Delivery plan creation method and delivery plan creation device

Abstract

To create a delivery plan for delivery people to easily understand.SOLUTION: A delivery plan creation method is executed by an information processor to create delivery plans for a plurality of packages. The delivery plan creation method includes a first determination step for determining a delivery order in units of blocks, which are areas including a plurality of delivery destinations; a second determination step for determining a delivery plan within each of the blocks; and a third determination step for determining an overall delivery plan by connecting delivery plans within each of the blocks.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The present invention relates to a delivery plan creation method and a delivery plan creation device. [Background technology]

[0002] Due to the increase in demand for delivery (home delivery) and labor shortages, there is a need to improve the efficiency of delivery operations, especially last-mile delivery. One measure to improve the efficiency of last-mile delivery is to improve delivery route planning.

[0003] Conventional delivery route planning methods emphasize optimizing (minimizing) the total delivery time, taking into account time-slot-specified delivery and the time constraints for completing all deliveries. There is also a trend toward emphasizing optimizing (minimizing) the total driving distance from the perspective of fuel costs and carbon dioxide emissions. However, delivery plans created using conventional technology often result in inefficient routes that cross major roads, railways (railroad crossings), and rivers (bridges), increasing the risk of getting caught in traffic jams. While there is also research being done to utilize road network information to avoid crossing major roads, railways, rivers, etc., optimization is performed for all destinations, resulting in complex route plans that are difficult for humans to understand.

[0004] In addition, deliveries to multiple destinations may be performed in a single parking action. In order to perform efficient deliveries, it is necessary to determine the appropriate parking location according to the destination.

[0005] Furthermore, in order to create an appropriate delivery plan, it is necessary to accurately estimate the time required for each action. While much research has been done on estimating the time required for vehicle movement, there has not been much research on accurately estimating the time required for delivery after parking. While there have been estimates of the time required for walking to the destination site, there are few precedents for estimating the time required for movement within the building site or for receiving and handing over the package. On the other hand, in delivery from the final base to the destination (last-mile delivery), the time required for actions after parking accounts for a larger proportion than the time required for vehicle movement. Therefore, in order to create an efficient delivery plan that is feasible for humans, it is necessary to appropriately estimate the time required after parking. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] Japanese Patent Application Laid-Open No. 2012-083882 [Patent Document 2] Japanese Patent Application Laid-Open No. 2010-78571 [Patent Document 3] Japanese Patent Publication No. 2020-144655 Summary of the Invention [Problem to be solved by the invention]

[0007] In view of such circumstances, an object of the present disclosure is to provide a technology that can create a delivery plan that is efficient and suitable for human execution. [Means for solving the problem]

[0008] One aspect of the present invention is a method for producing a medicament for the treatment of a pulmonary arthritis. A delivery plan creation method executed by an information processing device to create a delivery plan for a plurality of packages, A first determination step of determining a delivery order for each block, which is an area including a plurality of delivery destinations; a second determination step of determining a delivery plan within each of the blocks; a third decision step of connecting the delivery plans within each block to determine an overall delivery plan; A delivery plan creation method including:

[0009] According to this method, the route of the final delivery plan is roughly a traffic line moving through blocks, and a delivery plan that is easy for delivery personnel to understand is obtained. Furthermore, by defining the blocks so that they do not include areas where crossings are predicted to be difficult, such as major roads, railways, or rivers, the risk of getting caught in traffic jams can be reduced. In this way, according to this aspect, it is possible to create a more appropriate delivery plan that is easy for delivery personnel to understand and has a low risk of getting caught in traffic jams.

[0010] Another aspect of the present invention is A delivery plan creation method executed by an information processing device to determine a delivery plan for a plurality of packages, determining a parking location according to a delivery destination based on a parking location database previously acquired from delivery logs and positioning logs relating to past deliveries; In the parking space database, a priority is set for each parking space according to the number of times the vehicle has been parked, In the step of determining a parking location, a parking location is determined for each delivery destination according to the priority. The delivery plan creation method is characterized by the following.

[0011] According to this method, parking locations that have actually been used in past deliveries can be used in the delivery plan. When a parking location is determined using a predetermined algorithm based on the location of the delivery destination and road information, it may be possible to obtain a location that is actually unavailable for parking. However, because the parking location is extracted from past performance, it is possible to obtain a location that is more reliably available for parking. From this perspective, this aspect makes it possible to create an appropriate delivery plan.

