Estimation apparatus, estimation method, and program

Abstract

To provide an estimation apparatus, an estimation method, and a program for calculating doorstep time necessary for formulating a more accurate delivery plan.SOLUTION: An estimation apparatus 100 includes: a position information acquisition unit 110 which acquires the time series of position of a first deliverer; a moving speed estimation unit 120 which estimates the moving speed of the first deliverer at each position; a partial time series extraction unit 130 which extracts partial time series from the acquired time series, based on the estimated moving speed; a partial time series selection unit which selects a partial time series to be associated with a first delivery destination; and a doorstep time estimation unit 150 which specifies an alighting position and a boarding position, based on a preset rule, from the selected partial time series, estimates a first stopping / parking position based on the specified alighting position and boarding position, and estimates a doorstep time related to a combination of the first delivery destination and the first stopping / parking position, from the time at which the boarding position is detected and the time at which the alighting position is detected.SELECTED DRAWING: Figure 3

Description

[Technical field]

[0001] The present invention relates to an estimation device, an estimation method, and a program. [Background technology]

[0002] There is known a technology for creating a delivery plan for multiple packages delivered by a delivery vehicle. For example, Patent Document 1 discloses a presentation device that calculates a user's interest level based on the difference between the time when the delivery notification information was sent and the time when the delivery notification information was viewed, and sets the delivery order of multiple packages based on the calculated interest level. In this presentation device, the delivery schedule time of each package is determined by searching map information for a delivery route that minimizes the delivery cost using the delivery destination address of each package. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] International Publication No. 2019 / 138776 Summary of the Invention [Problem to be solved by the invention]

[0004] However, in general, a delivery person temporarily parks and stops the vehicle, gets off the vehicle, and delivers the package to the delivery destination. For example, the time from parking to delivering the package and getting on the vehicle to a delivery destination such as a high-rise apartment building and a single-family home facing a public road is significantly different. Here, this time is called the door step time. Therefore, there is a possibility that the delivery cannot be made according to the delivery plan created in advance. Therefore, by considering the door step time for each delivery destination, it is possible to create a delivery plan with higher accuracy.

[0005] The present invention has been made in consideration of the above-mentioned circumstances, and provides an estimation device, an estimation method, and a program for calculating a door step time required to create a more accurate delivery plan. [Means for solving the problem]

[0006] In order to solve the above problems, an estimation device according to the present invention comprises: An estimation device that estimates a door step time, which is a time from when a deliverer gets off a vehicle, delivers a package to a delivery destination, and gets on the vehicle, comprising: a location information acquisition unit that acquires a time series of a location of a first deliverer detected in a process in which the first deliverer delivers a first package to a first delivery destination using a first vehicle; a moving speed estimation unit that estimates a moving speed of the first deliverer at each of the positions from a time at which each of the positions included in the time series acquired by the position information acquisition unit was detected; a partial time series extraction unit that extracts a partial time series in which it is estimated that the first delivery person is not riding in the first vehicle from the time series acquired by the position information acquisition unit based on the moving speed estimated by the moving speed estimation unit; a partial time series selection unit that selects a partial time series to be associated with the first delivery destination from the partial time series extracted by the partial time series extraction unit; a door step time estimation unit that identifies, based on a preset rule, an alighting location where the first delivery person alights from the first vehicle and an aboarding location where the first delivery person aboards the first vehicle from the partial time series selected by the partial time series selection unit, and estimates a first parking location where the first delivery person parks the first vehicle to deliver the first package to the first delivery destination based on the identified alighting location and the aboarding location, and estimates a door step time for the combination of the first delivery destination and the first parking location from the time when the identified aboarding location was detected and the time when the identified alighting location was detected; Equipped with. Effect of the Invention

[0007] According to the present invention, it is possible to provide an estimation device, an estimation method, and a program for calculating a door step time required to create a more accurate delivery plan. [Brief description of the drawings]

[0008] [Figure 1] FIG. 2 is an explanatory diagram showing cooperation between the estimation device and other devices. [Diagram 2] FIG. 13 is a diagram showing an example of a delivery instruction list received by a deliverer terminal. [Diagram 3] FIG. 2 is an explanatory diagram showing a functional configuration of the estimation device. [Figure 4] 11 is a diagram showing an example of time-series data acquired by a position information acquisition unit; FIG. [Diagram 5] FIG. 11 is a diagram illustrating an example of time series data extracted by a partial time series extraction unit. [Figure 6] 13 is a diagram illustrating an example of a process in which a partial time series extraction unit links a partial time series with a delivery destination. FIG. [Figure 7] FIG. 1 is a diagram showing an example of time series data plotted on a map. [Figure 8] 8 is a diagram showing the beginning and end parts of a partial time series of the time series data illustrated in FIG. 7. FIG. [Figure 9] FIG. 13 is a diagram showing an example of a doorstep time table generated by a doorstep time estimation unit. [Figure 10] FIG. 11 is a diagram showing an example of a delivery schedule list received by a delivery destination information acquisition unit. [Figure 11] FIG. 13 is a diagram illustrating an example of a prediction result table generated by a prediction unit. [Figure 12] FIG. 2 is an explanatory diagram showing a physical configuration of the estimation device. [Figure 13] 13 is a flowchart of a doorstep time estimation process. [Figure 14] 13 is a flowchart of a doorstep time prediction process. [Figure 15] FIG. 13 is a diagram illustrating an example of a process performed by a prediction unit. [Figure 16] 13 is a diagram illustrating another example of processing by the prediction unit. FIG. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0009] Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings. Note that the embodiment is for illustrative purposes only and does not limit the scope of the present invention. Therefore, a person skilled in the art can adopt an embodiment in which each or all of these elements are replaced with an equivalent, and such an embodiment is also included in the scope of the present invention.

[0010] (Relationship between the estimation device, the delivery terminal, and the program) The estimation device according to the present embodiment estimates a door step time, which is the time it takes for a deliverer who drives a vehicle to deliver a package to a delivery destination to park the vehicle at a parking location, get off the vehicle, deliver the package, return to the parking location, and get back into the vehicle. The estimation device is composed of one or more server devices. The estimation device collects a time series of the location of a deliverer terminal, which is a computer operated by the deliverer, and calculates the door step time for a combination of a delivery destination and a parking location based on the collected time series.

[0011] The estimation device of this embodiment is generally realized by making a computer execute a program, but it is also possible to execute the processing by a dedicated electronic circuit. In addition, as an intermediate form between a computer and a dedicated electronic circuit, it is also possible to configure the estimation device of this embodiment by applying a technology such as FPGA (Field Programmable Gate Array) that compiles a program into a design script for an electronic circuit and dynamically configures the electronic circuit based on the design script.

