Transportation management system and transportation management method

Abstract

In the present invention, on the basis of transportation reservation information, transportation performance information, and event information of a transportation target area, the number of reservations for each time segment in a predetermined period following the present time is predicted for each point in the transportation target area, transportation time required for transportation between respective points is predicted on the basis of traffic information, arrival time at which transportation equipment for which an instruction for the transportation has already been output will arrive at a destination thereof is predicted on the basis of the transportation time, the number of units of the transportation equipment that can be used at each point is predicted for each time segment, the time required to handle the reservations and transportation distance of the transport equipment handling the reservation are predicted on the basis of the predicted number of reservations for each time segment at each point, and the number of available units of transportation equipment, a plan for the transportation is created using an objective function in which the required time and the transportation distance are evaluation indices, and transportation instructions are output to a plurality of units of the transport equipment, and also a mobile terminal of a customer is notified of the schedule.

Description

Mobility management system and mobility management method

[0001] The present invention relates to a movement management system and a movement management method for transport machines that transport people and goods.

[0002] In recent years, mobility services that utilize information and communication technology to transport people and goods on demand have been expanding. Examples of mobility services include ride-hailing services that allow users to request a taxi to pick them up using a mobile phone application, and food delivery services that allow users to request home delivery of food ordered from a restaurant.

[0003] In such mobility services, customers who receive the service are required to reduce the time it takes for the service to be provided, i.e., the waiting time, in order to improve service quality. On the other hand, service providers are required to reduce the distance traveled by transportation vehicles to provide the service, thereby reducing energy costs, labor costs, and greenhouse gas emissions, in order to reduce costs and the environmental impact. Therefore, transportation vehicle movement management systems must take these multiple indicators into consideration when creating transportation vehicle movement plans and issuing movement instructions to transportation vehicles.

[0004] In response to this, Patent Document 1 discloses a vehicle dispatch planning device including: "a reception unit that receives transportation request information including the respective locations of the transportation source and the transportation destination of the transportation object; and a creation unit that acquires vehicle information including location information of a plurality of vehicles traveling in each of a plurality of regions from the transportation source to the transportation destination, and creates a vehicle dispatch plan based on the acquired location information and the transportation request information received by the reception unit so that the vehicles are positioned within an area where each of a plurality of regions intersect within a predetermined time range."

[0005] Japanese Patent Application Laid-Open No. 2021-196958

[0006] In Patent Document 1, fixed values ​​are set in advance for the travel time and other parameters on the transportation route, and a vehicle dispatch plan is determined by calculating the travel time and other parameters based on the preset fixed values. However, in mobility services that require flexible response to on-demand demand, for example, the number of customer reservations increases and decreases in real time. Furthermore, travel times between destinations may fluctuate significantly due to constantly changing traffic conditions. A system is needed that can flexibly respond to such situations, appropriately evaluate indicators such as customer waiting times and travel distances of transportation vehicles, create transportation vehicle movement plans, and manage the movement of transportation vehicles.

[0007] An object of the present invention is to provide a movement management system and a movement management method that can create a movement plan for a transport machine in a flexible manner in response to fluctuations in the required travel time of the transport machine and reservations.

[0008] In order to achieve the above object, the present invention provides a movement management system including a reservation information receiving unit that receives reservation information related to transportation reservations, a movement plan creating unit that creates a movement plan for a plurality of transportation machines in accordance with the reservation information, and a movement instruction unit that outputs movement instructions to the plurality of transportation machines based on the movement plan, the movement management system including a demand status prediction unit that predicts the number of reservations for each time section for a predetermined period from the current time for each of a plurality of points set in the transportation target area based on the reservation information, transportation performance information, and event information for the transportation target area, a movement time prediction unit that calculates the movement time required for the transportation machine to move between each of the plurality of points based on traffic information for the transportation target area, and a demand status prediction unit that calculates the movement time predicted by the movement time prediction unit based on current status information of the plurality of transportation machines and the movement time predicted by the movement time prediction unit. and an arrival time prediction unit that predicts the arrival time at which the transport machine to which the movement instruction has been output will arrive at its destination point and predicts the number of the transport machines that will be available at each point for each time interval, the movement plan creation unit predicts the time required to fulfill the reservation and the movement distance of the transport machine that will fulfill the reservation based on the number of reservations for each time interval at each point predicted by the demand status prediction unit and the number of the transport machines that will be available at each point for each time interval predicted by the arrival time prediction unit, and creates the movement plan using an objective function that uses at least one of the required time and the movement distance as an evaluation index, and the movement instruction unit outputs movement instructions to the plurality of transport machines based on the movement plan and notifies customers of their schedules to their mobile terminals.

[0009] According to the present invention, a movement plan for a transport machine can be created in a flexible manner in response to fluctuations in the required travel time of the transport machine and in reservations.

[0010] 8A . FIG. 8B is a functional block diagram of a mobility management system according to a first embodiment of the present invention. FIG. 8C is a schematic diagram of a vehicle dispatch network in an on-demand taxi service to which the mobility management system according to the first embodiment of the present invention is applied. FIG. 8D is a diagram showing an example of a reservation screen (a screen for inputting a departure point and a destination point) on a customer's mobile device in the mobility management system according to the first embodiment of the present invention. FIG. 8E is a diagram showing an example of a reservation screen (a screen for inputting additional information) on a customer's mobile device in the mobility management system according to the first embodiment of the present invention. FIG. 8F is a diagram showing an example of a data table showing a list of reservation information in the mobility management system according to the first embodiment of the present invention. FIG. 8G is a diagram showing an example of transportation performance data, and is a graph showing the number of transportations for each time interval with a certain node as the departure point. FIG. 8H is a graph showing the number of reservations for each time interval with a certain node as the departure point, predicted using the transportation performance data. FIG. 8I is an example of a graph showing the relationship between traffic volume and travel time between certain points. FIG. 8I is a diagram showing an example of travel instructions sent to a transportation machine displayed on a terminal mounted on the transportation machine. FIG. 8I is a diagram showing an example of the positional relationship between a waiting customer, their destination, and the vehicle in a vehicle dispatch network. FIG. 8I is an explanatory diagram of travel instructions when a comparative example is applied to the example of FIG. 8A . FIG. 8I is an explanatory diagram of an example of travel instructions when a mobility management system according to the present invention is applied to the example of FIG. 8A . FIG. 8I is an explanatory diagram of an example of changes in travel instructions by the mobility management system according to the present invention according to traffic conditions. 1 is an explanatory diagram of an example of changes in travel instructions by the mobility management system according to the present invention in accordance with the time priority / price priority dispatch options specified by each customer in their reservation information, and is a diagram showing an example of the relative positions of waiting customers, their destinations, and vehicles; FIG. 2 is an explanatory diagram of an example of changes in travel instructions by the mobility management system according to the present invention in accordance with the time priority / price priority dispatch options specified by each customer in their reservation information, and is a diagram showing an example of an example where both waiting customers wish to prioritize time; FIG. 3 is an explanatory diagram of an example of changes in travel instructions by the mobility management system according to the time priority / price priority dispatch options specified by each customer in their reservation information, and is a diagram showing an example where one customer wishes to prioritize time and the other customer wishes to prioritize price;1 is an explanatory diagram of an example of changes in travel instructions by the mobility management system according to the present invention in accordance with the dispatch options of time priority / price priority specified by each customer in the reservation information, and is a diagram showing an example in which one customer desires price priority and the other customer desires time priority. FIG. 2 is a flowchart showing an example of the processing procedure of the mobility management system according to the first embodiment of the present invention. FIG. 3 is a schematic diagram showing an example of an in-warehouse transport system to which the second embodiment of the present invention is applied. FIG. 4 is a diagram showing an example of a data table showing a list of reservation information (work information) in the mobility management system according to the first embodiment of the present invention.

