VENDING MACHINE SERVICE SCHEDULING
Techniques are provided for calculating vending machines' service priorities and scheduling the vending machines for service taking into account a number of factors and thresholds (520). In some embodiments, the machines (110) are subdivided into subroutes (320). Each subroute has one or more machines, and at least one subroute has a plurality of machines. The vending machine service schedule is generated by selecting the highest priority subroute (410) and selecting the machines in that subroute (420). Other subroutes can be selected (430, 440) if there is time left in the Service Period.
The present application is a division of U.S. patent application Ser. No. 14/034,958, filed Sep. 24, 2013, incorporated herein by reference, which is a division of U.S. patent application Ser. No. 12/944,223, filed Nov. 11, 2010, incorporated herein by reference, now U.S. Pat. No. 8,571,705, which is a division of U.S. patent application Ser. No. 11/096,889, entitled “Vending Machine Service Scheduling,” filed on Mar. 31, 2005, incorporated herein by reference, now U.S. Pat. No. 7,894,938.
BACKGROUND OF THE INVENTIONThe present invention relates to vending machines, and more particularly to scheduling vending machine service to have the vending machines restocked.
Vending machines should be restocked promptly to ensure maximum sales and customer satisfaction. On the other hand, too frequent restocking involves wasteful service calls by service personnel.
To facilitate optimum restocking, vending machines 110 (
Improved vending machine service scheduling methods are desirable.
SUMMARYThis section summarizes some features of the invention. Other features are described in the subsequent sections. The invention is defined by the appended claims which are incorporated into this section by reference.
Due to service personnel limitations, it may be impossible or difficult to service all the machines at the same time or even on a single day. Some machines must be selected for service in preference to other machines. This selection can be difficult due to different factors that may have to be taken into account. For example, some machines may run out of a few products but be adequately stocked with other products for days to come. Other machines may be stocked with all of the products but if the product quantity is low then many products may run out quickly. Some embodiments of the present invention provide techniques for calculating the service priority for the vending machines taking a number of factors into account. Some embodiments provide fast and simple computation techniques easily implemented with a computer.
Further, the inventor has observed that if the machines are scattered over a large geographical area, it may be more efficient to service machines with a lower service priority before higher priority machines. For example, suppose that machines A and B have the highest service priority (say, they are almost empty) but they are far from each other. It may be more efficient to service the machine A and its adjacent machines on a given day, and service the machine B on a different day, even though the machines adjacent to A have a lower priority than B. Therefore, in some embodiments, the machines are subdivided into “subroutes”. Each subroute has one or more machines, and at least one subroute has a plurality of machines. In the example above, the machine A and its adjacent machines can be one subroute, and the machine B and its adjacent machines can be another subroute. The service schedule for a given day is generated by selecting the highest priority subroute and selecting the machines in that subroute. Other subroutes can be selected if there is time left in the day after servicing the first subroute. Each subroute can be a group of adjacent machines, or a group of machines positioned so as not to require an inordinate travel time between the machines. For example, a subroute may contain machines located far from each other but interconnected by a high speed road.
Also, if the service schedule generation takes into account the vending machines' locations and the driving times between the machines, the schedule generation may be too slow for a given computing power. Therefore, in some embodiments, simplified assumptions are made regarding the driving times. For example, in some embodiments, each machine is assumed to require the same fixed service time (including any driving time). The subroutes can sometimes be configured to generate good schedules with this assumption (if, for example, the driving time differences are negligible within each subroute.) The invention is not limited to any subroute configuration however.
The invention is not limited to the features and advantages described above. Other features are described below. The invention is defined by the appended claims.
The embodiments described in this section illustrate but do not limit the invention. The invention is defined by the appended claims.
To service the vending machines 110, a driver obtains supplies at a home office 310 (
The vending machines can be scattered over a large geographical area. It may be more efficient to service a group of adjacent machines together before servicing another group of machines. In
A more complex model can also be chosen. For example, in addition to the fixed time for servicing each vending machine, additional driving time can be allocated for driving to the first machine of each subroute. This additional time may be different for different subroutes. Also, as shown in
A subroute does not have to be in a contiguous area. Machines distant from each other but interconnected by a high speed highway may be placed in the same subroute, with another subroute between those machines. The subroutes can be generated manually by the user of system 140, or automatically by a system that examines the driving times between different locations 330. The invention is not limited to any subroute generation or configuration or to the advantages described herein.