[0012] Yet another aspect of the present invention is a method for producing a pharmaceutical composition comprising: A delivery plan creation method executed by an information processing device to determine a delivery plan for a plurality of packages, a step of estimating a required delivery time after parking based on delivery information of one or more packages and parking locations when delivering the one or more packages, using an estimation model determined based on the required delivery times in past deliveries obtained from a delivery log; The delivery plan creation method is characterized by the following.

[0013] This method makes it possible to calculate the time required for delivery after parking the vehicle in the delivery plan, and since it is calculated based on past performance, it is highly accurate. Therefore, the time accuracy of the resulting delivery plan is high, and deliveries can be carried out as planned.

[0014] The present invention can be understood as a delivery plan creation method including at least some of the above-described functions and processes. Also, the present invention can be understood as a delivery plan creation device that executes the above-described delivery plan. [Effects of the Invention]

[0015] An object of the present invention is to provide a technology that can create an appropriate delivery plan. [Brief explanation of the drawings]

[0016] [Figure 1] FIG. 1 is a schematic diagram illustrating the configuration of a delivery plan creation device according to an embodiment. [Figure 2] 10 is a flowchart of a delivery plan creation process according to the embodiment. [Figure 3] FIG. 4 is a diagram illustrating a delivery plan creation process according to the embodiment. [Figure 4] FIG. 4 is a diagram illustrating a delivery plan creation process according to the embodiment. [Figure 5] FIG. 4 is a diagram illustrating a delivery plan creation process according to the embodiment. [Figure 6] FIG. 4 is a diagram illustrating a delivery plan creation process according to the embodiment. [Figure 7] FIG. 4 is a diagram illustrating a delivery plan creation process according to the embodiment. [Figure 8] FIG. 4 is a diagram illustrating a delivery plan creation process according to the embodiment. [Figure 9] FIG. 4 is a diagram illustrating a delivery plan creation process according to the embodiment. [Figure 10] FIG. 4 is a diagram illustrating a delivery plan creation process according to the embodiment. [Figure 11] FIG. 4 is a diagram illustrating a delivery plan creation process according to the embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0017] Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the drawings. Note that the present invention is not limited to the described embodiments. Furthermore, not all of the components described in the embodiments are necessarily essential to the present invention.

[0018] (composition) This embodiment relates to a delivery plan creation device that creates a delivery plan for multiple packages. A delivery plan is a plan that indicates the order and route in which packages to be delivered will be delivered. A delivery plan typically includes the route and parking locations of a delivery vehicle, as well as the packages to be delivered at each parking location and their delivery destinations. Note that some packages have a specified delivery time. In this disclosure, such packages are referred to as packages with a specified delivery time or time-specified packages. The delivery plan needs to be created to meet time-specified requests and other service requirements.

[0019] FIG. 1(A) is a hardware configuration diagram of the delivery plan creation device 1, and FIG. 1(B) is a functional block diagram of the delivery plan creation device 1. As shown in FIG.

[0020] As shown in FIG. 1(A), the delivery plan creation device 1 is realized by a computer (information processing device). The hardware configuration of the delivery plan creation device 1 includes an arithmetic processor 11 such as a CPU, a communication device 12 capable of wireless or wired communication, an input device 13 such as a keyboard or mouse, an output device such as a display, and a storage device 15 including a main memory and an auxiliary storage device. The processor 11 executes a computer program stored in the storage device 15, causing the delivery plan creation device 1 to realize each of the functions shown in FIG. 1(B). Any or all of the functions of the delivery plan creation device 1 may be realized by a dedicated hardware device. Alternatively, the delivery plan creation device 1 may be realized by multiple computers that can communicate with each other.

[0021] As shown in FIG. 1(B), the delivery plan creation device 1 includes, as its functional units, a block visiting order determination unit 20, an intra-block delivery plan determination unit 30, and an overall delivery plan determination unit 40. The intra-block delivery plan determination unit 30 includes, as sub-functional units, a parking location determination unit 31 and a delivery time estimation unit 32. Details of each functional unit will be explained with reference to the flowcharts, so each functional unit will be briefly explained here.

