time interval T in FIG. 26) is calculated (step S46). Consequently, as seen in FIG. 26, the operable time Ta is calculated as (n-1).times.Tu+T, and the corresponding operable time to is calculated as ›present point of time!+(n-1).times.Tu+T (step S47).

Such processing as described above will be described more simply with reference to FIG. 26. First, in a section 1 shown in FIG. 26 (at steps S41 to S45 of FIG. 27), a predicted value for each prediction factor is successively added from the current point of time and a prediction factor (value of n) with which the sum value first exceeds the remaining cash amount of the ATM 11 at present is detected. Then, in another section 2 shown in FIG. 26 (at step S46 of FIG. 27), it is detected by calculation at which point of time of the nth prediction factor the predicted value and the remaining money amount at present coincide with each other. Consequently, the operable time point t.sub..0. and the operable time period Ta are calculated.

The operable time point t.sub..0. calculated in such a manner as described above is displayed together with other prediction results on the display unit 18L of the remaining money management terminal 18 as shown in FIG. 20 by the function of the remaining money management section 18d.

By the way, in a usual case, when stoppage of operation of any of the ATMs 11 to 13 because of consumption of cash occurs, an alarm is developed, and every time an alarm is developed, a staff member supplements cash into the stopping ATM. Accordingly, in order to supplement cash into each of the ATMs 11 to 13 before stoppage of it because of consumption of cash occurs, the remaining money amounts in the ATMs 11 to 13 must be supervised periodically, which makes a heavy burden to a staff member. Further, since the operation schedule is different among different days of the week or the like, it is difficult to grasp till which hour of which day each of the ATMs 11 to 13 can operate with the remaining money amount at present in it.

However, where the operable time point prediction function in the present embodiment described above is used, a staff member can readily grasp, only by referring to the display unit 18L of the remaining money management terminal 18, until when each ATM can operate with the remaining money amount at present, and can grasp an ATM which may possibly suffer from stoppage of operation because of consumption of cash before next periodical supplement of cash. Consequently, the staff member can perform supplement of money with certainty without the necessity for periodical supervision and need not take a countermeasure every time stoppage of operation because of consumption of cash occurs, resulting in significant reduction in burden to the staff member. Further, since an operable time point according to an operation schedule can be predicted and calculated, the staff member can know the operable time point immediately even if it does not know the operation schedule.

B-6. Initial Transaction Data Setting Technique

In the remaining money management system described above, in order to accurately predict the demanded cash numbers for the ATMs 11 to 13, transaction data of the ATMs 11 to 13 must be collected and stored by some amounts. However, when it is tried to introduce a remaining money management system into a business office to be opened newly, since no transaction data in the past are accumulated at all as yet, a data collection apparatus is connected to ATMs of the business office and data are collected actually for a certain period of time (for example, for three to six months). Unless the data collected in an actual operating condition are used, accurate prediction, and consequently, accurate operation of the remaining money management system, cannot be anticipated. Since the transaction of a bank has different characteristics among different business offices, particular data cannot be applied as data for an arbitrary business office.

Therefore, in the present embodiment, transaction data of a plurality of business offices in the past are analyzed to set a plurality of different transaction patterns to which the individual business offices may belong (such as, for example, of a bedroom town type, a terminal type, an office type, a rural district type, a general type and so forth) in advance.

Then, for a business office to be established newly, that is, for a business office with which transaction data in the past to be used for prediction of the demanded money amount by the remaining money management library 18E are not stored on the hard disk 18G, transaction data in the past of another business office which belongs to the same transaction form pattern as that of the new business office (that is, existing data collected in advance) are installed, upon shipment of the system, as default data into the hard disk 18G. This allows operation of the remaining money management system immediately after establishment of the business office.

In the following, a detailed method of setting initial transaction data to a business office to be established newly will be described with reference to the flow chart (steps S51 to S59) shown in FIG. 28.

If the business office to be established newly is of the bedroom town type (YES at step S51), then upon shipment of the system, data of a business office A of the bedroom town type collected already are set and stored as initial transaction data onto the hard disk 18G (step S52). Or, if the business office to be established newly is of the terminal type (YES at step S53), then upon shipment of the system, data of another business office B of the terminal type collected already are set and stored as initial transaction data onto the hard disk 18G (step S54) in a similar manner as described above.

Or, if the business office to be established newly is of the office type (YES at step S55), then upon shipment of the system, collected data of a further business office C of the office type are set and stored as initial transaction data onto the hard disk 18G (step 356).

Further, if the business office to be established newly is of the rural district type (YES at step S55), then upon shipment of the system, collected data of a still further business office D are set and stored as initial transaction data onto the hard disk 18G (step S58).

If the judgments at steps S51, S53, S55 and S57 are all NO, it is determined that the business office to be established newly is of the general type, and upon shipment of the system, collected data of a yet further business office E of the general type are set and stored as initial transaction data onto the hard disk 18G (step S59).

It is to be noted that, when data of a transaction form pattern are to be installed at step S52, S54, S56, S58 or S59, if the installed data are corrected based on the number of ATMs to be installed in the new business office, then the initial transaction data can be set so as to conform with a form of the transaction of the new business office.

In this manner, even where transaction data in the past are not stored at all on the hard disk 18G because a business office is to be opened newly or by some other reason, since transaction data in the past of another business office having a transaction form pattern similar to that of the business office in question are used as initial transaction data, the demanded money number can be predicted with a certain degree of