In addition, each subroute 320 is associated with a “Second To” (or “SecondTo”) list 510. This is a list of other subroutes each of which can immediately precede the subroute 320 in the list of subroutes serviced on the same day. Exemplary “Second To” lists for
In Table 1, subroute 320A can be the next subroute serviced after subroute 320B or 320C on the same day. Subroute 320B can be the next after subroute 320C but not after subroute 320A. The reason for this may be that the subroute 320B is farther from office 310 than subroute 320A and it would be inefficient for the driver to go from subroute 320A to subroute 320B. For the same or some other reasons, subroute 320C cannot be serviced after any other subroute. The “Second To” data can be generated by the user manually or automatically using some criteria. The invention is not limited to any criteria or “Second To” lists.
In addition, each subroute is associated with a Class parameter 512. Exemplary class parameters are shown in Table 1. In one embodiment, the class value can be 1 (for a “primary” subroute), 2 (for a “secondary” subroute), or 3 (“tertiary” subroute). A primary subroute (class 1) is a subroute which is Second To None and thus must be the first to be serviced, if at all, on any given day. This may be because the subroute is a great distance from home 310, or is very large, but these factors are not limiting. A secondary subroute (such as subroute 320B) can be either the first serviced subroute or can be serviced after another subroute on any give day. The tertiary subroutes (class 3) are those that cannot be the first serviced subroutes but can only be serviced after another subroute on any given day. These could be smaller subroutes close to home 310 that the user only wants to be selected as extra work at the end of the day.
Some embodiments do not have the class indicators. In other embodiments, the class indicators are present but the SecondTo lists are omitted. In some such embodiments, a secondary subroute is assumed to be Second To any primary or secondary subroute. A tertiary subroute is assumed to be Second To any other subroute. The Second To lists are thus not needed. Other class indicators can also be used to indicate a subroute service priority.
The pictorial data representations such as that of
Each machine is associated with a Service Threshold parameter or parameters 520 indicating some threshold that the machine should reach before being served. The following parameters are illustrated in
Empty Coils Threshold 520.1 is a threshold for the number of empty coils 220 in the machine. The machine will receive a higher service priority if the number of empty coils reaches the Threshold 520.1, as described below.
Empty Products Threshold 520.2. The machine will receive a higher service priority if the number of empty products in the machine reaches the Threshold 520.2.
Required Service Interval 520.3 is the maximum time that the machine should be allowed to operate without service. The Required Service Interval can be flexible (giving a somewhat higher service priority to the machine) or strict (giving a highest service priority to the machine). A strict parameter may be desirable, for example, if the machine sells perishable food or some dangerous condition may develop in the machine. The invention is not limited to a particular parameter usage, and the threshold parameters can be set by the user as desired. Some embodiments use both a flexible Required Service Interval and a strict Required Service Interval.
Cash Threshold 520.4 is a threshold for the amount of cash in the machine. The amount of cash is calculated from data periodically sent by the machine to database 120. The machine gets a higher service priority if the amount of cash reaches the threshold 520.4.
Some embodiments do not use all of the threshold parameters shown in
1. Inventory (current number of items in the coil).
2. Stock Level (“StockLevel”), also called “Par Level” (the number of items in the coil when the coil has just been restocked).
3. Alert Level (“AlertLevel”). The coil is considered empty when the inventory drops to the alert level, which may be positive rather than zero. This is because some vending machines cannot empty a coil and they stop vending when there is still some number of items in the coil. That number of items is the “alert level”.
4. Daily Average (“DailyAverage”), computed as the average sales (i.e. number of items sold) per day from that coil. In some embodiments, the Daily Average is taken over the past 60 days. In some embodiments, a weighted average is used, with the recent sales weighted more heavily than the older sales. In some embodiments, the Daily Average is cleaned up to account for faulty data received from the vending machine as described below.