[0022] The block visiting order determination unit 20 determines the delivery order for each block. A block is a section that includes multiple potential delivery destinations, and is typically divided in advance taking into account city blocks, road networks, etc. The blocks are configured so as not to include routes that require time to pass through, such as main roads, railways, or rivers, and are divided so that delivery workers can easily recognize them. The block visiting order determination unit 20 determines the delivery order for each block based on the destination and specified delivery time for the current delivery. The block visiting order determination unit 20 determines the order of visiting blocks taking these factors into consideration. The block visiting order determination unit 20 roughly determines whether or not service conditions such as time zone designation can be met at this point, and if the conditions cannot be met, excludes some packages from this delivery. The block visiting order determination unit 20 may also divide a pre-set block as necessary and determine the visiting order for the divided blocks.

[0023] The intra-block delivery plan determination unit 30 determines a delivery plan within each block. To determine a delivery plan within a block, the parking location determination unit 31 determines a parking location for each delivery destination. The parking location determination unit 31 determines the current parking location taking into account parking locations used in past deliveries. In addition, the delivery required time estimation unit 32 estimates the required delivery time within the block. The required delivery time within the block includes the time required for vehicle movement (including parking) and the time it takes for the delivery person to make the delivery on foot after parking (also referred to as the doorstep time). For the doorstep time, the delivery required time estimation unit 32 estimates the current required time taking into account the actual required time for past deliveries. In this way, by determining the required delivery time in more detail, a delivery plan can be created that satisfies delivery conditions.

[0024] The overall delivery plan determination unit 40 determines an overall delivery plan by connecting the delivery plans within each block determined by the intra-block delivery plan determination unit 30 in the block tour order determined by the block tour order determination unit 20.

[0025] Fig. 2 is a flowchart of the delivery plan creation method executed by the delivery plan creation device 1. Figs. 3 to 12 are diagrams for explaining the details of the processing of the delivery plan creation method. The delivery plan creation method according to this embodiment will be described below with reference to Figs. 2 to 12.

[0026] The delivery plan creation process shown in Figure 2 includes a first determination step (STP01 to STP08) that determines the delivery order for each block, a second determination step (loop process L1; STP09 to STP11) that determines a delivery plan within each block, and a third determination step (STP12) that connects the delivery plans within each block to determine an overall delivery plan. The first to third determination steps are executed by the block visiting order determination unit 20, the intra-block delivery plan determination unit 30, and the overall delivery plan determination unit 40, respectively.

[0027] In SPT01, the block circulating order determination unit 20 extracts parcels with delivery time specifications from the parcel delivery information, and determines the delivery order of the parcels with delivery time specifications.

[0028] Figure 3(A) is a diagram that schematically illustrates the destination of a parcel included in parcel delivery information. The parcel delivery information includes parcels for which no delivery time is specified, as indicated by white circles 301, and parcels for which a delivery time is specified, as indicated by black circles 302 containing white numbers. Here, the delivery time is specified as a time period, with "14" indicating a time period from 14:00 to 16:00, "16" indicating a time period from 16:00 to 18:00, and "18" indicating a time period from 18:00 to 20:00. The method for setting the delivery time is not particularly limited and can be arbitrary.

[0029] FIG. 3(B) shows the result of determining the delivery order for packages with specified delivery times, as determined in STP01. There are no particular limitations on how the delivery order is determined; it may be determined so as to satisfy the time specification and minimize travel costs (values ​​based on one or more of travel time, travel distance, fuel consumption, and carbon dioxide emissions). In this example, a route is determined from delivery start location 310 to deliveries to destinations 311-316 in this order, and then to delivery end location 317. Note that delivery start location 310 and delivery end location 317 may be determined as appropriate, and these two locations may be the same. Note that travel to delivery end location 317 does not need to be taken into consideration.