[0012] The estimation device according to this embodiment is realized by one or more server computers that communicate with the deliverer terminal executing each function realized by one or more server programs.

[0013] The deliverer terminal according to this embodiment is a terminal computer that realizes a smartphone, a tablet computer, a wearable terminal, etc., and can be realized by executing a terminal program provided in advance. As the terminal program, a so-called "app" can be adopted.

[0014] In addition, a general browser can be used as the terminal program, and a script program that runs on a browser can also be used as the terminal program.

[0015] Generally, programs executed by server computers and terminal computers can be recorded on computer-readable non-transitory information recording media such as compact disks, flexible disks, hard disks, magneto-optical disks, digital video disks, magnetic tapes, ROMs (Read Only Memory), EEPROMs (Electrically Erasable Programmable ROM), flash memories, semiconductor memories, etc. These information recording media can also be distributed and sold independently of the server computers.

[0016] In server computers and terminal computers, programs recorded on non-transitory information recording media such as flash memory and hard disks are read into a temporary storage device called a Random Access Memory (RAM), and then the instructions contained in the read program are executed by a Central Processing Unit (CPU). However, in architectures where ROM and RAM can be mapped and executed in a single memory space, the instructions contained in the program stored in ROM are directly read and executed by the CPU.

[0017] Furthermore, server programs and terminal programs can be distributed and sold to server computers and terminal computers, etc. from a distribution server, etc. managed by the operator, via a temporary transmission medium such as a computer communication network, independently of the computer on which the program is executed.

[0018] In addition, when the estimation device is composed of multiple computers, the programs running on the computers are multiple different server programs that have different functions and work together, and the combination of the multiple programs can be considered as a system program for realizing the prediction device.

[0019] (Overall composition) Fig. 1 is an explanatory diagram showing the cooperation between an estimation device and other devices. The following description will be made with reference to this diagram.

[0020] The estimation device 100 is connected to a deliverer terminal 200 via a computer communication network 300 such as the Internet.

[0021] The estimation device 100 is operated, for example, by a business providing a parcel delivery service, an online supermarket, etc., and estimates a door step time, which is the time from parking a vehicle at a parking position and getting off the vehicle, delivering the parcel, returning to the parking position, and getting on the vehicle, for each delivery destination, based on past delivery records. The estimation device 100 acquires time series data indicating the position of the deliverer terminal 200, and calculates the moving speed at each position. The estimation device 100 extracts a partial time series indicating a section where a deliverer is estimated to get off the vehicle based on the moving speed, and uses the extracted partial time series to identify an estimated parking position for each delivery destination and calculate the door step time. The estimation device 100 accumulates the calculation results in a database. When delivering a parcel to a new delivery destination not registered in the database, the estimation device 100 refers to the database and calculates a predicted value of the door step time for the new delivery destination based on information on a delivery destination with a matching or similar address. It should be noted that a delivery destination refers to a combination of an address and a name, and the estimation device 100 identifies delivery destinations with the same address but different names as different delivery destinations.

[0022] The deliverer terminal 200 is a smartphone, tablet, or other delivery-only terminal carried by a deliverer who delivers packages while working. The deliverer terminal 200 includes a position detection means for detecting the position of the deliverer terminal 200 and a display means for displaying information.

[0023] The position detection means includes a receiver that receives a signal from a satellite. The receiver is, for example, a Global Positioning System (GPS), which is a system for measuring a current position. The position detection means acquires the latitude and longitude, which are position information of the deliverer terminal 200, based on a signal received by the GPS at every preset time. The deliverer terminal 200 transmits the position information acquired by the position detection means to the estimation device 100 in association with the acquired time. The position detection means may acquire the position information of the deliverer terminal 200 based on a mobile phone base station communicating with the deliverer terminal 200. Alternatively, the position detection means may acquire the position information of the deliverer terminal 200 based on a wireless LAN base station communicating with the deliverer terminal 200.

[0024] The display means is, for example, a liquid crystal display or an organic electroluminescence display. For example, the deliverer displays a delivery instruction list showing a list of delivery destinations for each day, half day, etc. on the display means, checks the addresses of the delivery destinations, the delivery order, etc., and delivers the parcels. This delivery instruction list is created by a delivery plan planning system that plans a delivery plan, and is transmitted to the estimation device 100 and the deliverer terminal 200. As illustrated in FIG. 2, the delivery instruction list includes information such as a "delivery ID" for uniquely identifying delivery information, a "delivery destination ID" which is identification information for identifying the delivery destination, a "name" indicating the name of the delivery destination, an "address" indicating the address of the delivery destination, and a "target delivery time" indicating the target time of delivery for each delivery destination. The delivery instruction list may also include information such as a delivery time zone specified by the customer when the order is accepted, and a parcel ID for identifying the parcel to be delivered.

[0025] (Functional configuration of the estimation device) Next, the functional configuration of the estimation device 100 will be described with reference to Fig. 3. The estimation device 100 includes a position information acquisition unit 110, a travel speed estimation unit 120, a partial time series extraction unit 130, an alighting / boarding position specification unit 140, a doorstep time estimation unit 150, a doorstep time storage unit 160, and a doorstep time prediction unit 170.

[0026] The location information acquisition unit 110 acquires time series data including the location information of the deliverer terminal 200 and the acquisition time of the location information transmitted from the deliverer terminal 200. The time series data is stored in the memory of the estimation device 100. As illustrated in FIG. 4, the time series data includes information such as "date" indicating the date when the deliverer terminal 200 acquired the location information, "time t" indicating the acquisition time, and "location information" indicating the acquired location coordinates. The time series data may also include information such as a terminal ID that identifies the deliverer terminal 200 that collected the location information.

[0027] Returning to Fig. 3, the moving speed estimation unit 120 estimates the moving speed of the delivery person terminal 200. Specifically, the moving speed estimation unit 120 calculates the moving distance and moving time of the delivery person terminal 200 from the time information and position information included in the time series data illustrated in Fig. 4 stored in the memory of the estimation device 100, and calculates the moving speed of the delivery person terminal 200 from the calculated moving distance and moving time.