5. Product ID. In this example, a Universal Product Code (UPC) is used, but this is not limiting.
6. Days To Empty (“DaysToEmpty”) computed at step 410 as described below.
Step 410.1—Points Computation for a Given Vending Machine (See
Step 410.1—Sub-Step 810:
For each coil 220, calculate “DaysToEmpty”. This is a forecast estimate of the number of days until the coil reaches its Alert Level (“AlertLevel”). The calculation can be:
DaysToEmpty=(Inventory−AlertLevel)/DailyAverage. (1)
In some embodiments, DaysToEmpty is rounded to an integer. The rounding can be to the nearest integer, or it can be up or down.
Other computation methods can also be used. For example, if the vending machine is inaccessible on weekends, the number of weekend days during the “DaysToEmpty” period can be added to DaysToEmpty to get a new DaysToEmpty value. In another example, the machine is accessible during the weekends, but the weekend sales are different from the week day sales. Then two separate daily averages can be computed, one for the week days and one for the weekends. Each day (say, day d1) beginning with today or tomorrow is then examined, and a running total for the number of items to be sold starting today or tomorrow is incremented by the daily average value for day d1 (i.e., depending on whether the day d1 is a work day or a weekend day). When the running total first reaches or exceeds the value (Inventory-AlertLevel) for the coil, the corresponding day d1 is considered to be the day when the coil will become empty. DaysToEmpty are computed as the number of days until that day d1 (including d1 or not including d1). In other embodiments, separate daily averages are kept for other day categories for which substantially different sales averages are expected. For example, separate categories can be created for summer weekends, summer week days, non-summer weekends, and non-summer weekdays, to get a more accurate DaysToEmpty estimate.
Step 410.1—Sub-Step 820:
In a similar fashion, a forecast estimate “Days To Empty From Restock” (or “DaysToEmptyFromRestock”) is generated for the number of days until the coil becomes empty if it is restocked immediately. In one embodiment,
DaysToEmptyFromRestock=(StockLevel−AlertLevel)/DailyAverage (2)
Other computational methods described above for DaysToEmpty can also be used for DaysToEmptyFromRestock. The invention is not limited to any computational methods.
Step 410.1—Sub-Step 830:
The DaysToEmpty data are used to generate “Days To Service” (or “DaysToService”). This is an estimate for the number of days until the machine will reach its Empty Coils Threshold 520.1 (
Step 410.1—Sub-Step 830—Sub-Sub-Step 830.1:
The DaysToEmpty data are placed in an array. For the example data of
Array=[4,1,3,5,2,8,10,12,2] (3)
In this array, if the same product is on multiple coils (e.g. coils A1 and A2), only the maximum DaysToEmpty is taken. Thus, the coil A2 data are omitted.
Step 410.1—Sub-Step 830—Sub-Sub-Step 830.2:
The array is sorted. For the example of
Array=[1,2,2,3,4,5,8,10,12] (4)
Step 410.1—Sub-Step 830—Sub-Sub-Step 830.3:
DaysToService is calculated as Array (Empty Coils Threshold), where the array index varies from 1 to the number of coils. In the example above, Array(1)=1. If Empty Coils Threshold is 2 coils, then DaysToService=2.
Step 410.1—Sub-Step 840 (FIG.
8): As seen from formula (4), if EmptyCoilsThreshold=2, then three coils will become empty on the day corresponding to DaysToService because Array(3)=Array(2). At step 840, “Empty Coils” (or “EmptyCoils”) is calculated as the number of empty coils on the day corresponding to DaysToService. EmptyCoils is calculated as the index of the last array entry equal to Array(EmptyCoilsThreshold). For the example (4), EmptyCoils=3.
Step 410.1—Sub-Step 850:
The DaysToEmptyFromRestock data are used to generate “Days To Service From Restock” (or “DaysToServiceFromRestock”). This is an estimate for the number of days when the machine will reach its Empty Coils Threshold 520.1 if the machine is restocked immediately. In one embodiment, this computation is identical to step 830 except that DaysToEmptyFromRestock is used in place of DaysToEmpty.