[0030] In STP02, the block patrol order determination unit 20 determines whether a delivery order can be determined that satisfies delivery conditions such as the time slot designation, or in other words, whether the delivery plan will fail. FIG. 4(A) is a diagram showing an example of a plan that will fail. As shown in FIG. 4(A), if there are packages 401-403 with a time slot designation between 14:00 and 16:00 and the destinations are far apart, it can be determined that delivery within the designated delivery time slot is impossible. Note that this determination can be made in any way. For example, the delivery order determined in STP01 may be determined to satisfy the time constraints of the time slot designation by estimating the time required for vehicle travel and the time required for delivery on foot, or a simple determination may be made based on the number of packages and travel distance. Alternatively, the plan may be determined to have failed if a delivery order that satisfies the constraints cannot be determined in STP01. If the plan has failed, processing proceeds to STP03. If the constraints are satisfied, processing proceeds to STP04.

[0031] In STP03, the block patrol order determination unit 20 excludes some of the packages with designated time periods from the delivery plan. Typically, it is sufficient to exclude some of the packages with designated time periods where deliveries are concentrated. Then, the process returns to STP01 to again determine the delivery order for packages with designated delivery times. In the example of Figure 4(B), package 403 is excluded, and as a result of the replanning, the delivery order is determined so that package 404 is delivered after package 402.

[0032] The processes STP01 to STP03 are repeated until the constraints are met. Excluded packages are scheduled for delivery by another delivery person, but this is outside the scope of this disclosure. Furthermore, the excluded packages do not have to be limited to time-specified packages; packages with no time specification and destinations close to the time-specified packages may also be excluded.

[0033] In STP04, the block visiting order determination unit 20 projects the destination of a package with a specified delivery time onto a block, and associates the package with a specified delivery time with the block. More specifically, it determines that the block that includes the delivery destination of the package is the block that corresponds to the package.

[0034] A block is an area that includes multiple delivery destination candidates, and the entire delivery area is divided into several blocks. The block division is performed manually or by computer processing in advance and stored in the block information. It is preferable that a block is easily recognized by a delivery person as a single unit and does not include areas that require time to pass through, such as major roads, railway tracks, or rivers. Note that instead of defining blocks in advance, blocks may be dynamically determined depending on the distribution of the delivery destinations for the current delivery.

[0035] The block patrol order determination unit 20 determines the block to which each package destination belongs by referring to the delivery destination and block information. Figure 5(A) is a diagram showing the correspondence between destinations and blocks. Destinations 501 and 502 are associated with block 511, destination 503 is associated with block 512, destinations 504 and 505 are associated with block 513, and destination 506 is associated with block 514.

[0036] In STP04, when projecting the destinations of parcels with specified delivery times onto blocks, if the destinations (parcels) projected onto one block have different specified delivery times (if there are multiple parcels with delivery time differences of a predetermined value or more), the block patrol order determination unit 20 divides the block. For example, as shown in FIG. 5(B), assume that the specified time for destination 501 is 14:00-16:00 and the specified time for destination 502 is 18:00-20:00. Here, destinations 501 and 502 are both included in block 511. In this way, if there are multiple parcels (destinations) in one block with specified delivery times that differ by a predetermined value or more, the block patrol order determination unit 20 divides the block so that these multiple parcels (destinations) correspond to different blocks. In the example of Fig. 5(B), block 511 to which destinations 501 and 502 belong is divided into two blocks: block 511a to which destination 501 belongs, and block 511b to which destination 502 belongs. There are various possible specific division methods, but any division may be used here as long as it satisfies the above-mentioned conditions.

[0037] In STP05, the block patrol order determination unit 20 estimates the delivery time required within each block. If the block has been divided in STP04, the required time is estimated for each of the divided blocks. There are no particular limitations on the method for estimating the required time within a block; for example, it may be estimated based on the number of destinations, the number of packages, the range of destinations, and the travel distance between destinations. Note that there may be some blocks in which there are no packages with specified delivery times, and in STP05 the required time within such blocks is also estimated.

[0038] 6A shows an example of calculation of the required time for each block. Here, it is assumed that the required time for each of blocks 611 to 614 is estimated to be 110 minutes, 70 minutes, 95 minutes, and 90 minutes, respectively.

[0039] In STP06, the block patrol order determination unit 20 determines the block patrol order (delivery order for each block) according to the delivery order of packages with specified delivery times. The block patrol order can basically be determined in the same way as the delivery order of packages with specified delivery times determined in STP01. Note that if there is a block without packages with specified delivery times, the patrol order for all blocks is determined by inserting it in an appropriate position.