[0028] Returning to FIG. 3, the partial time series extraction unit 130 extracts a partial time series, which is time series data for a section where it is estimated that a delivery person is not in a vehicle, for each delivery destination from the time series data stored in the memory of the estimation device 100. Specifically, the partial time series extraction unit 130 extracts a partial time series for each delivery destination based on a preset rule. For example, as illustrated in FIG. 5, the partial time series extraction unit 130 extracts data from the time series data shown in FIG. 4 acquired by the location information acquisition unit 110, in which the moving speed calculated by the moving speed estimation unit 120 is equal to or less than a preset threshold, for example, 6 km / h or less. Note that the threshold for the moving speed is not limited to 6 km / h, and any speed at which it is estimated that a delivery person is not moving in a vehicle may be set.

[0029] Next, the partial time series extraction unit 130 divides the time series data in which the moving speed is equal to or less than a threshold into a plurality of partial time series that are continuous in time. For example, the time series data when packages are delivered to a plurality of delivery destinations includes at least a plurality of partial time series in which it is estimated that the delivery person is not in the vehicle for each delivery destination. The partial time series extraction unit 130 divides the extracted time series data into a plurality of partial time series according to an arbitrary rule, for example, classifying two time series data into the same partial time series if the time difference with the previous time series data is within one minute, and classifying two time series data into different partial time series if the time difference with the previous time series data is one minute or more.

[0030] Next, the partial time series extraction unit 130 links the partial time series with the delivery destination. Specifically, the partial time series extraction unit 130 calculates the distance between the position of each time series included in each partial time series and the position of each delivery destination included in the delivery instruction list shown in Fig. 2, and links the partial time series with the closest distance to the delivery destination, as shown in Fig. 6. The partial time series extraction unit 130 is an example of a partial time series selection unit.

[0031] Returning to FIG. 3, the disembarking / boarding position identifying unit 140 identifies the disembarking position and boarding position of the vehicle when the vehicle is delivered to the delivery destination based on the partial time series linked to the delivery destination by the partial time series extracting unit 130. Specifically, the disembarking / boarding position identifying unit 140 extracts, from each partial time series linked to the delivery destination, a leading portion that is position data acquired at an earlier time and a trailing portion that is position data acquired at a later time, the number of which is based on a preset rule. FIG. 8 is a diagram showing the leading portion and the trailing portion of the partial time series of the time series data exemplified in FIG. 7. As shown in the figure, for example, based on the rule that five leading portions and five trailing portions are extracted, the disembarking / boarding position identifying unit 140 extracts the first to fifth earliest acquired position data A1 to A5 as leading portions, and extracts the first to fifth latest acquired position data B1 to B5 as trailing portions. Next, the disembarking / boarding position specifying unit 140 calculates the distance between each of the position data A1 to A5 of the front portion and each of the position data B1 to B5 of the rear portion, specifies the position data of the front portion with the closest distance as the disembarking position, and specifies the position data of the rear portion with the closest distance as the boarding position. For example, if the distance between A5 and B2 is closer than the distance between other combinations, the disembarking / boarding position specifying unit 140 specifies A5 as the disembarking position and B2 as the boarding position.

[0032] Returning to FIG. 3, the door step time estimation unit 150 calculates the door step time associated with the combination of the vehicle parking position and the delivery destination. Specifically, the door step time estimation unit 150 identifies the center position of the coordinates of the disembarking position and the boarding position identified by the disembarking / boarding position identification unit 140 as the parking position. Next, the door step time estimation unit 150 calculates the door step time by subtracting the time when the identified disembarking position was detected from the time when the identified boarding position was detected. The door step time estimation unit 150 stores the calculated door step time in the door step time estimation unit 150 in association with the identified parking position and the delivery destination.

[0033] The doorstep time storage unit 160 stores a doorstep time table indicating the doorstep time for each delivery destination calculated by the doorstep time estimation unit 150. As illustrated in FIG. 9, the doorstep time table includes information such as a "delivery ID" for uniquely identifying delivery information, a "delivery destination ID" which is identification information for identifying the delivery destination, a "name" indicating the name of the delivery destination, an "address" indicating the address of the delivery destination, a "parking position" indicating the position information of the parking position for each delivery destination, and a "doorstep time" indicating the doorstep time for each delivery destination. As illustrated, the doorstep time estimation unit 150 may calculate the average of the time when the boarding position is detected and the time when the disembarking position is detected as an estimated time of the time when the package is delivered, and include the calculated "estimated delivery completion time" in the doorstep time table.

[0034] Returning to Figure 3, the doorstep time prediction unit 170 includes a delivery destination information acquisition unit 171, a doorstep time information extraction unit 172, a prediction unit 173, and an error estimation unit 174, and calculates a predicted value of the doorstep time of the scheduled delivery destination by referring to the doorstep time table stored in the doorstep time memory unit 160.

[0035] The delivery destination information acquisition unit 171 accepts the input of a delivery schedule list showing a list of delivery destinations scheduled for delivery by day, half day, etc., and acquires delivery information for each delivery destination. As illustrated in FIG. 10, the delivery schedule list includes information such as a "delivery ID" for uniquely identifying delivery information, a "delivery destination ID" which is identification information for identifying the delivery destination, a "name" indicating the name of the delivery destination, an "address" indicating the address of the delivery destination, and a "designated delivery time" indicating the designated delivery time for each delivery destination. If the user of the delivery destination does not specify a delivery time, the designated delivery time does not need to include time information, and if only either morning or afternoon is specified, information specifying whether it is morning or afternoon may be included. The delivery destination information acquisition unit 171 acquires information for each delivery destination from the input delivery schedule list, and judges whether each delivery destination is a new delivery destination for which the doorstep time has not been calculated by referring to the doorstep time table shown in FIG. 9. Specifically, whether or not a delivery destination is a new delivery destination is determined based on whether the delivery destination ID of each delivery destination included in the delivery schedule list is included in the doorstep timetable, or whether or not a combination of name and address of each delivery destination included in the delivery schedule list that matches the combination is included in the doorstep timetable.

[0036] Returning to Fig. 3, the door-step time information extraction unit 172 extracts door-step time information including the door-step time and the parking / stopping position, which is data used to predict the door-step time of each delivery destination, from the door-step time table shown in Fig. 9 stored in the door-step time storage unit 160. Specifically, when the delivery destination information acquisition unit 171 determines that the door-step time of the delivery destination to be calculated has already been calculated and is not a new delivery destination, the door-step time and the parking / stopping position of the same delivery destination as this delivery destination are extracted from the door-step time table.

[0037] Furthermore, when the delivery destination information acquisition unit 171 determines that the delivery destination to be calculated is a new delivery destination for which the door step time has not been calculated, the door step time information extraction unit 172 extracts the door step times and parking positions of multiple delivery destinations having the same or similar addresses as the delivery destination from the door step time table. Details of this process will be described later.