The following table describes data structures available in memory 140M and/or 150M for each machine 110 at the conclusion of step 860:
Step 410.1—Sub-Step 870:
The service priority (“Points”) is calculated for each machine as an increasing function of EmptyCoils and a decreasing function of DaysToService. In one embodiment,
Points=EmptyCoils/(DaysToService+1) (5)
The inventor has found experimentally that the following formula gives superior results for differentiating the machines by their need of service:
Points=EmptyCoils0.5/(DaysToService1.75+1) (6)
In some embodiments, the Points are calculated as a function of DaysToService only, for example:
Points=1/(DaysToService+1) (7)
The calculation (7) gives a higher priority to machines 110 with smaller DaysToService. EmptyCoils is not used, and step 840 is omitted. In formula (6), EmptyCoils is used to differentiate between machines with the same or similar DaysToService parameter. The DaysToService power of 1.5 is larger than the EmptyCoils power of 0.5 to give DaysToService a larger weight. The DaysToService power can be 2, and other powers and weighing techniques can also be used.
End of Step 410.1
In some embodiments, the Points computation takes into account other thresholds 520.2, 520.3, 520.4. For example, DaysToService can be calculated using Empty Products (the number of empty products in the machine) instead of Empty Coils. This is done by calculating DailyAverage in formula (1) as the average number of items of a product sold by the machine, and calculating the Inventory as a combined inventory for the product. The AlerLevel can be zero for example. In other embodiments, Points can be defined as some other increasing function of the number of empty products in the machine, or Points can be increased by some amount if the empty products number has reached the threshold 520.2. Likewise, Points can be increased if the time since the last service has reached the Required Service Interval 520.3. If this is a strict Required Service Interval, the Points can be set to some predetermined high value to increase the chances that the machine's subroute will be selected at step 410.3 (described below) and the machine itself will be selected at step 420 (
In some embodiments, the following method is used. Suppose at first that DaysToService is calculated as described above for step 830.3, based on the Empty Coils Threshold 520.1. Denote this DaysToService value as DaysToService_1. Another DaysToService value DaysToService_2 is calculated in a similar way based on the number of empty products in the machine rather than empty coils, based on the Empty Products Threshold 520.2. Then DaysToService can be calculated as:
DaysToService=min(DaysToService_1,DaysToService_2) (8)
Other computations are also possible, giving different weights to DaysToService_1 and DaysToService_2. Also, DaysToService can be calculated as:
DaysToService=min(DaysToService_1,Days remaining till the machine reaches a flexible or strict Required Service Interval 520.3) (9)
In another embodiment, DaysToService is calculated as the minimum of DaysToService_2 and the days remaining till the machine reaches a flexible or strict Required Service Interval, or as the minimum of DaysToService_1, DaysToService_2 and the days remaining till the machine reaches a flexible or strict Required Service Interval. Other computations of DaysToService can also be used, giving different weights to DaysToService_1, DaysToService_2, the number of days till the machine reaches a flexible Required Service Interval, and the number of days till the machine reaches a strict Required Service Interval. The DaysToService can also be a decreasing function of the cash amount in the machine, and/or DaysToService can be reduced if the machine has reached its Cash Threshold 520.4.
Step 410.2 (
The Points are totaled for the vending machines in each subroute 320, to obtain SubRoutePoints for each subroute.
Step 410.3:
Subroutes 320 are sorted in the decreasing order of SubRoutePoints to generate a subroute list 1010 (
Step 420 (
Let “Service Day” be the day for which the service schedule is generated. This may be today, tomorrow, the next business day, or some other day. For example, the schedule can be generated in the morning of the Service Day, or in the evening of the day immediately before the Service Day. In some embodiments, step 420 is performed as follows (
Step 420.1:
Pick the machine 110 having the maximum Points among the machines which have not been selected and which have their DaysToService≦MaxDays. Add the machine to the Service List (the list of machines scheduled for service on the Service Day; the list is initialized to empty before the first iteration of step 420 in the loop of
For each machine 110 added to the Service List, for the machine's location 330, calculate Min(DaysToServiceFromRestock) amongst all the machines at the location. This minimum indicates the earliest day when any machine at that location is likely to be scheduled for service if all the machines at that location are restocked immediately.
Repeat Step 420.1 until either one of the following conditions is true:
Condition 420.1—C1: The schedule is full. For example, if each machine is allocated a fixed amount of service time, and no other driving or servicing time is taken into account for the schedule generation, then the condition 420.1—C1 means that the Service List reaches the maximum number of machines to be serviced in one day. If this condition is true, terminate the step 420 and go to step 450 (
Condition 420.1—C2: All the machines in the subroute are in the Service List. In this case, go to step 430 (
Condition 420.1—C3: The subroute has unselected machines, but all of these machines have their DaysToService>MaxDays. In this case, proceed to Step 420.2.