[0011] Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

[0012] The movement management system of the present invention is one or more servers capable of communicating with multiple transport vehicles, or software such as a program executed on a server. The movement management system is a system that creates a movement plan for each transport vehicle, including information on the destination, arrival time at the destination, and route to the destination, and transmits movement instructions to each transport vehicle based on the movement plan, so that multiple transport vehicles under its management can efficiently perform transportation work.

[0013] Here, a transportation machine is a means of transportation capable of transporting people or goods, and includes vehicles such as taxis that transport people, as well as home delivery robots that transport goods outdoors and transport robots R that perform transport work in logistics warehouses. The scope of application is not limited to land transportation means such as automobiles, but also includes drones and the like. The transportation machine to be managed may be a machine that moves with a driver on board or remotely controlled, or may be a machine that moves automatically under computer control.

[0014] First Embodiment - Mobility Management System - FIG. 1 is a functional block diagram of a mobility management system according to a first embodiment of the present invention. In this embodiment, an example of applying the present invention to the dispatch management of an on-demand taxi is described. In an on-demand taxi, a vehicle V (taxi), an example of a transportation machine, performs transportation work in response to a request from a customer P. This transportation work is performed, for example, in response to a call via an application on a mobile device M that can connect to the Internet, such as a smartphone owned by the customer P. An available vehicle V travels to the customer P holding the mobile device M, picks up the customer P, picks up the customer P, and then travels with the customer P and drops off the customer P at the destination. To perform this series of transportation work, the mobility management system A100 receives various information, including reservation information, from the mobile device M of the customer P. The mobility management system A100 creates a travel plan for each vehicle V, including the destination points for each of the pick-up and transportation trips, as well as the travel route and target time from the departure point to the destination. Based on the travel plan, the mobility management system A100 transmits travel instructions to the vehicle V and notifies the driver. If the vehicle V is an autonomous vehicle, the travel instructions are issued by a controller (control device) of the vehicle V. However, in the following description, the vehicle V is assumed to be manually driven by an on-board driver. Furthermore, the mobility management system A100 notifies the customer P of the schedule, such as the arrival time of the vehicle 001, via the mobile terminal M. Note that communication between the mobility management system A100 and the mobile terminal M is performed via a necessary network (not shown), such as the Internet.

[0015] FIG. 2 is a schematic diagram of a dispatch network for vehicles V to customers P. The dispatch network shown in FIG. 2 is expressed in a graph structure. The nodes of the graph, indicated by dots in FIG. 2, indicate the boarding and disembarking points of customers P within the transportation target area, that is, within the transportation target area (service provision area) of the on-demand taxi. The edges (also called links) of the graph, indicated by lines in FIG. 2, indicate the routes between the nodes. Data on these nodes and edges is stored, for example, in the memory of the mobility management system A100.

[0016] A node is a specific location where a customer P actually gets on and off, and for example, a location near a station, a store, or a public facility where the customer P frequently calls the vehicle V is set as a node in advance. In addition to the preset locations, any location where the customer P calls the vehicle V can be temporarily or additionally set as a node based on the location information of the mobile terminal M of the customer P.

[0017] Furthermore, the node does not necessarily have to coincide with the destination point, but may be a representative point of a specific area. For example, the region covered by the on-demand taxi service can be divided into an appropriate number of areas, and the center point of each area can be set as a representative point and a node. In this case, the specific route from each node to the customer P's call point can be calculated at the stage of creating the travel plan based on the location information of the mobile terminal M of the customer P who made the call and the location information of the destination.

[0018] An edge is a representative route that can be traveled between nodes, and when there are multiple routes connecting nodes, for example, the shortest route or the route that can be traveled in the shortest time can be set as an edge.

[0019] In order to realize processes such as creating a movement plan, issuing movement instructions to vehicles V, and sending notifications to mobile terminals M, the movement management system A100 includes, as shown in FIG. 1 , a reservation information receiving unit A001, a demand status prediction unit A002, a movement time prediction unit A003, an arrival time prediction unit A004, an objective function setting unit A005, a constraint condition setting unit A006, a movement plan creation unit A007, and a movement instruction unit A008. These processing units can be realized as hardware elements such as circuits that execute each process, or software elements such as programs. In general, the movement management system A100 receives reservation information related to transportation reservations in the reservation information receiving unit A001, creates movement plans for multiple vehicles V in accordance with the reservation information in the movement plan creation unit A007, and outputs movement instructions to multiple vehicles V based on the movement plan in the movement instruction unit A008. The processing contents of each processing unit will be described in detail below.

[0020] - Reservation Information Receiving Unit - The reservation information receiving unit A001 receives reservation information regarding transportation reservations from the customer P and manages the received reservation information. The customer P operates an application corresponding to the on-demand taxi service on the mobile terminal M, connects to the mobility management system A100, and makes a reservation for the vehicle V.

[0021] 3A is a diagram showing an example of a reservation screen for a vehicle V of an application on a mobile terminal M of a customer P. On the screen of FIG. 3A, the customer P selects a departure point, which is the point where the customer will board the vehicle V, and a destination point, which is the point where the customer will disembark from the vehicle at their destination. The departure point and destination point can be specified by selecting them on a map displayed on the application screen, or by entering an address. Furthermore, if the mobile terminal M is equipped with a GNSS (Global Navigation Satellite System) receiver, the departure point and destination point can be entered by selecting the current location displayed on a map based on data from the GNSS receiver.

[0022] On the screen of FIG. 3B , customer P inputs additional information related to the reservation. The additional information includes information about the desired pick-up time and arrival time at the destination. The pick-up time is the desired arrival time of vehicle V at the departure point. The additional information can also include boarding requirements input by customer P related to vehicle selection, such as the number of passengers, whether or not there is large luggage or a wheelchair, and other items. The input format for the desired time and boarding requirements is not limited, and may be an input method using a pull-down menu or a method of directly inputting text.

[0023] In addition, the additional information may include a dispatch option as a selection condition according to the customer P's convenience and wishes. In the example of FIG. 3B, "time priority" and "price priority" are shown as dispatch option choices. "Time priority" is an option in which the customer wishes to prioritize service provision time over price, and wishes to arrive at the destination as quickly as possible even if it is a high price. "Price priority" is an option in which the customer wishes to prioritize price over service provision time, and wishes to travel to the destination as cheaply as possible even if it takes longer.

[0024] Although FIG. 3B illustrates a method for selecting either "time priority" or "price priority" using radio buttons, it is also possible to adopt a method for specifying the degree of importance given to each priority using a numerical value (e.g., a five-level number) that indicates the degree of importance given to each priority.

[0025] 3B is used as basic information for adjusting the parameters of the objective function and setting constraints when creating a travel plan, which will be described later. Various methods for adjusting the price are possible, but for example, if there are many reservations for the same time slot and the desired times of all customers P cannot be met, dynamic pricing can be performed, in which the service price is lowered in exchange for lowering the priority of vehicle dispatch for customers P with a relatively high price priority.

[0026] FIG. 4 is a diagram showing an example of a data table listing reservation information. The reservation information receiving unit A001 stores the received reservation information in a database in memory and uses it for subsequent processing. In the example of FIG. 4, the "reservation time," "departure point," "destination point," "desired pick-up time," "desired arrival time," "number of passengers," "other," and "dispatch options" are stored for each reservation. The "reservation time" is, for example, the time each reservation is received by the mobility management system A100. It may also be the reservation transmission time of customer P's mobile terminal M. The "departure point," "destination point," "desired pick-up time," "desired arrival time," "number of passengers," "other," and "dispatch options" are items entered in the example screens of FIGS. 3A and 3B. For the departure point and destination point, coordinates or latitude and longitude data representing specific locations may be used, or a node ID may be used by calculating a correspondence relationship with a node on the dispatch network shown in FIG. 2.