[0038] The prediction unit 173 calculates predictions of the door step times and parking positions of delivery destinations included in the delivery schedule list illustrated in Fig. 10 based on the door step times and parking positions extracted by the door step time information extraction unit 172. Specifically, in the case of a prediction of a delivery destination determined to be the same as a delivery destination included in the door step time table, the prediction unit 173 specifies the door step times and parking positions stored in the door step time table as the door step times and parking positions of the delivery to be calculated.

[0039] Furthermore, when predicting a delivery destination determined to be a new delivery destination, the prediction unit 173 calculates each prediction based on the door step times and parking positions of multiple delivery destinations having the same or similar address as the address of the delivery destination to be calculated, extracted by the door step time information extraction unit 172. Specifically, the prediction unit 173 obtains the average door step times of the multiple extracted delivery destinations and the central position of the parking positions, and calculates the obtained average door step time and the central position of the parking positions as the door step time and parking position of the delivery destination to be calculated.

[0040] The error estimation unit 174 calculates the error between the door step time and the parking stop position of each delivery destination calculated by the prediction unit 173, and creates a prediction result table showing the prediction of the door step time of each delivery destination based on the prediction result by the prediction unit 173 and the calculated error. As illustrated in FIG. 11, the prediction result table includes information such as a "delivery destination ID" which is identification information for identifying the delivery destination, a "name" which indicates the name of the delivery destination, an "address" which indicates the address of the delivery destination, a "door step time" which indicates the door step time for each delivery destination, a "door step time error" which indicates the error of the door step time, "parking position 1", "parking position 2", "parking position 3" which indicates the position information of the parking position for each delivery destination, and a "parking position error" which indicates the error of the parking position. Details of the processing by the error estimation unit 174 will be described later.

[0041] (Physical configuration of the estimation device) The estimation device 100 having the functional configuration described above physically comprises, as shown in FIG. 12, a CPU 11 that executes processing according to a program, a RAM 12 which is a volatile memory, a ROM (Read Only Memory) 13 which is a non-volatile memory, a storage unit 14 which stores data, an input unit 15 which accepts input of information, a display unit 16 which visualizes and displays information, and a communication unit 17 which transmits and receives information, all of which are connected via an internal bus 99.

[0042] The CPU 11 executes various processes by reading out the programs stored in the storage unit 14 into the RAM 12 and executing them. The CPU 11 executes each process by a travel speed estimation unit 120, a partial time series extraction unit 130, an alighting / boarding position identification unit 140, a doorstep time estimation unit 150, and a doorstep time prediction unit 170 as main functions provided by the programs.

[0043] The RAM 12 is used as a work area for the CPU 11. The ROM 13 stores a control program executed by the CPU 11 for the basic operation of the estimation device 100, a BIOS (Basic Input Output System), and the like.

[0044] The storage unit 14 includes a hard disk drive, stores the programs executed by the CPU 11, and stores various data used when the programs are executed. The storage unit 14 functions as a doorstep time storage unit 160.

[0045] The input unit 15 is a user interface including a keyboard, a mouse, a communication device, etc. The display unit 16 is a display device such as a liquid crystal display or an organic EL (Electro Luminescence) display that visualizes and displays information.

[0046] The communication unit 17 is a network termination device or a wireless communication device that connects to a network, and a serial interface or a LAN (Local Area Network) interface that connects to them.

[0047] (Doorstep time estimation process) Next, a description will be given of the operation of the estimation device 100. First, a door step time estimation process for estimating a door step time for each delivery destination based on past delivery records will be described with reference to FIG.

[0048] As a preparation, the storage unit 14 of the estimation device 100 stores in advance the time series data illustrated in Fig. 4. The location information acquisition unit 110 of the estimation device 100 receives the location information transmitted from the delivery person terminal 200 and the time when the location information was acquired, generates the time series data illustrated in the figure, and stores it in the storage unit 14.

[0049] When the user operates the input unit 15 of the estimation device 100 to set a period for which time-series data is desired to be referenced and instructs the estimation device 100 to start the doorstep time estimation process, the estimation device 100 starts the process.

[0050] The location information acquisition unit 110 of the estimation device 100 acquires time series data to be processed (step S101). Specifically, the location information acquisition unit 110 acquires time series data for a set period from the storage unit 14 of the estimation device 100.

[0051] Next, the moving speed estimation unit 120 estimates the moving speed of the delivery terminal 200 (step S102). Specifically, the moving speed estimation unit 120 calculates the moving distance and moving time of the delivery terminal 200 from the time information and position information included in the time series data illustrated in FIG. 4 acquired in step S101, and calculates the moving speed of the delivery terminal 200 from the calculated moving distance and moving time. For example, when calculating the moving speed at the illustrated position information (a2, b2), the moving speed estimation unit 120 calculates the moving distance, which is the distance between the position information (a1, b1) and (a2, b2) of the previous time series data, and the moving time, which is the difference between the acquisition time t2 = 10:13:22 of the position information (a2, b2) and the acquisition time t1 = 10:13:10 of the position information (a1, b1), and calculates the moving speed by dividing the moving distance by the moving time. Also, for example, the movement speed estimation unit 120 may use the previous and next time series data to determine a movement distance, which is the sum of the distance between the position information (a1, b1) and (a2, b2) of the previous time series data and the distance between the position information (a3, b3) and (a2, b2) of the next time series data, and a movement time, which is the difference between the acquisition time t3 = 10:13:30 of the position information (a3, b3) and the acquisition time t1 = 10:13:10 of the position information (a1, b1), and calculate the movement speed of the position information (a2, b2) by a method such as dividing the movement distance by the movement time.

[0052] 13, next, the partial time series extraction unit 130 extracts, for each delivery destination, a partial time series that is time series data for a section where it is estimated that a delivery person is not in the vehicle from the time series data based on the moving speed calculated in step S102 (step S103). Specifically, the partial time series extraction unit 130 extracts data where the moving speed calculated by the moving speed estimation unit 120 is equal to or less than a preset threshold, for example, 6 km / h or less, from the time series data.

[0053] Next, the partial time series extraction unit 130 divides the time series data whose moving speed is equal to or less than a threshold into a plurality of temporally continuous partial time series. The partial time series extraction unit 130 divides the time series data whose moving speed is equal to or less than a threshold into a plurality of partial time series according to an arbitrary rule, such as classifying two pieces of time series data into the same partial time series if the time difference with the previous time series data is within one minute, and classifying two pieces of time series data into different partial time series if the time difference with the previous time series data is more than one minute.