Step 420.2:
For each location 330 for which a machine was selected in the subroute at step 420.1, for each non-selected machine at that location, check for the following condition:
Condition 420.2—C1: The machine's DaysToService is less than Min(DaysToServiceFromRestock) for that location (see step 420.1). If this condition is true, add the machine to the Service List. (This condition indicates that the machine will reach its Empty Coils Threshold 520.1 before the next service would be scheduled for this location if all the machines at the location were restocked immediately.)
Note.
In some embodiments, the machines satisfying the condition 420.2—C1 are added to the Service List in the order of the decreasing Points values. In other embodiments, the machines are added as they are examined, or in some other order. In other embodiments, before adding any machine to the Service List, the non-selected machines satisfying the condition 420.2—C1 from all the selected locations 330 are placed in a separate list. Then the machines on the separate list are added to the Service List in the order of the decreasing Points values. Other embodiments are also possible for machine selection.
When adding any machine to the Service List, check for the following conditions:
Condition 420.2—C1: The schedule is full (same as 420.1—C1 above). In this case, go to step 450 (
Condition 420.2—C2: All the machines have been selected at all of the locations selected at step 420.1. In this case, go to step 430 (
Step 420.3:
This step is reached if there are still unselected machines at the locations 330 selected at step 420.1 but the schedule is not yet full. In this case, for the locations 330 selected at step 420.1, traverse the unselected machines, and for each unselected machine, add it to the Service List if (a) the machine will reach its Required Service Interval Threshold 520.3 on the Service Day (the day for which the schedule is being generated), and/or (ii) the machine has reached its Cash Threshold 520.4 (
The Note to Step 420.2 applies also to step 420.3.
When adding any machine to the Service List, check for the following conditions:
Condition 420.3—C1: The schedule is full (same as 420.1—C1 above). In this case, go to step 450 (
Condition 420.3—C2: There are no more machines to select at step 420.3. In this case, go to step 430 (
End of Step 420
Step 440 (
In some embodiments, this step is performed as follows. Let us denote the current subroute (the subroute for which the step 420 has just been executed) as S1. Then the subroutes that have not yet been selected are scanned to see if S1 is on their “Second To” list. Among all such subroutes, a subroute is selected with the maximum SubRoutePoints. If there are no subroutes that have not already been selected which have S1 on their Second To list, go to step 450.
End of Step 440
Then the machines for each subroute are listed. Only an incomplete listing for only one subroute (“Orange County”) is shown in
Additional features of some embodiments will now be described.
Sales Data Clean Up for Calculating the Inventory and Daily Average
In some embodiments, the sales data can be “cleaned up” by server 150 and/or client 140 when the inventory and daily average data are generated for
In
This technique can also be used for
In
In some embodiments, the same techniques are used to calculate DailyAverage for a given product rather than a coil.
The invention is not limited to the embodiments described above. For example, in some embodiments, the subroute selection operations at steps 410.3 (
The Days to Service parameter is an estimate for a time left until the time when the machine will need service (until the machine's “need-service” time). Days to Service can be defined as starting from any point, for example, from today until the need-service time, from tomorrow until the need-service time, or from any other time point. Also, the Days to Service estimate can be viewed as an estimate defining the need-service time itself. Alternatively, the estimated need-service time defines Days to Service. Other estimates can also be used that define the Days to Service time or the need-service time. The invention is not limited to any mathematical computations or expressions for estimating Days to Service, Days to Empty, or other parameters.
