Warehouse work planning learning system, warehouse work planning learning method and program

Abstract

[Problem] The objective of the present invention is to be able to provide an optimal work plan for a warehouse, even if the work plan for the day is changed. [Solution] A warehouse work plan learning system 1 that generates a work plan for a warehouse is provided, wherein upon acquiring data regarding a change of work plan (such as insufficient workforce for the day), the system determines whether the changed work plan is feasible, and if determined to be unfeasible, the system analyzes the cause of the infeasibility and generates a plurality of recovery plans in which the workforce and / or work details are changed on the basis of the analyzed cause. The system then determines a recovery plan satisfying a predetermined condition from among the generated plurality of recovery plans, and if the work is completed according to plan as laid out in the determined recovery plan, the system learns the recovery plan in association with the analyzed cause or the work that led to the cause.

Description

Technical Field

[0001] The present invention relates to an effective technique for outputting an optimal work plan for warehouse operations.

Background Art

[0002] Conventionally, systems for optimizing warehouse operations have been proposed. For example, in Patent Document 1, a work plan system capable of formulating and optimizing an overall work plan including resource allocation for operations corresponding to a plurality of shipping orders is disclosed.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] However, even if a work plan for warehouse operations is established, on the same day, there may be workers who cannot come to work due to poor physical condition or the like, resulting in changes in the warehouse operation staff, or there may be shipping orders that require emergency response, leading to changes in the workload, etc., and the planned warehouse work plan may not be able to be carried out on the same day. In such a case, it is required to change the original work plan to respond to the shipping order. Also, it is required that the profit, work efficiency, or planned actual performance of the work plan in the warehouse operation on the same day does not deviate significantly from the original work plan. Therefore, the inventors of the present invention focused on a mechanism that, for changes in the work plan of warehouse operations, when it is expected that the warehouse operations will not proceed as originally planned and the warehouse operations will not be completed as they are, associates the cause factors for the occurrence of the cause of the incomplete warehouse operations with the work plan changed along with this factor, and utilizes the learning results for the same cases after the next time.

[0005] The present invention aims to provide a warehouse work plan learning system, a warehouse work plan learning method, and a program that enable the provision of an optimal work plan. [Means for solving the problem]

[0006] This invention relates to warehouses One day A warehouse work plan learning system that generates work plans, wherein data relating to changes in the work plans Data regarding the number of workers and / or the nature of the work. If obtained, after the change to the above A determination unit that determines whether the work plan is achievable, and if the determination unit determines that it is not achievable, The differences in the number of workers and / or work content extracted from the comparison between the work plan and the acquired data are analyzed for each work process. The aforementioned factors that make it impossible to achieve identification The analysis department, and the analysis performed by the said analysis department Identified Based on the factors, In the work plan A generation unit that generates multiple recovery plans with changed personnel and / or work content, and among the multiple recovery plans generated by the generation unit, Regarding profit margin and / or work efficiency, specified conditions satisfies , the aforementioned day A decision unit that determines the recovery plan, and the recovery plan determined by the decision unit The aforementioned day If the work is completed as planned, the recovery plan and the analysis unit will identification The present invention provides a warehouse work planning learning system comprising: a learning unit that learns by associating the factors that occurred or the work in which those factors occurred.

[0007] According to the present invention, if a change in the work plan makes it unlikely that the work plan in the warehouse will be achieved, the factors that prevented the work plan from being achieved are analyzed, and a recovery plan is generated. If the work is completed as planned using the recovery plan, the factors and the recovery plan are associated and learned. This means that, for example, if there is a change in the work plan before the start of work and it is unlikely that the work will be completed as planned, the factors that prevented the plan from being completed and the recovery plan at that time are associated and learned, so even if a similar plan change occurs in warehouse work after learning, it becomes possible to provide an optimal recovery plan.