[0040] Fig. 6(B) shows an example of the block cyclic order determined in STP06. This example is a block cyclic order determined in accordance with the cyclic order of Fig. 3(B) or Fig. 5(A).

[0041] In STP07, the block visiting order determination unit 20 determines whether the visiting order determined up to this point is flawed. Because the block visiting order and the required time within each block have been obtained, it is possible to determine based on this information whether the time constraints for parcels with specified delivery times can be met. If the time constraints are not met (the plan fails), the process proceeds to STP08, where the block is divided and the delivery order is re-planned on a block-by-block basis; if the constraints are met, the process proceeds to STP09, where a delivery plan is made within the block.

[0042] FIG. 7(A) is an example of a plan that fails. In this example, the blocks 711, 712, 713, and 714 are visited in this order, and the estimated time required within each block is 110 minutes, 120 minutes, 95 minutes, and 90 minutes. Based on this estimate, it is assumed that package 701, which has a specified delivery time between 16:00 and 18:00, cannot be delivered within the specified time. In other words, the plan in FIG. 7(A) is a failed plan that cannot satisfy the constraints on the time period specified for package 701.

[0043] In STP08, the block visit order determination unit 20 divides the blocks. More specifically, it divides one or more blocks that are located earlier in the visit order than the package that cannot meet the time slot specification, and then rearranges the delivery order of some of the blocks after the division. Specifically, in the example of FIG. 7(A), block 712, which is located earlier in the visit order than block 713 containing package 701 that cannot be delivered within the specified time, is divided into two blocks, block 712a and block 712b, as shown in FIG. 7(B). Blocks to be divided can be selected, for example, from blocks where the required time within the block exceeds a threshold, or blocks where the destinations of packages with time-specified delivery times are concentrated in a certain area, etc. Furthermore, it is preferable to divide the blocks so that there are blocks (712a) containing packages with time slots specified and blocks (712b) containing no packages with time slots specified.

[0044] After the block division in STP08, the process returns to STP05 to estimate the required time within the block and determine the block patrol order in STP06, and then in STP07 it is determined again whether the plan satisfies the constraints. In the example of Figure 7(B), block 712 is divided into two, block 712a and block 712b, and the patrol order of block 712b is changed to be last, so that the time zone specification of package 701 (and other packages) can be respected. Therefore, the blocks and block patrol order shown in Figure 7(B) are the results obtained by the block patrol order determination unit 20.

[0045] The processing from STP09 to STP11 is performed for each block. Here, the processing of each block is described as being performed sequentially as loop processing L1, but the processing for multiple blocks may be executed in parallel.

[0046] In STP09, the intra-block delivery plan determination unit 30 determines a parking location for each delivery destination based on the delivery log and positioning log related to past deliveries, using the parking location determination unit 31. Figure 8(A) is a flowchart showing the details of the parking location determination process in STP09.

[0047] In STP09-1, the parking location determination unit 31 extracts parking locations for each delivery destination related to past deliveries based on the delivery log and positioning log related to past deliveries. The delivery log records the time of delivery start, delivery completion, and return for each package. The positioning log records the location of the delivery vehicle by time. By matching the delivery log with the positioning log, the location of the delivery vehicle at the time of delivery completion can be determined, and the parking location actually used can be determined for each delivery destination. Note that while it is preferable for the positioning log to store the location of the delivery vehicle, parking locations can also be determined from a log storing the location of the delivery person, so this can be used instead.

[0048] In the processing of STP09-1, if multiple deliveries are made to one delivery destination, multiple parking locations are determined. Therefore, the parking location determination unit 31 determines the priority of each parking location for one delivery destination based on the number of times each parking location has been used (how many times that parking location has been used). The priority may be determined according to any standard as long as the priority is determined so that the more times a parking location has been used, the higher the priority. One example of a method for calculating the priority is to use the percentage of times that the parking location has been used.