The relationships “less than” and “less than or equal to” are interchangeable. For example, at step 420.1 in
DaysToService≦MaxDays
is equivalent to
DaysToService<MaxDays+1,
so the “less than” relationship can be used to compute the “less than or equal to” relationship. The invention is not limited to any particular way to compute the “less than” or “less than or equal to” relationships, or “greater than” or “greater than or equal to” relationships, or other logical or arithmetic values. Also, MaxDays can be defined as incremented by 1. The invention can be practiced with the configuration of
Claims
1-30. (canceled)
31. A method for servicing a plurality of vending machines, the method comprising:
- obtaining, by a computer system, state data for the vending machines, the state data indicating changes, or absence of changes, of the vending machines' states over time;
- determining by the computer system, for one or more of the vending machines, the vending machines' first timing estimates, the vending machines' first timing estimates being determined by using the vending machines' state data and one or more predefined service threshold parameters for the vending machines, wherein each first timing estimate is indicative of an estimated time until a corresponding vending machine will need service after a predefined service period assuming that one or more of the plurality of the vending machines are serviced in the predefined service period; and
- selecting, by the computer system, one or more of the vending machines for service, the one or more of the vending machines being selected by using at least one first timing estimate, and identifying each selected vending machine by the computer system to initiate service for each selected vending machine.
32. The method of claim 31 further comprising determining by the computer system, for at least two of the vending machines, the vending machines' second timing estimates, the vending machines' second timing estimates being determined by using the state data and the one or more predefined service threshold parameters for the vending machines, wherein each second timing estimate is indicative of an estimated time until a corresponding vending machine will need service without assuming that any vending machine is serviced in the predefined service period;
- wherein the computer system uses at least one second timing estimate to perform said selecting.
33. The method of claim 31 wherein said selecting comprises selecting which of the vending machines is to be serviced in said predefined service period.
34. The method of claim 33 wherein the predefined service period is insufficient to service all the vending machines.
35. The method of claim 32 wherein said selecting comprises:
- selecting one or more of the vending machines for service by using at least one second timing estimate but not using any first timing estimate; and then
- selecting one or more additional vending machines for service by using at least one second timing estimate.
36. The method of claim 35 wherein the one or more additional vending machines are selected by using information on the one or more vending machines selected without using any first timing estimate.
37. The method of claim 31 wherein said obtaining the state data comprises receiving inventory data from each vending machine over a network at different times, the inventory data for each vending machine being indicative of a number of products in the vending machines' product containers.
38. The method of claim 31 wherein at least one vending machine comprises a plurality of product containers, and determining at least one first timing estimate comprises:
- determining, for one or more of the product containers of one or more of the vending machines, the product containers' first container-based estimates, the product containers' first container-based estimates being determined by using the vending machines' state data, wherein each first container-based estimate is indicative of an estimated time until a number of products in a corresponding product container will reach at least a predefined alert level for the product container;
- wherein at least one first timing estimate is determined using at least one first container-based estimate.
39. The method of claim 31 wherein the predefined service threshold parameters define, for each vending machine, one or more of thresholds Th1 and Th2, wherein:
- (a) the vending machine's threshold Th1 is a threshold for a number of the vending machine's product containers satisfying a condition C1 which is a condition that an amount of a product in the container is at most a predefined alert level for the container;
- (b) the vending machine's threshold Th2 is a threshold for a number of the vending machine's products vended by the vending machine that satisfy a condition C2 which is a condition that an amount of the product in the vending machine is at most a predefined alert level for the product.
40. The method of claim 31 further comprising repeatedly performing, by each vending machine, operations of:
- obtaining, by a controller in the vending machine, the state data for the vending machine, the state data being indicative of the number of products in each product container of the vending machine;
- transmitting, by a transmitter in the vending machine, the state data to the computer system over a network.
41. A method for restocking a plurality of vending machines each of which comprises one or more product containers, the method comprising:
- obtaining, by a computer system, state data for the vending machines, the state data indicating the state data indicating changes, or absence of changes, of the vending machines' states over time;
- determining by the computer system, for one or more of the product containers, the product containers' first container-based estimates, the product containers' first container-based estimates being determined by using the vending machines' state data, wherein each first container-based estimate is indicative of an estimated time until a number of products in a corresponding product container will reach at least a predefined alert level for the product container after a predefined service period assuming that one or more of the plurality of the vending machines are restocked in the predefined service period; and
- selecting, by the computer system, one or more of the vending machines for restocking, the one or more of the vending machines being selected by using at least one first container-based estimate, and identifying each selected vending machine by the computer system to initiate restocking for each selected vending machine.