[0008] Although this invention falls under the category of a system, similar effects and benefits can be achieved with methods and programs. [Effects of the Invention]

[0009] According to the present invention, even when the work plan for a warehouse is changed on the day, an optimal work plan can be provided. [Brief explanation of the drawing]

[0010] [Figure 1] This is a diagram illustrating the overview of Warehouse Work Planning Learning System 1. [Figure 2] This diagram shows the functional configuration of the warehouse work planning learning system 1. [Figure 3] This diagram shows a flowchart of the work plan acquisition process performed by the warehouse work plan learning system 1. [Figure 4] This diagram shows the flowchart of the shipping order acquisition process performed by the warehouse work planning learning system 1. [Figure 5] This diagram shows a flowchart of the shift data acquisition process performed by the warehouse work planning learning system 1. [Figure 6] This diagram shows the flowchart of the recovery plan generation process (personnel shortage) performed by the warehouse work planning learning system 1. [Figure 7] This diagram shows the flowchart of the recovery plan generation process (shipping order discrepancy) performed by the warehouse work planning learning system 1. [Figure 8] This diagram shows a flowchart of the recovery plan determination process performed by the warehouse work planning learning system 1. [Figure 9] This diagram schematically shows an example of the simulation results for multiple recovery plans. [Figure 10] This diagram shows a flowchart of the learning process performed by the warehouse work planning learning system 1. [Figure 11] This diagram shows a flowchart of the correction process performed by the warehouse work plan learning system 1.

Best Mode for Carrying Out the Invention

[0011] Hereinafter, with reference to the accompanying drawings, embodiments for carrying out the present invention (hereinafter referred to as embodiments) will be described in detail. In the following figures, the same elements are denoted by the same numbers or symbols throughout the description of the embodiments.

[0012] [Basic Concept / Basic Configuration] FIG. 1 is a diagram for explaining the outline of the warehouse work plan learning system 1. The warehouse work plan learning system 1 includes at least a computer 10. In the present embodiment, the warehouse work plan learning system 1 includes a computer 10, a WMS (Warehouse Management System) 20 for managing shipping orders, an RCS (Resource Control System) 30 for managing work plans, a shipper terminal 40 for transmitting shipping orders to the WMS, and a shift management system 50 which is an attendance management system for managing the employment status such as attendance, working hours, lateness, early departure, breaks, absenteeism, etc. for each worker by shift data, etc. These components are connected to each other such that these configurations are capable of data communication.

[0013] The work plan of the present embodiment is created in advance by the RCS 30, and the computer 10 acquires and stores this work plan at a predetermined timing. Here, it is assumed that this work plan may be changed due to the circumstances on the day before the work, for example, shortage of personnel or an increase in shipping orders.

[0014] An outline of the processing executed by the warehouse work plan learning system 1 will be described based on FIGS. 1 and 2.

[0015] The computer 10 acquires the work plan, shipping orders, and shift data for the day of warehouse work (step S1). This acquisition timing may be, for example, the timing when work starts on the day. A work plan consists of at least the work processes in warehouse operations, the number of workers assigned to each process, and the work content (e.g., work time, workload). Here, the processes in warehouse operations are assumed to be receiving, warehousing, picking, distribution processing, packing, and shipping, but the processes are not limited to these. A shipping order is data that includes, for example, a product ID, shipping quantity, and deadline date and time, and is data related to the collection of goods at the warehouse in question. Shift data shows the work schedule for each worker performing the task in question. WMS20 receives shipping orders from the shipper terminal 40. A shipping order is data used by a shipper to instruct warehouse personnel to ship goods, and includes the order number, store number, product number, product name, quantity, and order date. This shipping order is sent for each product. WMS20 also receives shipping orders from multiple shippers. WMS20 sends the shipping orders for the day from the received orders to the computer 10 (step S0). WMS20 calculates the shipping date by adding a predetermined number of days to the order date and determines the shipping orders for the day based on the shipping date. Note that WMS20 may also be configured to send shipping orders obtained from customer systems managed by customers such as shippers to the computer 10. RCS30 sends the work plan for the current day from among the work plans that have been created or saved to computer 10 (step S0). The shift management system 50 transmits the shift data for the current day to the computer 10 (step S0). As described above, computer 10 acquires shipping orders, work plans, and shift data from WMS20, RCS30, or shift management system 50, respectively.

[0016] The determination unit 11 of the computer 10 compares the acquired work plan with the planned shipping orders and shift data to check if there are any changes to the work plan. For example, on the day the work starts, the number of shipping orders may suddenly increase, causing the shipping volume indicated by the shipping orders to exceed the previously planned work volume, or on the day the work starts, there may be a shortage of workers, resulting in fewer workers indicated by the shift data. In such cases, changes to the plan may occur at the start of work. Therefore, if the shipping orders and shift data for the day indicate that a change will occur in the work plan that was created in advance, the computer 10 considers this data as data related to a change in the work plan and determines whether the work plan can still be achieved as planned even under these changed circumstances (step S2).