[0054] Next, the partial time series extraction unit 130 links the divided partial time series to the delivery destinations based on the delivery instruction list illustrated in FIG. 2. Specifically, the partial time series extraction unit 130 refers to the delivery instruction list for the same day as the acquisition date of the partial time series, identifies the delivery destination that is closest to the position of each time series included in the partial time series, and links it to the partial time series. The distance may be calculated by any method, such as calculating the average distance between each position included in the partial time series and the delivery destination, or calculating the distance between the first position or the last position of the partial time series and the delivery destination. As illustrated in FIG. 6, for example, when the partial time series extraction unit 130 determines that the distance between the partial time series 1 and the position of the delivery destination ID "A001" is the closest, the partial time series extraction unit 1 links the partial time series 1 to the delivery destination ID "A001".

[0055] Returning to FIG. 13, next, the alighting / boarding position identifying unit 140 selects a partial time series of a delivery destination for which the doorstep time is to be calculated (step S104). Specifically, the alighting / boarding position identifying unit 140 identifies an unprocessed delivery destination from the delivery instruction list, and selects a partial time series of the identified delivery destination. For example, if it is identified that the delivery destination ID "A001" is unprocessed, the alighting / boarding position identifying unit 140 selects a partial time series linked to the delivery destination ID "A001" from the partial time series generated in step S104.

[0056] Next, the disembarking / boarding position identifying unit 140 identifies the disembarking position and the boarding position of the vehicle when the vehicle is delivered to the delivery destination based on the partial time series selected in step S104 (step S105). Specifically, the disembarking / boarding position identifying unit 140 extracts, from the partial time series, a leading portion that is position data acquired at an earlier time and a trailing portion that is position data acquired at a later time, the number of which is based on a preset rule. A set number of position data may be extracted from the leading portion and the trailing portion, or a number of pieces of data based on a set ratio with respect to the total number of data included in the partial time series may be extracted from each of the leading portion and the trailing portion. As illustrated in FIG. 8, the disembarking / boarding position identifying unit 140 extracts, for example, the first to fifth earliest acquired position data A1 to A5 as the leading portion, and the first to fifth latest acquired position data B1 to B5 as the trailing portion, based on the rule of extracting five leading portions and five trailing portions. Next, the disembarking / boarding position specifying unit 140 calculates the distance between each of the position data A1-A5 of the front portion and each of the position data B1-B5 of the rear portion, specifies the position data of the front portion with the closest distance as the disembarking position, and specifies the position data of the rear portion with the closest distance as the boarding position. For example, if the distance between the position data A5 and the position data B2 is closer than the distance between other combinations, the disembarking / boarding position specifying unit 140 specifies the position data A5 as the disembarking position and the position data B2 as the boarding position.

[0057] Returning to Fig. 13, the door step time estimation unit 150 calculates the door step time related to the combination of the vehicle parking position and the delivery destination. Specifically, the door step time estimation unit 150 specifies the center position of the coordinates of the disembarking position and the boarding position specified by the disembarking / boarding position specification unit 140 as the parking position (step S106). In the example of Fig. 8, when the position data A5 is specified as the disembarking position and the position data B2 is specified as the boarding position, the center position of the position data A5 and the position data B2 is specified as the parking position.

[0058] Next, the door step time estimation unit 150 calculates the door step time by subtracting the time when the specified disembarking position is detected from the time when the specified boarding position is detected (step S107). For example, when the acquisition time t of the position data A5, which is the disembarking position, is 10:13:53, and the acquisition time t of the position data B2, which is the boarding position, is 10:14:55, the door step time estimation unit 150 calculates 62 seconds as the door step time. The door step time estimation unit 150 associates the calculated door step time with the specified parking position and delivery destination to create a door step time table shown in FIG. 9, stores it in the door step time storage unit 160 (step S108), and ends the process.

[0059] (Doorstep time prediction processing) Next, a door-step time prediction process for calculating a predicted door-step time for each scheduled delivery destination based on the door-step timetable created in the door-step time estimation process will be described with reference to Fig. 14. In the following description, a delivery schedule list showing a list of scheduled delivery destinations, as exemplified in Fig. 10, is input to the estimation device 100, and a process for predicting the door-step time for each delivery destination included in the delivery schedule list will be described as an example.

[0060] The delivery destination information acquisition unit 171 acquires delivery destination information, which is information on the delivery destination to be processed (step S201). Specifically, the delivery destination information acquisition unit 171 acquires information on each delivery destination from the delivery schedule list illustrated in FIG. 10, and judges whether each delivery destination is a new delivery destination for which the doorstep time has not been calculated, by referring to the doorstep time table illustrated in FIG. 9. In the delivery schedule list illustrated in FIG. 10, the delivery destination IDs "A001", "A002", and "A003" are included in the doorstep time in FIG. 9, so the delivery destination information acquisition unit 171 judges that the doorstep time has been calculated for these delivery destinations. In addition, the delivery destination ID "B001" is not included in the doorstep time, so the delivery destination information acquisition unit 171 judges that the delivery destination ID "B001" is a new delivery destination for which the doorstep time has not been calculated. Note that the delivery destination information acquisition unit 171 may judge whether a delivery destination is included in the doorstep time table based on whether the combination of the name and address is the same, regardless of the delivery destination ID.

[0061] Next, the doorstep time information extraction unit 172 extracts doorstep time information including the doorstep time and the parking position, which is data used for predicting the doorstep time of each delivery destination, from the doorstep time table shown in FIG. 9. Specifically, if it is determined in step S201 that the delivery destination to be processed is a delivery destination included in the doorstep time table and is not a new delivery destination (step S202; No), the doorstep time and the parking position of the same delivery destination as this delivery destination are extracted from the doorstep time table (step S203). For example, if the delivery destination ID to be processed is "A001", the doorstep time information extraction unit 172 extracts the doorstep time "300 seconds" and the parking position "(X1, Y1)" of the delivery destination ID "A001" from the doorstep time table shown in FIG. 9. Note that if there are multiple pieces of doorstep time information for the same delivery destination in the doorstep time table, the doorstep time information extraction unit 172 extracts multiple pieces of doorstep time information.