42. The method of claim 41 further comprising determining by the computer system, for at least two of the product containers, the product containers' second container-based estimates, the product containers' second container-based timing estimates being determined by using the state data, wherein each second container-based estimate is indicative of an estimated time until a number of products in a corresponding product container will reach at least the predefined alert level for the product container without assuming that any vending machine is restocked in the predefined service period;
- wherein the computer system uses at least one second container-based estimate to perform said selecting.
43. The method of claim 41 further comprising repeatedly performing, by each vending machine, operations of:
- obtaining, by a controller in the vending machine, the state data for the vending machine, the state data being indicative of the number of products in each product container of the vending machine;
- transmitting, by a transmitter in the vending machine, the state data to the computer system over a network.
44. One or more computer readable media comprising computer instructions operable to cause a computer system to perform a method for servicing a plurality of vending machines, the method comprising:
- obtaining, by the computer system, state data for the vending machines, the state data indicating the state data indicating changes, or absence of changes, of the vending machines' states over time;
- determining by the computer system, for one or more of the vending machines, the vending machines' first timing estimates, the vending machines' first timing estimates being determined by using the vending machines' state data and one or more predefined service threshold parameters for the vending machines, wherein each first timing estimate is indicative of an estimated time until a corresponding vending machine will need service after a predefined service period assuming that one or more of the plurality of the vending machines are serviced in the predefined service period; and
- selecting, by the computer system, one or more of the vending machines for service, the one or more of the vending machines being selected by using at least one first timing estimate, and identifying each selected vending machine by the computer system to initiate service for each selected vending machine.
45. The one or more computer readable media of claim 44 wherein the method further comprises determining by the computer system, for at least two of the vending machines, the vending machines' second timing estimates, the vending machines' second timing estimates being determined by using the state data and the one or more predefined service threshold parameters for the vending machines, wherein, in the method, each second timing estimate is indicative of an estimated time until a corresponding vending machine will need service without assuming that any vending machine is serviced in the predefined service period;
- wherein, in the method, the computer system uses at least one second timing estimate to perform said selecting.
46. The one or more computer readable media of claim 44 wherein, in the method, said selecting comprises selecting which of the vending machines is to be serviced in said predefined service period.
47. The one or more computer readable media of claim 46 wherein the predefined service period is insufficient to service all the vending machines.
48. The one or more computer readable media of claim 45 wherein, in the method, said selecting comprises:
- selecting one or more of the vending machines for service by using at least one second timing estimate but not using any first timing estimate; and then
- selecting one or more additional vending machines for service by using at least one second timing estimate.
49. The one or more computer readable media of claim 48 wherein, in the method, the one or more additional vending machines are selected by using information on the one or more vending machines selected without using any first timing estimate.
50. The one or more computer readable media of claim 44 wherein, in the method, said obtaining the state data comprises receiving inventory data from each vending machine over a network at different times, the inventory data for each vending machine being indicative of a number of products in the vending machines' product containers.
51. The one or more computer readable media of claim 44 wherein, in the method, at least one vending machine comprises a plurality of product containers, and determining at least one first timing estimate comprises:
- determining, for one or more of the product containers of one or more of the vending machines, the product containers' first container-based estimates, the product containers' first container-based estimates being determined by using the vending machines' state data, wherein, in the method, each first container-based estimate is indicative of an estimated time until a number of products in a corresponding product container will reach at least a predefined alert level for the product container;
- wherein, in the method, at least one first timing estimate is determined using at least one first container-based estimate.
52. The one or more computer readable media of claim 44 wherein, in the method, the predefined service threshold parameters define, for each vending machine, one or more of thresholds Th1 and Th2, wherein:
- (a) the vending machine's threshold Th1 is a threshold for a number of the vending machine's product containers satisfying a condition C1 which is a condition that an amount of a product in the container is at most a predefined alert level for the container;
- (b) the vending machine's threshold Th2 is a threshold for a number of the vending machine's products vended by the vending machine that satisfy a condition C2 which is a condition that an amount of the product in the vending machine is at most a predefined alert level for the product.
53. The one or more computer readable media of claim 44 wherein the method further comprises repeatedly performing, by each vending machine, operations of:
- obtaining, by a controller in the vending machine, the state data for the vending machine, the state data being indicative of the number of products in each product container of the vending machine;
- transmitting, by a transmitter in the vending machine, the state data to the computer system over a network.