[0049] It is desirable to set multiple priorities according to the attributes of the delivery date. Examples of attributes of the delivery date include weekday, holiday, and weather (heavy rain). The reason for using different priorities for each attribute of the delivery date in this way is because it is expected that available parking locations will change depending on the attributes of the delivery date. For example, on a certain road, there is a high probability that there will be no available parking spaces on weekdays due to heavy traffic, but there is a high probability that there will be available parking spaces on holidays, or vice versa. Also, during heavy rain, it is conceivable that the delivery person will park in a location close to the delivery destination to minimize the walking distance. In this way, it is expected that the appropriate parking location will change depending on the attributes of the delivery date, so a priority is determined for each attribute of the delivery date.

[0050] As a result of STP09-2, the parking space determination unit 31 stores the parking space and its priority for each delivery destination in the parking space DB. Figure 9 is a diagram explaining the parking space DB. Table 901 stores the address, building name, floor / room number, name, parking space ID, and priority of each parking space, with the delivery destination ID as the primary key. Table 902 stores the latitude, longitude, and required time calculation formula, with the parking space ID as the primary key. The required time calculation formula is determined in STP10.

[0051] In table 901, priority 1 is, for example, the priority for weekdays, priority 2 is the priority for holidays, and priority 3 is the priority for heavy rain. For delivery destination D001, parking location P001 has a higher priority on weekdays, parking location P002 has a higher priority on holidays, and parking location P0003 has a higher priority for heavy rain. However, it is possible that a specific parking location may be calculated to have a higher priority regardless of the attributes of the delivery date.

[0052] In STP09-3, the parking location determination unit 31 determines a parking location for each delivery destination of the current delivery by referring to the parking location DB. More specifically, for each delivery destination, the priority of the delivery date attribute of the current delivery is referred to, and the parking location with the highest priority is determined as the parking location to be used for the current delivery.

[0053] It is also possible that the delivery destination is a new delivery destination and has no record in the parking space DB. In such a case, the location closest to the delivery destination can be determined as the parking space from among the parking spaces that have been used in the past and are stored in table 902. It is also possible to set a delivery range for each parking space in table 902, and not use the parking space for delivery destinations that are outside this range. The delivery range may be set manually by an administrator, or may be automatically determined by a computer taking into account distance and map information. It is also possible to determine a parking space for a new delivery destination from map information without referring to the parking space DB.

[0054] In this way, a parking location can be determined for each delivery destination based on past performance. Because it is based on past performance, it is possible to prevent a location that is not actually available from being determined as a parking location. This method can also be used for bulk deliveries, where deliveries are made to multiple delivery destinations in a single parking session. In other words, if the same parking location is selected for multiple delivery destinations in this method, bulk deliveries can be made to these multiple delivery destinations.

[0055] For the sake of simplicity, the parking space DB is created in STP09, but the parking space DB may be created in advance.

[0056] FIG. 8(B) is a diagram showing an example of parking locations for each delivery destination determined by the processing in STP09. In the diagram, white circles represent delivery destinations, and "P" represents parking locations. In this way, one parking location is linked to one delivery destination, and one parking location is linked to one or more delivery destinations. In other words, there is a many-to-one correspondence between delivery destinations and parking locations.

[0057] In STP10, the intra-block delivery plan determination unit 30 estimates the required delivery time after parking for each parking location using the required delivery time estimation unit 32. This estimation uses an estimation model created based on delivery logs related to past deliveries. Figure 10(A) is a flowchart showing the details of the estimation model construction and delivery time estimation steps of STP10.

[0058] In STP10-1, the delivery time estimation unit 32 references the delivery log to acquire information about past deliveries. Since the purpose here is to build an estimation model that estimates the delivery time after parking, the information about past deliveries requires the delivery time (objective variable) and various information that serves as its explanatory variables. Examples of explanatory variables include the number of delivery destination buildings to which deliveries are made after one parking session, the number of delivery destinations, and the number of packages. The number of delivery destination buildings may be aggregated by delivery destination building type. Examples of building types include detached houses, small apartment buildings, large apartment buildings, small office buildings, medium-sized office buildings, large office buildings, and extra-large commercial facilities. Other building types may also include the presence or absence of elevators, the presence or absence of reception desks, and the business-to-customer ratio (B / C ratio, the ratio of businesses / stores to residential homes in the building). Information about buildings can be acquired by referencing a building information database.