[0027] Demand Status Prediction Unit: The demand status prediction unit A002 predicts the number of reservations for each time interval for a predetermined period from the current time for each of multiple locations (nodes in this example) set in the transportation target area based on the received reservation information, past transportation performance information, and event information for the transportation target area. That is, the demand status prediction unit A002 predicts the expected number of reservations for each node as the departure point for each time interval. Transportation performance information is data recorded in memory regarding transportation work performed by vehicles V starting from each node, and indicates the number of transportation work performed for each time interval corresponding to each time interval from the current time during a certain period in the past. Event information is information regarding sudden or non-regular events that occur in the transportation target area and affect the movement of vehicles V, including, for example, information on the suspension or delay of public transportation such as railways, information on major events, and road traffic information such as accidents, congestion, and traffic restrictions. For example, if a railway is suspended, an increase in the number of reservations for departure points from stations in the immediate time period is expected. The time intervals are preset fixed time intervals, such as five-minute or fifteen-minute intervals. The length of the time interval may be the same as the output period of movement instructions to the vehicle V according to the movement plan created by the movement plan creation unit A007, or may be different from the output period.

[0028] FIG. 5A is a diagram showing an example of transportation performance data, and is a graph showing the number of transportation items for each time interval starting from a certain node. The transportation performance data can be statistical data of data from multiple working days. In this case, for example, the average number of transportation items for each time period on multiple working days can be used, or it can be expressed probabilistically using variance. Furthermore, the transportation performance data can be data by day of the week, or data distinguished between weekdays and holidays.

[0029] FIG. 5B is a graph showing the number of reservations for each time interval starting from a certain node, predicted using transportation performance data. The nodes and days of the week in FIG. 5B correspond to the nodes and days of the week in FIG. 5A. First, the demand status prediction unit A002 can tally up the number of reservations for each time interval in which a vehicle is desired to be picked up at a certain location as the number of confirmed reservations based on the reservation information managed by the reservation information receiving unit A001. In addition, by using the transportation performance data shown in FIG. 5A, the number of additional reservations that will occur in each time interval can be estimated and the number of reservations that will occur in each time interval before the arrival of each time interval can be predicted. FIG. 5B shows an example in which the number of reservations that are insufficient compared to the transportation performance data for each time interval is simply supplemented as the predicted number of additional reservations. However, the algorithm for predicting the number of reservations can be modified as appropriate. For example, if there is a time interval in which the number of confirmed reservations already exceeds the transportation performance data, an algorithm can be applied in which the difference obtained by subtracting the transportation performance data from the number of confirmed reservations is subtracted from the predicted number of additional reservations for the nearby time interval.

[0030] As mentioned above, event information in the target transportation area can also be reflected in the forecast of reservation numbers. For example, an increase in the number of reservations departing from a station can be predicted based on information about train service suspensions. The actual expected increase in reservation numbers can be set as a parameter in advance, or it can be calculated based on the number of train users during normal times during the same time period by, for example, sharing information with the railway company.

[0031] - Travel Time Prediction Unit - The travel time prediction unit A003 calculates the travel time required for the transport machine to travel between each of the multiple points, specifically the running time of each edge, based on traffic information for the transportation target area. For example, the traffic volume on the route represented by each edge is used as traffic information. When there are multiple routes represented by an edge, the traffic volume for each route can be taken into account, but a simple example is a method using the traffic volume of a predetermined representative route. Traffic volume is the number of passing vehicles per unit time at a fixed point on the edge, and can be calculated based on information obtained from infrastructure systems (cameras, etc.) installed in the transportation target area. Traffic information used in general car navigation systems can also be used.

[0032] FIG. 6 shows the relationship between traffic volume and travel time at a certain edge. The travel time prediction unit A003 also calculates actual travel time data for each vehicle V, for example, from the position information for each vehicle V, and records this data in memory along with the traffic volume for each edge when the vehicle V traveled. The relationship between traffic volume and edge travel time can then be calculated by deriving an approximate straight line from the actual data. As shown by the dotted line in FIG. 6 , it is possible to obtain an approximate straight line in which the edge travel time also drops to nearly zero when traffic volume approaches zero. Therefore, the shortest travel time Tmin for each edge when traveling at the legal speed is calculated, and the lower limit of the edge travel time is limited by Tmin. In this way, the travel time prediction unit A003 can predict the travel time for each edge based on the traffic volume for each edge predicted for each time period using an approximate straight line based on the actual data.

[0033] The relationship between traffic volume and edge travel time does not necessarily have to be an approximate straight line, i.e., a linear function, as shown in FIG. 6, but may be expressed by a higher-order equation, such as a quadratic function.

[0034] The arrival time prediction unit A004 predicts the arrival time at the destination of each vehicle V for which a movement instruction has been output, i.e., each vehicle V currently performing transportation work, based on the current status information of each vehicle V and the travel time of each edge predicted by the movement time prediction unit A003. As a result, the arrival time prediction unit A004 predicts the number of vehicles V available at each node for each time interval.

[0035] The current status information of each vehicle V used to calculate the arrival time includes the movement instructions given to each vehicle V, and the state and position of the vehicle V. The state of the vehicle V is information on whether the vehicle V is performing transportation work or waiting. If the vehicle V is performing transportation work, the position of the vehicle V is a combination of the node ID where the vehicle V is located, or an edge ID and the progress rate on the edge. If the vehicle V is waiting, the position is the node ID where the vehicle V is located. Note that this expression of vehicle position using node IDs and edge DI is based on the assumption that the vehicle dispatch network is represented in a graph structure, but the vehicle position can also be expressed using a coordinate system or latitude and longitude that can be used in the transportation target area.

[0036] Based on this current status information for each vehicle V, for a vehicle V currently performing transportation work, it is possible to calculate the time and location (node) at which the vehicle V will complete the movement instruction and become available again, based on the movement instruction being executed and the location information of the vehicle V. Also, for a vehicle V that is waiting, it can be determined at the waiting node that it will be available from the current time onwards. In this way, it is possible to predict the number of vehicles V available at each node for each time interval.

[0037] - Objective Function Setting Unit - The objective function setting unit A005 sets an objective function that reflects the reservation information managed by the reservation information receiving unit A001, for example, based on the dispatch options (FIG. 3B) that are the selected and desired conditions of the customer P. The objective function set by the objective function setting unit A005 in accordance with the selected and desired conditions of the customer P is used by the movement plan creation unit A007 to create a movement plan.

[0038] The objective function will be described in detail later, but one of the basic evaluation indexes taken into account in the objective function is the waiting time of the customer P. The waiting time of the customer P can be calculated, for example, as the difference between the desired pick-up time set for each reservation and the predicted time when the vehicle V actually arrives at the pick-up point. In the objective function, an evaluation weight can be assigned to the waiting time of the customer P for each reservation, which is an evaluation index. In this case, a large evaluation weight is set for the waiting time for a customer P who relatively prioritizes time over price based on the dispatch option set by the customer P, and a small evaluation weight is set for the waiting time for a customer P who relatively prioritizes price over time.

[0039] Constraint Condition Setting Unit The constraint condition setting unit A006 sets constraint conditions used by the movement plan creation unit A007 to create a movement plan based on the reservation information managed by the reservation information receiving unit A001, for example, boarding requirements ( FIG. 3B ) such as the number of passengers or baggage information entered by the customer P. For example, the constraint condition setting unit A006 sets constraint conditions such as selecting a type of vehicle V that can perform transportation work for the reservation and assigning it to the reservation, based on the boarding requirements such as the number of passengers and whether or not there is large baggage entered by the customer P as reservation information.