[0062] On the other hand, if it is determined in step S201 that the delivery destination to be processed is a new delivery destination not included in the doorstep timetable (step S202; Yes), the doorstep time information extraction unit 172 extracts the doorstep times and parking positions of delivery destinations having the same or similar address as the delivery destination (step S204). Specifically, the doorstep time information extraction unit 172 identifies delivery destinations having the same address but different names, or delivery destinations having highly similar addresses, from the doorstep timetable, and extracts the doorstep times and parking positions of the identified multiple delivery destinations. A method of identifying addresses with a high degree of similarity is, for example, to compare addresses 1 and 2, scan the character strings of each address from a wide area to a narrow area to find a common character string, and use the number of characters a of address 1, the number of characters b of address 2, and the number of characters c of the common character string to obtain 2×c / (a+b), c 2 The similarity is calculated using a formula such as / (a×b), and if the similarity is equal to or greater than a threshold, it is determined that the address has a high similarity. If there are multiple delivery destinations with the same or similar addresses, the door-step time information extraction unit 172 extracts door-step time information for multiple delivery destinations that satisfy the conditions from the door-step time table.

[0063] Next, the prediction unit 173 calculates predictions of the door step time and parking position of the delivery destination included in the delivery schedule list illustrated in Fig. 10 based on the door step time information extracted by the door step time information extraction unit 172 (step S205). Specifically, in the case of a prediction of a delivery destination determined in step S202 to be the same as a delivery destination included in the door step time table, the prediction unit 173 calculates the door step time and parking position extracted in step S203 as the door step time and parking position of the delivery destination to be processed.

[0064] In addition, when multiple door step time information is extracted in step S202, the average of multiple door step times and the center position of multiple parking and stopping positions may be calculated as the door step time and parking and stopping positions of the delivery destination to be calculated, respectively. In addition, multiple parking and stopping positions may be calculated by grouping parking and stopping positions that are close to each other as one parking and stopping position. Specifically, for example, an upper limit value of the distance between parking and stopping positions may be set, and parking and stopping positions whose mutual distance falls within the upper limit value may be grouped together. As illustrated in FIG. 15, when parking and stopping positions C1 to C6 are extracted, the respective center positions C7 and C8 of parking and stopping positions C1 to C3 and parking and stopping positions C4 to C6 whose mutual distance falls within the upper limit value may be calculated as the parking and stopping positions of the delivery destination to be processed, and the average of the door step times of parking and stopping positions C1 to C3 and the average of the door step times of parking and stopping positions C4 to C6 may be calculated as the door step times of the respective parking and stopping positions.

[0065] On the other hand, in the case of predicting a delivery destination determined to be a new delivery destination in step S202, the prediction unit 173 calculates the average of the door step times of the multiple delivery destinations extracted in step S204, and calculates the calculated average as the door step time of the delivery destination to be processed. In addition, the prediction unit 173 calculates the center position of the parking positions of the multiple delivery destinations extracted in step S204 as the parking position of the delivery destination to be processed. Specifically, as illustrated in FIG. 16, for example, when the new delivery destination is the 3rd floor of the building D3, and the 1st floors of the building D1, the building D2, and the building D3 are specified as delivery destinations with similar addresses, the prediction unit 173 calculates the average of the door step times of the 1st floors of the building D1, the building D2, and the building D3 as the door step time of the 3rd floor of the building D3, which is the new delivery destination to be calculated. Furthermore, the prediction unit 173 calculates the center position C12 of the parking position C9 of the building D1, the parking position C10 of the building D2, and the parking position C11 of the 1st floor of the building D3 as the parking position of the 3rd floor of the building D3, which is the new delivery destination to be calculated. The average is not limited to a simple average, and a weight may be set for each delivery destination according to the similarity of the addresses to obtain a weighted average, which is used as the door step time of the new delivery destination. For example, the weighted average may be obtained by setting a weight larger than that of the building D1 and the building D2 for the door step time of the 1st floor of the building D3, which is different from the 3rd floor of the building D3, which is the new delivery destination to be calculated, only in terms of the number of floors.

[0066] Next, the error estimation unit 174 calculates the error between the door step time of the delivery destination and the parking stop position calculated by the prediction unit 173 (step S206). Specifically, when a delivery destination determined to be the same as the delivery destination included in the door step time table in step S202 and a plurality of parking stop positions are specified, the error estimation unit 174 calculates the error between the door step time and each parking stop position based on the door step time and position information of each parking stop position. For example, when two parking stop positions are specified, the error of the door step time is calculated by adding the average or maximum value of the door step times of parking stop positions 1 and 2 to the standard deviation between the door step time of parking stop position 1 and the door step time of parking stop position 2. In addition, the distance between parking stop position 1 and parking stop position 2 is calculated, and the calculated distance is calculated as the error of the parking stop position. When there are three or more parking positions, the longest distance between the parking positions may be determined as the parking position error, or the standard deviation of the distances between the parking positions may be determined as the parking position error.

[0067] Furthermore, when calculating the error of a delivery destination determined to be a new delivery destination in step S202, the error estimation unit 174 calculates each error based on the door step times and parking / stopping positions of multiple delivery destinations having the same or similar address, using a method similar to that described above. Specifically, when calculating the error in the door step time, the error in the door step time is calculated by adding the average or maximum value of the door step times of multiple delivery destinations having the same or similar address as the new delivery destination to the standard difference in the door step times of the multiple delivery destinations. Furthermore, when calculating the error in the parking / stopping position, the longest distance between the parking / stopping positions among the parking / stopping positions of multiple delivery destinations having the same or similar address as the new delivery destination, or the standard deviation of the distance between the parking / stopping positions is calculated.

[0068] The error estimation unit 174 generates a prediction result table as shown in Fig. 11 based on the prediction result by the prediction unit 173 and the errors between the door step time and the parking position calculated in step S206 (step S207). The generated prediction result table is sent to a delivery plan creation system that creates a delivery plan including the delivery route, delivery order, number of deliveries, etc., and the delivery plan creation system creates a delivery plan taking into account the door step time and parking position of each delivery destination included in the prediction result table.

[0069] Specifically, for example, the delivery plan creation system searches for a route for a vehicle to travel from a departure position to a delivery destination using a route search means installed in a conventional car navigation system. The delivery plan creation system identifies a parking position in the obtained search result, and acquires a door step time and its error from the parking position, thereby calculating the time required to move from the departure position to the delivery destination, complete delivery to the delivery destination, and use the parking position of the delivery destination as a new departure position for the next delivery destination, and its error. When creating delivery plans for multiple delivery destinations, the delivery plan creation system not only shortens the required time, but also creates a delivery plan in which the sum of the errors in the required times for each delivery destination falls within the range of error allowed for a delivery slot, for example, from loading the package at the office and starting delivery to returning to the office.