[0059] In STP10-2, the delivery time estimation unit 32 calculates a required time estimation model for estimating the required delivery time from the explanatory variables based on the explanatory variables and the required delivery time (objective variable) for past deliveries. If the estimation model is a regression formula, it is obtained by regression analysis. Alternatively, if the estimation model is a machine learning model, it is obtained by an algorithm such as a neural network, SVM, or decision tree. The estimation model may be calculated individually for each parking space, or the same estimation model may be calculated for multiple parking spaces with similar attributes. Parking space attributes may include road type (size, traffic volume, etc.) and types of surrounding buildings. The estimation model calculated by STP10-2 is stored in the parking space database in association with each parking space (e.g., required delivery time calculation formula in table 902 in FIG. 9). In this embodiment, the objective variable of the estimation model is the required delivery time, but other variables may also be used as the objective variable, and there may be multiple objective variables.

[0060] For the sake of simplicity, the description is given here assuming that the estimation model is created within STP 10, but the estimation model may be created in advance.

[0061] In STP10-3, the delivery time estimation unit 32 references the estimation model to estimate the delivery time required after parking for each of the parking locations determined in STP09. Specifically, the delivery time estimation unit 32 acquires explanatory variables from the delivery information for each parking location and applies the estimation model to the explanatory variables, thereby estimating the time required for delivery after parking at each parking location.

[0062] FIG. 10(B) is a diagram showing an example of the estimated results of the required delivery time after parking at each parking position shown in FIG. 8(B).

[0063] In STP 11, intra-block delivery plan determination unit 30 creates a route plan connecting each parking location. Figure 11(A) is a diagram showing an example of an intra-block route plan, in which route plans 1111 to 1114 are determined for each of blocks 1101 to 1104.

[0064] In STP12, the overall delivery plan determination unit 40 connects the route plans within each block determined in STP11 to create an overall delivery plan. Figure 11(B) shows an example of the delivery plan that is finally obtained. Note that this delivery plan is checked to see if it meets conditions such as time specifications, and if it does not, it may be possible to re-execute processing such as dividing the block in STP08 or excluding some packages in STP03.

[0065] Advantages of the delivery planning method according to this embodiment will be described.

[0066] First, the order in which the blocks are visited is determined first, and then the delivery route within the block is determined, resulting in a final delivery plan that visits each block in sequence. Therefore, the final delivery plan is intuitively easy for delivery personnel to understand, as it visits each block in sequence. Even if blocks are divided, this intuitive understanding is not affected much. Furthermore, if a block is determined so as not to include areas where travel time may be significant, such as major roads, railways, or rivers, the final delivery plan can reduce the number of times the delivery personnel pass through such areas, thereby reducing the possibility of unexpected delays.

[0067] Secondly, since the parking location is determined based on past parking records, it is possible to select a location that is actually available for parking and suitable for delivery. If the parking location is determined by calculation based on factors such as the distance from the building entrance, it may be impossible to actually park or a location that is inappropriate for delivery may be selected as the parking location, but this can be avoided by determining the parking location based on past records. Furthermore, for each delivery destination, the system can select a location that is suitable for delivery according to the attributes of the delivery date. By setting multiple priorities depending on the delivery destination, it is possible to determine the appropriate parking location according to the attributes at the time of delivery. Furthermore, since a parking location is determined for each delivery destination, if the same parking location is obtained for different delivery destinations, combined deliveries can be made, so combined deliveries can be made without any special processing.

[0068] Third, the time required for delivery after parking is estimated based on past delivery performance. While it is possible to calculate the required time using a formula given in advance for the required delivery time, it is difficult to provide a formula that can always provide an appropriate estimate because the required time is determined by various factors. In this embodiment, an estimation model is constructed based on past performance, making it possible to accurately estimate the required time.