[0040] -Movement plan creation unit- The movement plan creation unit A007 predicts the time required for vehicle V to arrive at the departure point associated with the reservation and the travel distance of vehicle V that accepts the reservation based on the number of reservations for each time interval with each node as the departure point predicted by the demand status prediction unit A002, the travel time of each edge predicted by the travel time prediction unit A003, and the number of available vehicles for each time interval at each node predicted by the arrival time prediction unit A004, and creates a movement plan using an objective function that uses at least one of the required time and travel distance as an evaluation index. In this embodiment, the objective function is set to reflect the selection items of the dispatch options input as the desired selection conditions in the objective function setting unit A005, as described above. Therefore, the movement plan creation unit A007 solves an optimization problem for this objective function and creates a movement plan by searching for conditions that minimize the objective function. Furthermore, the movement plan creation unit A007 creates a movement plan by, for example, selecting the type of vehicle V corresponding to the constraint conditions (number of passengers or baggage information) so as to satisfy the constraint conditions set by the constraint condition setting unit A006. That is, in the case of this embodiment, the movement plan creation unit A007 searches for conditions for each reservation that satisfy the constraint conditions set by the constraint condition setting unit A006 and that minimize the objective function set by the objective function setting unit A005, and creates a movement plan.

[0041] The movement plan calculated by the movement plan creation unit A007 is a set of virtual movement instructions including the departure point and destination point for each vehicle V at the start time of each time interval within a predetermined period, the period starting from the current time being N time intervals, and the travel route between these two points. N is the number of time intervals (number of predicted steps), the total number of nodes in the vehicle dispatch network is Ns, the current time is time 0, and the length of the time interval is Δt. The movement plan creation unit A007 creates a movement plan including virtual movement instructions for each vehicle V at times 0, Δt, 2Δt, ..., (N-1)Δt for all combinations of nodes 1, ..., Ns as departure points and nodes 1, ..., Ns as destination points. The movement instructions for each time can be calculated, for example, by solving an optimization problem as described below.

[0042] However, the movement plan is merely a collection of virtual movement instructions and is data temporarily stored in the memory of the movement management system A100, and at the stage of creation, it is not actually transmitted to the vehicles V. At the stage of creating the movement plan, it is not necessarily necessary to decide which movement instructions will actually be sent to which vehicles V; it is sufficient to be able to calculate a plan for the number of vehicles moving between nodes.

[0043] -Movement instruction unit- The movement instruction unit A008 actually outputs movement instructions to each target vehicle V based on the movement plan created by the movement plan creation unit A007, and notifies the mobile terminal M of the customer P of schedules such as the estimated arrival time of the vehicle V at the specified departure point and the estimated arrival time at the destination. In this embodiment, virtual movement instructions for each of N time intervals are created as the movement plan for each vehicle V waiting at each point, but the movement instruction unit A008 actually transmits only the movement instruction calculated for the first time interval (i.e., time 0) to the vehicle V via a communication line.

[0044] 7 is a diagram showing an example of a state in which a travel instruction transmitted to the vehicle V by the travel instruction unit A008 is displayed on the in-vehicle terminal T. The in-vehicle terminal T is, for example, a terminal such as a tablet that can be connected to the Internet and can be checked by the driver inside the vehicle V. In the example of FIG. 7, vehicle dispatch information and a map are displayed on the screen of the in-vehicle terminal.

[0045] In the example of Figure 7, the dispatch information displays the pick-up point and destination point corresponding to the reservation information, as well as the target pick-up time and target arrival time, which are the target times of arrival at those points. Each target time may not match the desired pick-up time and desired arrival time in the reservation information entered by customer P. The dispatch information displays the target pick-up time and target arrival time that are finally calculated under various constraints such as the reservation volume, traffic conditions, and the reservation details of customer P. These target pick-up time and target arrival time are also notified to customer P's mobile terminal M as the scheduled pick-up time and scheduled arrival time.

[0046] In addition, the map displayed on the in-vehicle terminal T displays the locations on the map of the pick-up point and destination point, as well as the driving route from the current location to the pick-up point and the route from the pick-up point to the destination point.

[0047] After the transportation work is completed, the driver can notify the mobility management system A100 of the completion of the transportation work by operating the in-vehicle terminal T (the operation screen is not shown). By receiving the notification from the in-vehicle terminal T, the mobility management system A100 can determine that the vehicle V has transitioned to a standby state.

[0048] --Movement plan creation details-- The mobility management system A100 periodically and repeatedly predicts the demand situation and the number of available vehicles for each time interval at each node up to a predetermined time period into the future, and therefore predicts each time a change in the number of customers and vehicles for each time interval that occurs at each node as a result of issuing a movement instruction to vehicle V, creates a movement instruction for each time interval as a movement plan, and transmits a movement instruction for the first time interval to vehicle V each time. This can be realized within the framework of model predictive control (hereinafter referred to as MPC), which periodically solves an optimal control problem up to a finite time into the future starting from the current time, and determines the current input to each vehicle V under management.

[0049] An example of the specific calculation contents of the movement plan creation unit A007, which has the function of constructing an optimal control problem within the framework of MPC and finding a solution, will be described below.

[0050] First, as control inputs, let pijk and qijk denote the number of vehicles in empty state and in transit state (with customer P aboard) departing from node i to node j at time k, respectively. Nodes i and j (i ≠ j) are i, j = 1, ..., Ns. As described above, Ns is the number of nodes on the vehicle dispatch network. Time k is k = 0, ..., N-1, with the current time being 0. N is the number of time intervals (number of prediction steps). Here, nodes i and j do not need to be adjacent on the vehicle dispatch network. For example, p314 represents a variable representing the number of vehicles departing in empty state from node 3 to node 1 at time k = 4. Similarly, q582 represents a variable representing the number of vehicles departing in transit state from node 5 to node 8 at time k = 2. These control inputs are used to construct a prediction model for the number of customers and vehicles at each node.

[0051] The prediction model for the number of customers at each node is given by the following equation 1, where rijk is the number of customers waiting at each node i with node j as their destination at time k.

[0052]

[0053] Here, the second term on the right side of Equation 1 is the total number of vehicles departing in transport mode from node i as the departure point to another node at time k. Also, Δrijk on the third term on the right side is the number of customers whose destination is node j, which newly appears at node i between times k and k+1. The value of Δrijk can be determined using the calculation results of the demand status forecasting unit A002.

[0054] The prediction model for the number of vehicles at each node is given by Equation 2, where vi is the number of vehicles present at node i at time k.

[0055]

[0056] Here, the second term on the right side of Equation 2 is the sum of the number of vehicles in an empty or transporting state departing from node i to another node at time k. Furthermore, the third term on the right side is the number of vehicles that arrive at node i between times k and k+1 among the vehicles that departed from any node after time k=0. Here, Kji is the number of time steps (unit time) required to travel from node j to node i, and can be calculated based on the travel time of each edge calculated by the travel time prediction unit A003 and the time interval increment Δt.

[0057] The set S i k is given by Equation 3.

[0058]

[0059] This set S i k represents a set of nodes that can be reached by a vehicle that departs after time k=0 to node i at time k.

[0060] Δvik in the fourth term on the right side of Equation 2 is the number of vehicles that arrive at node i between times k and k+1 among vehicles that departed from any node before time k = 0. The value of Δvik can be determined using the calculation results of the arrival time prediction unit A004.

[0061] Here, the number of customers rijk and the number of vehicles vik, pijk, qijk must be integers, and this condition is given by Equation 4.

[0062]

[0063] Here, Z0+ represents a set of integers equal to or greater than 0.

[0064] Furthermore, if the number of vehicles managed by the mobility management system A100 is Nv, then an upper limit constraint is given to the variables related to the number of vehicles as shown in Equation 5.

[0065]

[0066] The number of customers rijk and the number of vehicles vik, pijk, qijk are optimization variables that are determined to minimize the objective function. In this embodiment, the waiting time of customer P and the travel distance of vehicle V are used as evaluation indices for the components of the objective function.

[0067] The total waiting time Jw of each customer P within the N time intervals is given by Equation 6.

[0068]

[0069] The total value Jw is the total number of waiting customers multiplied by the time interval Δt.

[0070] Furthermore, the total value Jd of the travel distance of each vehicle V within the N time intervals can also be used as an evaluation index constituting the objective function. The total value Jd of the travel distance of each vehicle V is given by Equation 7.

[0071]

[0072] Here, D ij is the distance traveled between node i and node j, i.e., it is the number of vehicles traveling between each node multiplied by the distance traveled between the nodes.