54. One or more computer readable media comprising computer instructions operable to cause a computer system to perform a method for restocking a plurality of vending machines each of which comprises one or more product containers, the method comprising:
- obtaining, by the computer system, state data for the vending machines, the state data indicating changes, or absence of changes, of the vending machines' states over time;
- determining by the computer system, for one or more of the product containers, the product containers' first container-based estimates, the product containers' first container-based estimates being determined by using the vending machines' state data, wherein each first container-based estimate is indicative of an estimated time until a number of products in a corresponding product container will reach at least a predefined alert level for the product container after a predefined service period assuming that one or more of the plurality of the vending machines are restocked in the predefined service period; and
- selecting, by the computer system, one or more of the vending machines for restocking, the one or more of the vending machines being selected by using at least one first container-based estimate, and identifying each selected vending machine by the computer system to initiate restocking for each selected vending machine.
55. The one or more computer readable media of claim 54 wherein, in the method, the method further comprises determining by the computer system, for at least two of the product containers, the product containers' second container-based estimates, the product containers' second container-based timing estimates being determined by using the state data, wherein, in the method, each second container-based estimate is indicative of an estimated time until a number of products in a corresponding product container will reach at least the predefined alert level for the product container without assuming that any vending machine is restocked in the predefined service period;
- wherein, in the method, the computer system uses at least one second container-based estimate to perform said selecting.
56. The one or more computer readable media of claim 54 wherein the method further comprises repeatedly performing, by each vending machine, operations of:
- obtaining, by a controller in the vending machine, the state data for the vending machine, the state data being indicative of the number of products in each product container of the vending machine;
- transmitting, by a transmitter in the vending machine, the state data to the computer system over a network.
57. The one or more computer readable media of claim 44 in combination with one or more computer processors coupled to the one or more computer readable media to be operable to execute the computer instructions to perform said method.
58. The one or more computer readable media of claim 45 in combination with one or more computer processors coupled to the one or more computer readable media to be operable to execute the computer instructions to perform said method.
59. The one or more computer readable media of claim 46 in combination with one or more computer processors coupled to the one or more computer readable media to be operable to execute the computer instructions to perform said method.
60. The one or more computer readable media of claim 47 in combination with one or more computer processors coupled to the one or more computer readable media to be operable to execute the computer instructions to perform said method.
61. The one or more computer readable media of claim 48 in combination with one or more computer processors coupled to the one or more computer readable media to be operable to execute the computer instructions to perform said method.
62. The one or more computer readable media of claim 49 in combination with one or more computer processors coupled to the one or more computer readable media to be operable to execute the computer instructions to perform said method.
63. The one or more computer readable media of claim 50 in combination with one or more computer processors coupled to the one or more computer readable media to be operable to execute the computer instructions to perform said method.
64. The one or more computer readable media of claim 51 in combination with one or more computer processors coupled to the one or more computer readable media to be operable to execute the computer instructions to perform said method.
65. The one or more computer readable media of claim 52 in combination with one or more computer processors coupled to the one or more computer readable media to be operable to execute the computer instructions to perform said method.
66. The one or more computer readable media of claim 53 in combination with: (a) one or more computer processors coupled to the one or more computer readable media, and (b) the vending machines, to be operable to execute the computer instructions to perform said method.
67. The one or more computer readable media of claim 54 in combination with one or more computer processors coupled to the one or more computer readable media to be operable to execute the computer instructions to perform said method.
68. The one or more computer readable media of claim 55 in combination with one or more computer processors coupled to the one or more computer readable media to be operable to execute the computer instructions to perform said method.
69. The one or more computer readable media of claim 56 in combination with: (a) one or more computer processors coupled to the one or more computer readable media, and (b) the vending machines, to be operable to execute the computer instructions to perform said method.
Type: Application
Filed: Jan 28, 2016
Publication Date: Aug 25, 2016
Inventors: Mandeep Singh Arora (Danville, CA), Anant Agrawal (Diamond Bar, CA), Fred Cheng (Aliso Viejo, CA), Eric Matthew Chu (Laguna Niguel, CA), Amedee Louis Beaudoin (Oakland, CA)
Application Number: 15/009,482