[0069] As described above, according to this embodiment, it is possible to create a delivery plan that is easier for delivery personnel to understand, has less deviation from the estimated required time, and is more efficient than conventional delivery plans. [Explanation of symbols]

[0070] 1: Delivery plan creation device 20: Block traversal order determination unit 30: Intra-block delivery plan determination unit 31: Parking location determination unit 32: Delivery required time estimation unit 40: Overall delivery plan determination unit

Claims

1. A delivery plan creation method executed by an information processing device to create a delivery plan for a plurality of packages, a first determination step of determining a delivery order for each block, which is an area including a plurality of delivery destinations; a second determination step of determining a delivery plan within each of the blocks; a third determination step of determining an overall delivery plan by connecting the delivery plans within each block; How to create a delivery plan, including:

2. The first determining step determining a delivery order for packages with specified delivery times; a step of associating the package having a specified delivery time with a block; determining a delivery order for the blocks according to a delivery order for the parcels with specified delivery times; The delivery plan creation method according to claim 1, further comprising:

3. The block is a predefined area, In the step of associating the package having a specified delivery time with a block, a block including a delivery destination of the package is determined to be the block corresponding to the package. The delivery plan creation method according to claim 2 .

4. In the step of associating the parcels with designated delivery times with the blocks, if there are a plurality of parcels whose designated delivery times differ by a predetermined value or more for one block, the block is divided so that the plurality of parcels correspond to different blocks.

4. The delivery plan creation method according to claim 3.

5. The first determining step For each of the blocks, estimating a delivery time within the block; determining whether a time constraint is satisfied based on a delivery time required within the block; re-planning the delivery sequence in units of blocks if the time constraint cannot be satisfied; The delivery plan creation method according to claim 2 , comprising:

6. In the step of re-planning the delivery sequence in units of blocks, a block in which the required delivery time exceeds a threshold is divided and re-planned.

6. The delivery plan creation method according to claim 5.

7. The second determination step includes a step of determining a parking location for each delivery destination based on a delivery log and a positioning log related to past deliveries. The delivery plan creation method according to claim 1 .

8. In the step of determining a parking location, a parking location for each delivery destination is determined based on a parking location database previously acquired from delivery logs and positioning logs related to past deliveries; In the parking space database, a priority is set for each parking space according to the number of times the vehicle has been parked, determining a parking location for each delivery destination according to the priority; The delivery plan creation method according to claim 7 .

9. a plurality of priorities are set according to attributes of the delivery date; In the step of determining the parking location, a parking location for each delivery destination is determined by referring to a priority according to an attribute of the delivery date of the current delivery.

9. The delivery plan creation method according to claim 8.

10. The second determination step includes a required time estimation step of estimating a required delivery time after parking for each parking location. The delivery plan creation method according to claim 7 .

11. the required time estimation step estimates the required delivery time using an estimation model determined based on required times for past deliveries obtained from the delivery log; The delivery plan creation method according to claim 10.

12. The estimation model is an estimation model in which the number of delivery destination buildings, the number of delivery destinations, and the number of packages to be delivered after one parking session are used as explanatory variables, and the required delivery time is used as a response variable. The delivery plan creation method according to claim 11 .

13. The second determination step includes creating a delivery plan that connects parking locations within each block; In the third determination step, the delivery plans within the block are connected to determine an overall delivery plan. The delivery plan creation method according to claim 7 .

14. A delivery plan creation method executed by an information processing device to determine a delivery plan for a plurality of packages, determining a parking location according to a delivery destination based on a parking location database previously acquired from delivery logs and positioning logs relating to past deliveries; In the parking space database, a priority is set for each parking space according to the number of times the vehicle has been parked, In the step of determining a parking location, a parking location is determined for each delivery destination according to the priority. A delivery plan creation method characterized by:

15. a plurality of priorities are set according to attributes of the delivery date; In the step of determining the parking location, a parking location for each delivery destination is determined by referring to a priority according to an attribute of the delivery date of the current delivery. The delivery plan creation method according to claim 14.

16. A delivery plan creation method executed by an information processing device to determine a delivery plan for a plurality of packages, and estimating a required delivery time after parking based on delivery information of one or more packages and parking locations when delivering the one or more packages, using an estimation model determined based on required delivery times in past deliveries obtained from a delivery log. A delivery plan creation method characterized by:

17. The estimation model is an estimation model in which the number of delivery destination buildings, the number of delivery destinations, and the number of packages to be delivered after one parking session are used as explanatory variables, and the required delivery time is used as a response variable.

17. The delivery plan creation method according to claim 16.

18. a memory that stores a program for executing each step of the delivery plan creation method according to any one of claims 1 to 17; an arithmetic processor; A delivery plan creation device comprising:

Images