[0073] The objective function for weighting and evaluating these evaluation indexes (total waiting time Jw of each customer P, total travel distance Jd of each vehicle V) is given by Equation 8.

[0074]

[0075] Here, ww and wd are evaluation weights relating to the waiting time of customer P and the travel distance of vehicle V, respectively. In this way, the objective function includes the product of the evaluation index and the evaluation weight relating to the desired selection conditions (dispatch options). Note that in this example, an example is described in which both time and travel distance (price) are considered as indices, but an algorithm can also be used that considers only one of time and travel distance as an index.

[0076] Based on the above preparations, the movement plan creation unit A007 solves the optimization problem of Equation 9 and determines the numbers of vehicles p ijk and q ijk to be dispatched, which are control inputs.

[0077]

[0078] Here, r, v, p, and q are vectors that list all variables related to i, j, and k in rijk, vik, pijk, and qijk, respectively. The number of vehicles moving between each node in an empty and loaded state at each time step, obtained by solving the optimization problem of Equation 9, serves as basic information for planning a series of movement instructions. Based on this number of vehicles, vehicles V to which movement instructions are actually given are assigned based on the state of each vehicle V (empty state / loaded state). However, since the movement instruction unit A008 only actually sends movement instructions to vehicles V at the initial time k=0, i.e., the start time of the first time interval among N time intervals, it is sufficient to assign vehicles V based on the solution of the above optimization problem only for k=0.

[0079] - Setting the objective function based on reservation information - In the above, for simplicity, an example has been explained in which all customers P are counted together and treated equally, but the variable for the number of customers can be divided and treated according to the characteristics of the customers P.

[0080] First, when time priority and price priority can be selected as dispatch options (Figure 3B) in reservation information, a customer P who selects time priority can have a relatively larger evaluation weight imposed on their waiting time compared to other customers P who do not select time priority.

[0081] For example, the number of customers rijk can be divided into the number r1ijk of customers P in a first group who have selected time priority and the number r2ijk of customers P in a second group who have selected price priority, and a customer number prediction model such as Formula 1 can be formulated for each group. In this case, the waiting time of customers P in each group can be evaluated as shown in Formula 6. If the evaluation values ​​of the waiting times of customers P in the first and second groups are Jw1 and Jw2, respectively, and the evaluation weights are ww1 and ww2, respectively, the objective function is given by Formula 10.

[0082]

[0083] However, although explanation has been omitted for simplicity, strictly speaking, it is also necessary to divide the variable for the number of vehicles corresponding to each customer P. Also, although an example in which customers P are divided into two groups has been explained, the objective function can be set using the same idea even if the number of groups is three or more. Therefore, the above procedure for setting the objective function is applicable even if the degree of time priority is further increased.

[0084] The above processing can be executed in the objective function setting unit A005 based on the reservation information.

[0085] Setting constraints based on reservation information When the available vehicle types are limited by constraints related to the number of passengers or baggage information in the boarding requirements of the reservation information (FIG. 3B), a travel plan can be created that takes this into consideration. For example, a vehicle V with a large capacity may be assigned to a reservation with a large number of passengers, a vehicle V with a large loading space may be assigned to a reservation with large baggage, or a vehicle V that can accommodate a wheelchair may be assigned to a wheelchair-using customer P.

[0086] To achieve this, in the same manner as described above for setting the objective function, the variable for the number of customers is divided according to the boarding requirements, and the variable for the number of vehicles is further divided for each vehicle type corresponding to the boarding requirements. Then, a prediction model of Equations 1 and 2 is formulated for the number of customers and the number of vehicles with corresponding characteristics, and the optimization problem is solved to calculate the travel plan. In this case, an optimization problem may be constructed for each group divided by characteristics to obtain a solution, or a single optimization problem that includes variables related to all characteristics may be constructed to obtain a solution. Note that when creating a single optimization problem, some vehicle types may be able to accommodate multiple characteristics, and therefore, by formulating a prediction model that takes these conditions into account, it is possible to create an even more efficient travel plan.

[0087] The above processing can be executed in the constraint setting unit A006 based on the reservation information.

[0088] --Example of Basic Operation--The operation realized by the movement plan of the method described above will be explained using a specific example.

[0089] Figure 8A shows a simplified example of the location of a waiting customer P, his destination, and vehicle V in a vehicle dispatch network. In Figure 8A, 16 grid-like lattice points are nodes of the vehicle dispatch network, and dotted lines are edges. It is assumed that waiting customers P1 and P2 have specified destinations G1 and G2, respectively, when reserving a vehicle. It is also assumed that both vehicles V1 and V2 are available for use. Here, it is assumed that all edges have the same distance, and that the travel time for all edges is the same regardless of whether vehicle V1 or V2 is used.

[0090] Under these conditions, first, the operation of a comparative example in which vehicles V1 and V2 are assigned to customers P1 and P2, respectively, without performing the vehicle dispatch management of the present invention is shown in Figure 8B. Here, vehicle V1 is assigned to customer P1 and vehicle V2 is assigned to customer P2, as a combination that minimizes the total traveling distance or time of vehicles V1 and V2. If the traveling distance of each edge is counted as 1, in this case, the traveling distance of vehicle V1 is 3, the traveling distance of vehicle V2 is 8, and the total traveling distance of vehicles V1 and V2 is 11. Similarly, if the traveling time of each edge is counted as 1, and attention is paid to the waiting time until customers P1 and P2 are picked up, the waiting time of customer P1 is 1, and the waiting time of customer P2 is 5.

[0091] In contrast, an example of travel instructions when the travel management system A100 of this embodiment is applied is shown in FIG. 8C . As described above, the travel plan creation unit A007 evaluates the waiting times of customers P1 and P2 and the travel distances of vehicles V1 and V2 for N × time intervals using an objective function, and calculates a travel plan that minimizes the objective function. Therefore, by setting N sufficiently large, it is possible to minimize the overall waiting times of customers P1 and P2 and the travel distances of vehicles V1 and V2. As a result, as shown in FIG. 8C , an operation is generated in which the same vehicle V1 is assigned to customers P1 and P2, rather than assigning both vehicles V1 and V2 to transportation tasks. That is, vehicle V1 first picks up and transports customer P1, then, starting from customer P1's destination point G1, continues to pick up and transport customer P2. Meanwhile, vehicle V2 remains at its initial position without moving. As a result, the total travel distance of vehicles V1 and V2 is 7, which is 4 less than the example of FIG. 8B . Furthermore, with regard to waiting time, the waiting time for customer P1 is 1, which is the same as in the case of Figure 8B, but the waiting time for customer P2 is 4, which is reduced by 1 compared to the case of Figure 8B. However, in this example, the boarding and disembarking time for customer P is not taken into consideration. As described above, when the mobility management system A100 of the present invention is applied, both the waiting time for customers P1 and P2 and the traveling distance of vehicles V1 and V2 are reduced compared to when a simple assignment is performed taking into consideration only the relationship between the initial positions of customers P1 and P2 and vehicles V1 and V2.

[0092] In addition, in the above example, the condition is that customer P is waiting from the initial time, but if customer P's desired pick-up time is specified in the reservation information, the vehicle travel time is taken into consideration and vehicle V is made to arrive at the departure point at the specified desired pick-up time, thereby eliminating or minimizing waiting time for customer P.

[0093] However, depending on the conditions of the reservation information, the location of the vehicle, and the status of the transportation work, the vehicle V may not arrive in time for the customer P's desired pick-up time, resulting in a waiting time for the customer P. In such cases, based on the movement plan created by the movement plan creation unit A007, it is possible to notify the customer P of the scheduled pick-up time and take measures to reduce the customer P's dissatisfaction, such as adjusting the service provision price.