[0070] As described above, the estimation device 100 identifies the parking location for each delivery destination from the time series data of the location of the delivery person terminal 200 detected during the delivery process, and calculates the door step time. The estimation device 100 associates the calculated door step time and the parking location with the delivery destination and stores them in the database. The estimation device 100 acquires the door step time and the parking location of the delivery destination of the delivery schedule from the accumulated database. When a new delivery destination not registered in the database is included in the delivery schedule list, the estimation device 100 calculates the prediction of the door step time and the parking location of the new delivery destination from the door step times and the parking locations of multiple delivery destinations with the same or similar addresses. Therefore, it is possible to calculate the door step time required to make a delivery plan with higher accuracy.

[0071] (Modification) In the above embodiment, the partial time series extraction unit 130 has been described as linking a partial time series to a delivery destination that is closest to each position included in the partial time series, but this is not limited to the above. For example, in an operation in which a delivery person registers information indicating completion of delivery every time he delivers a package to a delivery destination, it is sufficient to link a partial time series to a delivery destination whose delivery completion time is closest to the acquisition time of each position included in the partial time series.

[0072] In addition, the programs executed by the estimation device 100 and the like can be stored and distributed in computer-readable recording media such as CD-ROMs (Compact Disc Read Only Memory), DVDs (Digital Versatile Discs), MOs (Magneto-Optical Disks), USB memories, memory cards, etc. By installing such programs in a specific or general-purpose computer, the computer can be made to function as the estimation device 100 in the above embodiment.

[0073] The above-mentioned program may also be stored in a disk device of a server device on a communication network such as the Internet, and downloaded to a computer by, for example, superimposing the program on a carrier wave. The above-mentioned processing can also be achieved by starting and executing the program while transferring it via the communication network. Furthermore, the above-mentioned processing can also be achieved by executing all or a part of the program on a server device, and executing the program while the computer transmits and receives information related to the processing via the communication network.

[0074] In addition, when the above-mentioned functions are shared and realized by the OS (Operating System) or by the OS working together with an application, etc., only the parts other than the OS may be stored on the above-mentioned recording medium and distributed, or may be downloaded to a computer.

[0075] Various aspects of the present disclosure are summarized below as appendices.

[0076] (Appendix 1) An estimation device that estimates a door step time, which is a time from when a deliverer gets off a vehicle, delivers a package to a delivery destination, and gets on the vehicle, comprising: a location information acquisition unit that acquires a time series of a location of a first deliverer detected in a process in which the first deliverer delivers a first package to a first delivery destination using a first vehicle; a moving speed estimation unit that estimates a moving speed of the first deliverer at each of the positions from a time at which each of the positions included in the time series acquired by the position information acquisition unit was detected; a partial time series extraction unit that extracts a partial time series in which it is estimated that the first delivery person is not riding in the first vehicle from the time series acquired by the position information acquisition unit based on the moving speed estimated by the moving speed estimation unit; a partial time series selection unit that selects a partial time series to be associated with the first delivery destination from the partial time series extracted by the partial time series extraction unit; a door step time estimation unit that identifies, based on a preset rule, an alighting location where the first delivery person alights from the first vehicle and an aboarding location where the first delivery person aboards the first vehicle from the partial time series selected by the partial time series selection unit, and estimates a first parking location where the first delivery person parks the first vehicle to deliver the first package to the first delivery destination based on the identified alighting location and the aboarding location, and estimates a door step time for the combination of the first delivery destination and the first parking location from the time when the identified aboarding location was detected and the time when the identified alighting location was detected; An estimation device comprising:

[0077] (Appendix 2) the doorstep time estimation unit specifies the disembarking position based on a leading portion of the position data having an earlier detection time included in the partial time series selected by the partial time series selection unit, and specifies the boarding position based on a trailing portion of the position data having a later detection time. 2. An estimation apparatus as described in claim 1.

[0078] (Appendix 3) the doorstep time estimation unit extracts, from the selected partial time series, a set number of pieces of position data of the beginning portion and a set number of pieces of position data of the end portion, the number of pieces being based on a set ratio with respect to the total number of data included in the partial time series, and specifies one of the extracted pieces of position data of the beginning portion as the disembarking position, and specifies one of the extracted pieces of position data of the end portion as the boarding position. 3. An estimation apparatus as described in claim 2.

[0079] (Appendix 4) the door step time estimation unit identifies a combination of the position data of the leading portion and the position data of the trailing portion that are closest in distance from among the extracted position data of the leading portion and the position data of the trailing portion, and identifies the identified position data of the leading portion and the identified position data of the trailing portion as the disembarking position and the boarding position, respectively. 4. An estimation apparatus according to claim 3.

[0080] (Appendix 5) The door step time estimation unit specifies a center position between the specified disembarking position and the boarding position as the first parking position. 5. An estimation apparatus according to any one of claims 1 to 4.

[0081] (Appendix 6) the partial time series selection unit selects a partial time series that is closest to the address of the first delivery destination and associates it with the first delivery destination; 6. An estimation apparatus according to any one of claims 1 to 5.

[0082] (Appendix 7) A computer that estimates a door step time, which is a time from when a delivery person gets off a vehicle, delivers a package to a delivery destination, and gets on the vehicle, A step of acquiring a time series of a position of a first deliverer detected in a process in which the first deliverer delivers a first package to a first delivery destination using a first vehicle; estimating a moving speed of the first deliverer at each of the locations from a time at which each of the locations was detected, the time series being acquired; Extracting a partial time series in which it is estimated that the first delivery person is not riding in the first vehicle from the acquired time series based on the estimated moving speed; selecting a sub-time series to be associated with the first delivery destination from the extracted sub-time series; from the selected partial time series, based on a preset rule, identifying a drop-off location where the first delivery person dropped off the first vehicle and a boarding location where the first delivery person boarded the first vehicle, estimating a first parking location where the first delivery person parked the first vehicle to deliver the first package to the first delivery destination based on the identified drop-off location and the boarding location, and estimating a door step time for the combination of the first delivery destination and the first parking location from the time when the identified boarding location was detected and the time when the identified drop-off location was detected; An estimation method including:

[0083] (Appendix 8) A computer that estimates a door step time, which is the time from when a delivery person gets off a vehicle, when the delivery person delivers a package to a delivery destination, to when the delivery person gets on the vehicle, A process of acquiring a time series of a position of a first deliverer detected in a process in which the first deliverer delivers a first package to a first delivery destination using a first vehicle; A process of estimating a moving speed of the first deliverer at each of the locations from a time at which each of the locations was detected, the time series being acquired; extracting a partial time series in which it is estimated that the first delivery person is not riding in the first vehicle from the acquired time series based on the estimated moving speed; selecting a partial time series to be associated with the first delivery destination from the extracted partial time series; a process of identifying, from the selected partial time series based on a preset rule, a drop-off location where the first delivery person drops off from the first vehicle and a boarding location where the first delivery person boards the first vehicle, estimating a first parking location where the first delivery person parks the first vehicle to deliver the first package to the first delivery destination based on the identified drop-off location and the boarding location, and estimating a door step time for the combination of the first delivery destination and the first parking location based on the time when the identified boarding location was detected and the time when the identified drop-off location was detected; A program that executes the following.