[0094] Furthermore, even if no reservation information is actually received, proactive vehicle dispatching can be performed based on past transportation work information and event information. For example, the likelihood of receiving reservations with a train station as the departure point increases during rush hours in the morning and evening, weekend nights, or immediately after a train service suspension. If such time periods exist within a predetermined period of N x time segments, vehicles can be deployed to the required location in preparation for expected reservations by considering dummy reservation information with a provisional destination point. Furthermore, because travel planning is performed periodically, even if a prospective reservation is not actually received, another travel instruction is immediately issued based on the virtual future reservation information at that time, allowing for flexible vehicle dispatching in response to ever-changing conditions.

[0095] --Example of operation taking traffic conditions into consideration--Next, we will explain how travel instructions change depending on traffic conditions using Figure 9. The example in Figure 9 assumes a situation in which a road closure occurs on some edges between customer P1 and his destination point G1, and congestion occurs on edges that avoid the road closure. It is assumed that the travel time for edges experiencing congestion is three times longer than usual. The positional relationships between customers P1 and P2, destination points G1 and G2, and vehicles V1 and V2 are the same as in the example in Figure 8A.

[0096] In the example of Fig. 9, if the transportation work of customers P1 and P2 were performed using only vehicle V1 as in the example of Fig. 8C, it would take time for vehicle V1 to pass the edge of the congestion, and the waiting time of customer P2 would increase. Specifically, the waiting time of customer P2 would increase from 4 in the example of Fig. 8C to 6.

[0097] The movement plan creation unit A007 also takes into account increases in travel time due to the influence of such traffic conditions. Conditions that reduce the evaluation value of the objective function are searched for, and a more optimal movement plan is created. In the example of Figure 9, a movement plan is created in which vehicle V1 is assigned to customer P1 and vehicle V2 is assigned to customer P2. As a result, the waiting time for customer P2 is 5, which is an appropriate reduction in the waiting time for customers P1 and P2 depending on the situation, compared to the example of Figure 8C, in which transportation work for customers P1 and P2 is performed consecutively using only vehicle V1.

[0098] - Example of operation reflecting dispatch options (time priority / price priority) - Next, using Figures 10A to 10D, we will explain an example of changes in travel instructions due to the dispatch options of time priority / price priority specified by customers P1 and P2 in their reservation information.

[0099] 10A shows the positional relationship between customers P1 and P2, their destinations G1 and G2, and vehicles V1 and V2 in this example. As in the examples of FIGS. 8A to 8C, in this example too, the travel distance and travel time of each edge are counted as 1.

[0100] First, an example of travel instructions calculated when both customers P1 and P2 specify time priority is shown in Figure 10B. When both customers P1 and P2 specify time priority, a travel plan is created that minimizes the waiting time for customers P1 and P2. If there is no traffic congestion, a travel plan is created in which vehicle V1 transports customer P2 and vehicle V2 transports customer P1. In this case, the waiting time for customer P1 is 3, and the waiting time for customer P2 is 4. Furthermore, the total travel distance for vehicles V1 and V2 is 11.

[0101] Next, Figure 10C shows an example of travel instructions calculated when customer P1 specifies time priority and customer P2 specifies price priority. The travel plan created in this case will vary depending on the evaluation weights set in the objective function. In the example of Figure 10C, however, we consider a case where the evaluation weight for distance traveled is sufficiently large. Such an objective function setting can be increased, for example, in situations where it is desirable to reduce travel distance as much as possible due to rising fuel costs. In the situation shown in Figure 10C, minimizing customer P1's waiting time and the total travel distance of vehicles V1 and V2 is prioritized, and a travel plan is created in which vehicle V2 transports customer P1 first, followed by customer P2. Vehicle V1 does not travel. In this case, customer P1's waiting time is 3, customer P2's waiting time is 5, and the total travel distance of vehicles V1 and V2 is 8. As a result, compared to the situation shown in Figure 10B, where customers P1 and P2 both specify time priority, the total travel distance of vehicles V1 and V2 is reduced by 3, while customer P2's waiting time increases by 1.

[0102] In this example, for customer P2, it is possible to offer a discount on the service price to compensate for the reduced fuel costs, in exchange for an increased waiting time, thereby benefiting both customers P1 and P2 and the service provider.

[0103] Finally, FIG. 10D shows an example of travel instructions calculated when customer P1 specifies price priority and customer P2 specifies time priority. In the example of FIG. 10D , as in the example of FIG. 10C , the case is considered in which the evaluation weight for mileage is sufficiently large. Under these conditions, priority is given to reducing customer P2's waiting time and the total mileage of vehicles V1 and V2, and a travel plan is created in which vehicle V1 transports customer P2 first, followed by customer P1. Vehicle V2 does not travel. In this case, customer P1's waiting time is 8, customer P2's waiting time is 4, and the total mileage of vehicles V1 and V2 is 9. As a result, compared to the situation in FIG. 10B where both customers P1 and P2 specify time priority, the total mileage of vehicles V1 and V2 is reduced by 2, while customer P1's waiting time increases by 5.

[0104] In this example, it is possible to offer a discount on the service price for customer P1, in exchange for an increase in waiting time, by refunding the reduced fuel costs. However, in this case, since the increase in waiting time for customer P1 is significant and the amount of fuel costs that can be refunded is not that great, it is also possible to ask customer P1 to decide whether such a schedule is acceptable. For example, when a travel plan is created and customer P1 is notified of the scheduled pick-up time, information on the scheduled pick-up time and estimated price for an alternative plan in which vehicle V2 transports the vehicle can be provided, and customer P1 can choose whether the current travel plan is acceptable or prefers the alternative plan. If the alternative plan is preferred, the travel plan can be changed.

[0105] - Flowchart - The processing procedure of the mobility management system A100 is shown in the flowchart of FIG.

[0106] First, in step S001, the mobility management system A100 updates various information such as reservation information, transport performance information, event information, traffic information, and current status information.

[0107] In step S002, the mobility management system A100 calculates, using the demand status prediction unit A002, predicted information on the number of customers at each node on the vehicle dispatch network based on the reservation information, transportation performance information, and event information acquired by the reservation information receiving unit A001.

[0108] In step S003, the mobility management system A100 calculates the required travel time for each edge on the vehicle dispatch network based on the traffic information using the travel time prediction unit A003.

[0109] In step S004, the mobility management system A100 calculates, using the arrival time prediction unit A004, predicted information on the number of vehicles that will be available at each node based on the travel time information calculated by the travel time prediction unit A003 and the current status information of each vehicle V.

[0110] In step S005, the mobility management system A100 sets the variables in the objective function and the corresponding evaluation weights in accordance with customer P's reservation information (dispatch options) based on the reservation information acquired by the reservation information receiving unit A001 using the objective function setting unit A005.

[0111] In step S006, the mobility management system A100 sets constraints in accordance with the reservation information (boarding requirements) of customer P using the constraint setting unit A006 based on the reservation information acquired by the reservation information receiving unit A001.

[0112] In step S007, the mobility management system A100 uses the parameters calculated in the demand status prediction unit A002, travel time prediction unit A003, and arrival time prediction unit A004, and takes into consideration the objective function and limiting conditions set in the objective function setting unit A005 and constraint condition setting unit A006, to find a solution to the optimization problem shown in Equation 9, and thereby calculates a mobility plan for each vehicle V.

[0113] In step S008, the movement management system A100 causes the movement instruction unit A008 to transmit a movement instruction to the corresponding vehicle V based on the movement plan calculated in step S007.

[0114] The mobility management system A100 repeatedly executes the above processing periodically at predetermined time intervals.

[0115] Effect: In the mobility management system A100 of this embodiment, the demand status prediction unit A002 predicts the number of future reservations based on customer P's reservation information, transportation performance information, etc., the travel time prediction unit A003 predicts the required travel time for each edge based on traffic information, and the arrival time prediction unit A004 calculates the arrival time of a moving vehicle V at its destination and the number of available vehicles for each time interval at each node. Next, the mobility plan creation unit A007 calculates the required time until transportation work can be provided in response to each reservation and the travel distance of the vehicle V that will fulfill the reservation based on the predicted reservation status for each future time interval and the number of available vehicles for each time interval at each node, and creates a mobility plan based on an objective function that uses at least one of the required time and travel distance as an evaluation index. Then, the mobility instruction unit A008 sends a mobility instruction to the corresponding vehicle V based on the mobility plan, and also sends the schedule to the customer P's mobile terminal M.

[0116] In a typical system in which travel instructions are issued each time a reservation is made, a nearby vacant vehicle is assigned to each reservation based on the current location of each vehicle V. However, according to this embodiment, as described above, travel instructions are sent after simulating the reservation status for each time interval up to a predetermined time period in the future and the location status of each vehicle V. For example, since travel instructions are output with the status after each vehicle V has been moved understood, for example, in a conventional system, for a reservation by customer P waiting at point X, the vacant vehicle Vx closest to point X at the time of the reservation would be assigned. However, if there is a vehicle Vy that moves closer to point X than the vacant vehicle Vx after handling another reservation, vehicle Vy may be assigned at the request of customer P. As in this example, according to this embodiment, a rational travel plan for each vehicle V can be created in response to future reservation status and fluctuations in the required travel time of the vehicle V on each route, resulting in various benefits such as improved service quality and reduced costs and environmental impact.

[0117] Furthermore, the objective function setting unit A005 appropriately sets the objective function taking into consideration the desired conditions specified by the customer P when making a reservation, specifically whether priority is given to time or price. In this way, the objective function set according to the desired conditions of the customer P is used to create the travel plan in the travel plan creation unit A007, which can contribute to improving customer satisfaction.

[0118] Furthermore, the constraint setting unit A006 sets constraints for each reservation as needed, taking into consideration the input items entered by the customer P at the time of reservation, specifically the boarding requirements related to information on the number of passengers and luggage. These constraints are reflected in the travel plan created by the travel plan creation unit A007, so that a vehicle V that meets the needs of each customer P can be appropriately and efficiently dispatched.

[0119] - Variations - Although the present embodiment shows an example of on-demand taxi dispatch management using reservation information, traffic information, and the like, there is other information that can be taken into consideration when creating a travel plan.

[0120] For example, when a customer P inputs reservation information, an environmentally friendly vehicle V, i.e., a vehicle V with low CO2 emissions per unit travel distance, can be selected or prioritized as an option or boarding requirement. In this case, the types of vehicles V that can be dispatched may be limited, which may cause the customer P to wait, but for the business operator, this reduces CO2 emissions and promotes its contribution to society, and the indirect benefit of this can be passed on to the customer P, which may also lead to a reduction in the price of the service provided.

[0121] Furthermore, when inputting reservation information, it is possible to select whether or not to allow carpooling, and to create a travel plan that takes carpooling into consideration. To do this, in the optimization problem for creating a travel plan, the capacity of vehicle V can be further considered, and the variable related to the number of vehicles can be divided according to the number of additional passengers that can be accommodated in each vehicle V. In this case, allowing carpooling can reduce the total travel distance of the managed vehicles V relative to the number of reservations. Furthermore, since the number of available vehicles increases, waiting times for customers P can also be reduced.

[0122] Furthermore, when calculating the required travel time for each edge on the vehicle dispatch network, travel times according to characteristics may be calculated taking into account not only traffic information but also the type of vehicle V and the skill of the driver. Driver skill can be evaluated, for example, by comparing the average travel times of each driver measured for the same edge during the same time period or under similar traffic volume conditions. Actual travel time data may be accumulated for each type of vehicle V and driver skill, and an approximate straight line may be calculated as shown in FIG. 6 and used to estimate travel times. Using such estimated values ​​can further improve the accuracy of the prediction model of Equation 2, enabling more realistic optimization of travel plans.

[0123] The mobility management system of this embodiment can also be easily applied to the operation management of food delivery services that transport food. In the case of food delivery, the starting point and destination of a customer P in an on-demand taxi are equivalent to the location of a restaurant and the delivery destination, respectively. In this case, the objective function of the optimization problem solved by the mobility plan creation unit A007 may be configured to include evaluation of the transportation time of the food. Furthermore, the evaluation weight may be adjusted depending on the type of food ordered. For example, customer P's satisfaction can be improved by adjusting the evaluation weight to be larger for food that is prone to temperature changes or spoilage during transportation.

[0124] <Second embodiment> A second embodiment of the present invention will be described. The second embodiment is an example of application of the movement management system A100 of the present invention to a warehouse transportation system. In this embodiment, the configuration of the movement management system A100 is the same as in the first embodiment, and descriptions of the same functions will be omitted as appropriate.

[0125] - Overview of Intra-Warehouse Transportation System - Figure 12 is a schematic diagram showing an example of an intra-warehouse transportation system to which this embodiment is applied. Generally, there are various types and uses of transport robots R operating in a warehouse, but here, it is assumed that the transport robot R is a transport machine that transports inventory shelves C in the warehouse. The inventory shelves C are equipped with casters and are towed and transported by the transport robot R. In the warehouse, an order (reservation) is received from a customer P, and the inventory shelves C containing the products specified in the order are transported by the transport robot R to a pick station S, where the products are removed (picked) from the inventory shelves C, packed into shipping boxes corresponding to the order, and shipped.

[0126] Generally, one order contains multiple products, so it is necessary to pick the products from multiple inventory shelves C. To perform this task efficiently, a transport robot R transports the inventory shelves C containing the products to be placed in the shipping box for each order to a pick station S. At the pick station S, a worker or a pick robot picks the necessary products from the inventory shelves C that the transport robot R has transported and packs them into the shipping box. Once the picking work is complete, the inventory shelves C are returned to their original positions by the transport robot R. By using the transport robot R in this way, it is possible to save labor by eliminating the need for workers to travel back and forth between the inventory shelves C and the pick station S when packing shipping boxes according to orders.

[0127] Generally, a plurality of transport robots R work in cooperation with one another, so there is a server that controls these transport robots R. The movement management system A100 of this embodiment is applied as the server of the transport robot R or as a function thereof.

[0128] The following describes the case where the movement management system A100 is applied to a warehouse transport system, focusing on the differences from the first embodiment.

[0129] First, the dispatch network of the transport robot R is represented by nodes (dots) and edges (lines) shown in FIG. 12. The nodes are located at potential destinations of the transport robot R or at branching points on the route. Potential destinations include each inventory shelf C and each pick station S, as well as a charging station (not shown). The edges are routes for movement between nodes. In FIG. 12, the nodes indicate the vicinity of the destination of the transport robot R, and do not indicate the locations where the transport robot R actually performs work. Furthermore, the linear shapes of the edges between the nodes shown in FIG. 12 do not represent the trajectory of the actual route, but merely indicate the connection relationships between the nodes.

[0130] In this embodiment, the reservation information acquired by the reservation information receiving unit A001 corresponds to task information for collecting products to be packed into shipping boxes for orders. FIG. 13 shows an example of a table listing task information as reservation information. Each task information includes, for example, "task number," "order ID," "stock shelf ID," "item ID," "pick station," "pick deadline time," and "priority." By referencing this task information, the target stock shelf C and pick station S can be identified, determining the starting and destination points of the transport task. The movement management system A100 can then issue a transport instruction including this data to each transport robot R. In accordance with the transport instruction from the movement management system A100, the transport robot R first heads to the node of the stock shelf C specified in the transport instruction, begins transporting the stock shelf C, moves to the node of the pick station S specified in the transport instruction, and waits until the worker has finished picking the target item from the stock shelf C. After the picking task is completed, the transport robot R moves from the node of the pick station S to the node of the original stock shelf C, returns the stock shelf C, and completes the transport of the stock shelf C.