[0084] Various embodiments and modifications of the present disclosure are possible without departing from the broad spirit and scope of the present disclosure. The above-described embodiments are for explaining the present disclosure and do not limit the scope of the present disclosure. That is, the scope of the present disclosure is indicated by the claims, not the embodiments. Various modifications made within the scope of the claims and the scope of the disclosure equivalent thereto are considered to be within the scope of the present disclosure. [Industrial Applicability]

[0085] The present invention can be suitably adopted in an estimation device, an estimation method, and a program that estimate the door step time of a delivery destination based on time series data indicating the position of a delivery person terminal. [Explanation of symbols]

[0086] 100 Estimation device, 200 Delivery person terminal, 300 Computer communication network, 110 Location information acquisition unit, 120 Travel speed estimation unit, 130 Partial time series extraction unit, 140 Drop-off / boarding position identification unit, 150 Door step time estimation unit, 160 Door step time memory unit, 170 Door step time prediction unit, 171 Delivery destination information acquisition unit, 172 Door step time information extraction unit, 173 Prediction unit, 174 Error estimation unit, 11 CPU, 12 RAM, 13 ROM, 14 Memory unit, 15 Input unit, 16 Display unit, 17 Communication unit, 99 Internal bus, C1, C2, C3, C4, C5, C6, C9, C10, C11 Parking position, C7, C8, C12 Center position, D1, D2, D3 Building.

Claims

1. An estimation device that estimates a door step time, which is a time from when a deliverer gets off a vehicle, delivers a package to a delivery destination, and gets on the vehicle, comprising: a location information acquisition unit that acquires a time series of a location of a first deliverer detected in a process in which the first deliverer delivers a first package to a first delivery destination using a first vehicle; a moving speed estimation unit that estimates a moving speed of the first deliverer at each of the positions from a time at which each of the positions included in the time series acquired by the position information acquisition unit was detected; a partial time series extraction unit that extracts a partial time series in which it is estimated that the first delivery person is not riding in the first vehicle from the time series acquired by the position information acquisition unit based on the moving speed estimated by the moving speed estimation unit; a partial time series selection unit that selects a partial time series to be associated with the first delivery destination from the partial time series extracted by the partial time series extraction unit; a door step time estimation unit that identifies, based on a preset rule, an alighting location where the first delivery person alights from the first vehicle and an aboarding location where the first delivery person aboards the first vehicle from the partial time series selected by the partial time series selection unit, and estimates a first parking location where the first delivery person parks the first vehicle to deliver the first package to the first delivery destination based on the identified alighting location and the aboarding location, and estimates a door step time for the combination of the first delivery destination and the first parking location from the time when the identified aboarding location was detected and the time when the identified alighting location was detected; An estimation device comprising:

2. the doorstep time estimation unit specifies the disembarking position based on a leading portion of the position data having an earlier detection time included in the partial time series selected by the partial time series selection unit, and specifies the boarding position based on a trailing portion of the position data having a later detection time. The estimation device according to claim 1 .

3. the doorstep time estimation unit extracts, from the selected partial time series, a set number of pieces of position data of the beginning portion and a set number of pieces of position data of the end portion, the number of pieces being based on a set ratio with respect to the total number of data included in the partial time series, and specifies one of the extracted pieces of position data of the beginning portion as the disembarking position, and specifies one of the extracted pieces of position data of the end portion as the boarding position. The estimation device according to claim 2 .

4. the door step time estimation unit identifies a combination of the position data of the leading portion and the position data of the trailing portion that are closest in distance from among the extracted position data of the leading portion and the position data of the trailing portion, and identifies the identified position data of the leading portion and the identified position data of the trailing portion as the disembarking position and the boarding position, respectively. The estimation device according to claim 3 .

5. The door step time estimation unit specifies a center position between the specified disembarking position and the boarding position as the first parking position. The estimation device according to any one of claims 1 to 4.

6. the partial time series selection unit selects a partial time series that is closest to the address of the first delivery destination and associates it with the first delivery destination; The estimation device according to any one of claims 1 to 4.

7. A computer that estimates a door step time, which is a time from when a delivery person gets off a vehicle, delivers a package to a delivery destination, and gets on the vehicle, A step of acquiring a time series of a position of a first deliverer detected during a process in which the first deliverer delivers a first package to a first delivery destination using a first vehicle; estimating a moving speed of the first deliverer at each of the locations from a time at which each of the locations was detected, the time series being acquired; Extracting a partial time series in which it is estimated that the first delivery person is not riding in the first vehicle from the acquired time series based on the estimated moving speed; selecting a sub-time series to be associated with the first delivery destination from the extracted sub-time series; from the selected partial time series, based on a preset rule, identifying a drop-off location where the first delivery person dropped off the first vehicle and a boarding location where the first delivery person boarded the first vehicle, estimating a first parking location where the first delivery person parked the first vehicle to deliver the first package to the first delivery destination based on the identified drop-off location and the boarding location, and estimating a door step time for the combination of the first delivery destination and the first parking location from the time when the identified boarding location was detected and the time when the identified drop-off location was detected; An estimation method including:

8. A computer that estimates a door step time, which is the time from when a delivery person gets off a vehicle, when the delivery person delivers a package to a delivery destination, to when the delivery person gets on the vehicle, A process of acquiring a time series of a position of a first deliverer detected in a process in which the first deliverer delivers a first package to a first delivery destination using a first vehicle; A process of estimating a moving speed of the first deliverer at each of the positions from a time at which each of the positions included in the acquired time series was detected; extracting a partial time series in which it is estimated that the first delivery person is not riding in the first vehicle from the acquired time series based on the estimated moving speed; selecting a partial time series to be associated with the first delivery destination from the extracted partial time series; a process of identifying, from the selected partial time series based on a preset rule, a drop-off location where the first delivery person drops off from the first vehicle and a boarding location where the first delivery person boards the first vehicle, estimating a first parking location where the first delivery person parks the first vehicle to deliver the first package to the first delivery destination based on the identified drop-off location and the boarding location, and estimating a door step time for the combination of the first delivery destination and the first parking location based on the time when the identified boarding location was detected and the time when the identified drop-off location was detected; A program that executes the following.

Images