[0131] In the first embodiment, the demand status forecasting unit A002 predicted future changes in reservations based on past performance information and event information, but in this embodiment, work is planned with a high degree of certainty for some time into the future, and it is not necessarily necessary to reflect transportation performance, etc. in the forecast. Therefore, here, it is possible to apply blanks to transportation performance information and event information, and obtain future forecast information for the number of tasks based essentially only on the task information received by the reservation information receiving unit A001.

[0132] In the first embodiment, the travel time prediction unit A003 calculated the travel time of each edge based on traffic information. However, in this embodiment, the warehouse is an indoor, well-maintained work environment with little external disturbance, so it is basically possible to calculate the travel time with high accuracy based on the actual length of each edge and the performance of the transport robot R. However, in some cases, transport robots R may approach each other and interact, causing a delay in one of the transport robots R. In anticipation of such a scenario, for edges that may overlap with a route that the transport robot R is already scheduled to travel, the travel time can be calculated taking into account the delay time due to the interaction. That is, in this embodiment, information on the planned travel route of the transport robot R corresponds to traffic information.

[0133] The objective function setting unit A005 sets an objective function used by the movement plan creation unit A007 to create a movement plan based on the work information managed by the reservation information receiving unit A001. In this embodiment, since the goal is to achieve high productivity in warehouse shipping operations, it is important to be able to pack shipping boxes for orders quickly overall. To achieve this, for example, the objective function is created using the arrival time at the pick station S as an evaluation index, based on the pick deadline time set for each order. Furthermore, because some orders require express shipping, priority is included in the work information to represent such targets, as shown in FIG. 13 . By utilizing this information, evaluation weights can be set so that the evaluation of work related to higher priority orders is higher if the pick deadline time can be brought forward (i.e., the evaluation value of the objective function is smaller).

[0134] The constraint setting unit A006 sets constraints based on the work information managed by the reservation information receiving unit A001. For example, if there is a pick station S where a pick robot performs work and the types of transport robots R that can cooperate with the pick robot are limited, constraints can be set so that a corresponding type of transport robot R can be assigned to the pick station S.

[0135] Based on the above information, the movement plan creation unit A007 creates a movement plan for the transport robot R. Although a detailed explanation of the optimization problem will be omitted, the optimization problem can be roughly executed in the same manner as in the first embodiment by utilizing Equation 9.

[0136] Finally, the movement instruction unit A008 transmits movement instructions to each transport robot R based on the movement plan created by the movement plan creation unit A007. At this time, it is also possible to transmit to the mobile terminal M of the customer P who has ordered the shipment of the product the shipping schedule of the product specified in the order or the estimated date and time when the product will arrive at the specified delivery destination.

[0137] -Effect- As described above, even when applied to a warehouse transport system, the movement management system A100 can give efficient movement instructions to each transport robot R, which is a transport machine, taking into account subsequent work information and predictions of the behavior of the transport robot R.

[0138] Although the embodiments of the present invention have been described in detail above using on-demand taxis, food delivery systems, and warehouse transport systems that travel on public roads as representative examples, the application of the present invention is not limited to these. For example, the mobility management system A100 of the present invention can also be applied to mobile or flying delivery robots, food delivery robots in restaurants, and the like.

[0139] Furthermore, the mobility management system A100 according to the present invention may have functions other than those of the blocks illustrated in FIG. 1 . The present invention is not limited to the above-described embodiments and may include various modifications. For example, the above-described embodiments have been described in detail to clearly explain the present invention, and the embodiments of the present invention are not necessarily limited to forms that include all of the described configurations. It is also possible to replace part of the configuration of one embodiment with the configuration of another embodiment, or to add the configuration of another embodiment to the configuration of one embodiment. It is also possible to add, delete, or replace part of the configuration of each embodiment.

[0140] A001... reservation information receiving unit, A002... demand status prediction unit, A003... travel time prediction unit, A004... arrival time prediction unit, A005... objective function setting unit, A006... constraint condition setting unit, A007... travel plan creation unit, A008... travel instruction unit, A100... travel management system, G1, G2... destination point, M... mobile terminal, P... customer, R... transfer robot (transportation machine), V... vehicle (transportation machine)

Claims

1. A mobility management system comprising: a reservation information receiving unit which receives reservation information relating to reservations for transportation; a movement plan creating unit which creates a movement plan for a plurality of transportation machines in accordance with the reservation information; and a movement instruction unit which outputs movement instructions to the plurality of transportation machines based on the movement plan; a demand status prediction unit which predicts the number of reservations for each time interval for a predetermined period from a current time for each of a plurality of points set in the transportation target area based on the reservation information, transportation performance information, and event information for the transportation target area; a movement time prediction unit which calculates the movement time required for the transportation machine to move between each of the plurality of points based on traffic information for the transportation target area; and an arrival time prediction unit which predicts the arrival time at which the transportation machine to which the movement instruction has been output will arrive at its destination point based on current status information of the plurality of transportation machines and the movement time predicted by the movement time prediction unit, and predicts the number of the transportation machines available at each point for each time interval; the movement plan creation unit predicts the time required to fulfill the reservation and the travel distance of the transport machine to fulfill the reservation based on the number of reservations for each time interval at each location predicted by the demand status prediction unit and the number of available transport machines for each time interval at each location predicted by the arrival time prediction unit, and creates the movement plan using an objective function that has at least one of the required time and the travel distance as an evaluation index; and the movement instruction unit outputs movement instructions to the multiple transport machines based on the movement plan, and notifies a customer's mobile terminal of the schedule.

2. A mobility management system as described in claim 1, characterized in that the mobility plan created by the mobility plan creation unit is a set of virtual mobility instructions including a departure point and destination point for each transport machine at the start time of each time section within a predetermined period, the predetermined period being a period starting from the current time consisting of N time sections, and a mobility route between the departure point and the destination point.

3. A mobility management system as described in claim 2, characterized in that the mobility instruction unit transmits to the transport machine only the mobility instruction calculated for the first time section among the virtual mobility instructions calculated for each time section within the specified period.

4. A mobility management system as described in claim 1, further comprising an objective function setting unit which sets the objective function by reflecting the customer's desired selection conditions contained in the reservation information, and the mobility plan creation unit solves an optimization problem for the objective function which reflects the desired selection conditions, searches for conditions which minimize the objective function, and creates the mobility plan.

5. A mobility management system according to claim 4, wherein the objective function includes a product of an evaluation index and an evaluation weight related to the desired selection condition.

6. A mobility management system according to claim 4, wherein the options for the desired selection conditions include time priority and price priority.

7. A mobility management system as described in claim 1, further comprising a constraint condition setting unit which sets constraint conditions based on customer input information included in the reservation information, and the mobility plan creation unit creates the mobility plan so as to satisfy the constraint conditions.

8. A mobility management system as described in claim 7, characterized in that the constraint condition setting unit sets the constraint condition to select a type of transportation vehicle corresponding to the number of passengers or baggage information included in the input items and assign it to the reservation, and the movement plan creation unit selects a type of transportation vehicle corresponding to the constraint condition and creates the movement plan.

9. A movement management method for creating a movement plan for a plurality of transport machines in accordance with reservation information related to transport reservations and outputting movement instructions to the plurality of transport machines based on the movement plan, comprising: predicting the number of reservations for each time interval for a predetermined period from the current time for each of a plurality of points set in the transport target area based on the reservation information, transport performance information, and event information for the transport target area; predicting the travel time required for the transport machines to move between each of the plurality of points based on traffic information for the transport target area; predicting the arrival time at which the transport machine to which the movement instruction has been output will arrive at its destination point based on current status information of the plurality of transport machines and the predicted travel time, predicting the number of the transport machines available at each point for each time interval; predicting the time required to respond to the reservations and the travel distance of the transport machines that respond to the reservations based on the predicted number of reservations for each time interval for each point and the predicted number of the transport machines available for each time interval for each point; creating the movement plan using an objective function that uses at least one of the required time and the travel distance as an evaluation index; A movement management method comprising: outputting movement instructions to the plurality of transport machines; and notifying a customer of the schedule to a mobile terminal.

Images