[0017] Next, if the analysis unit 12 of the computer 10 determines that the work plan cannot be achieved as planned, it analyzes the factors that make it impossible to achieve (step S3). For example, if the shift data shows that there is a shortage of workers for the packing process, the factor that makes it impossible to achieve is a shortage of workers. In this case, the analysis unit 12 compares the number of workers in the work plan with the number of workers in the shift data and calculates the number of workers that are needed. Furthermore, the analysis unit 12 identifies the relevant factor. For example, by referring to shift data, if it is predicted that there will be a shortage of workers for the packing work on the day, causing a delay in the cut time, the analysis unit 12 will compare this with the planned number of workers and identify the factor as a reason why it is impossible to achieve the plan as intended, as there are two fewer people for the packing work. In addition, it will identify the work that caused this factor as the packing work.

[0018] Next, the recovery plan generation unit 13 of the computer 10 generates multiple recovery plans that change the number of workers and / or the work content based on the factors analyzed by the analysis unit 12 (step S4). Here, a recovery plan is a recovery plan to achieve the work plan objectives. The recovery plan may also be generated by receiving input from the manager via the shipper terminal 40, or it may be generated based on a model learned by the warehouse work plan learning system 1. The work plan objectives mean that the number of shipping orders processed and the profit reach the planned values ​​at the time of planning.

[0019] The computer 10 then runs a simulation of each of the generated recovery plans and outputs them as shown in Figure 9. In response, the decision unit 14 of the computer 10 selects and decides on one recovery plan that satisfies predetermined conditions, such as the work being completed by the work deadline, based on input from the manager via the shipper terminal 40 (step S6).

[0020] Furthermore, if the work is completed as planned using the recovery plan determined by the decision unit 14, the learning unit 15 of the computer 10 learns by associating this determined recovery plan with the factors analyzed by the analysis unit 12 (step S7).

[0021] According to this warehouse work plan learning system 1, if there is a change in the work plan at the start of work and it is expected that the work will not be completed as planned, the system learns by associating the factors that will prevent the plan from being completed with a recovery plan. Therefore, even if there is a change in the work plan during subsequent warehouse operations, it will be possible to provide an optimal recovery plan.

[0022] [Functional Configuration] Based on Figure 2, the functional configuration of the warehouse work planning learning system 1 will be explained. The warehouse work planning learning system 1 includes at least a computer 10, which is connected to a WMS 20 that manages shipping orders, an RCS 30 that manages work plans, a shipper terminal 40 owned by an administrator that manages workers and work details, and a shift management system 50 that is an attendance management system that manages the work status of each worker using shift data, etc., via a network 9 such as a public telephone network or an intranet, enabling data communication. The warehouse work planning learning system 1 may include, in addition to the computer 10, a WMS 20, an RCS 30, a consignor terminal 40, a shift management system 50, and other terminals and devices. In this case, the warehouse work planning learning system 1 will perform the processing described later using one or more combinations of the included terminals, devices, and systems.

[0023] Computer 10 is a computer with server functionality or a personal computer, and it generates a trained model for outputting the optimal work plan for warehouse operations. Computer 10 may be implemented as a single computer, for example, or as a cloud computer, as it is implemented as multiple computers. In this specification, a cloud computer may be one that uses any computer scalably to perform a particular function, or one that includes multiple functional modules to implement a certain system and uses those functions in any combination.

[0024] Computer 10 includes a control unit consisting of a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), RAM (Random Access Memory), ROM (Read Only Memory), etc., and a communication unit consisting of devices that enable communication with other terminals and equipment. Furthermore, the computer 10 includes a data storage unit, such as a hard disk, semiconductor memory, storage medium, or memory card, as a memory unit. Furthermore, the computer 10 includes, as a processing unit, a determination unit 11, an analysis unit 12, a recovery plan generation unit 13, a decision unit 14 that determines the optimal recovery plan from a predetermined perspective, a learning unit 15, and the like.

[0025] WMS20 is a system that records the status of goods received, stored, and shipped to the warehouse. The recorded data includes product ID, quantity, and shipper's name. WMS20 also has a function to receive shipping instructions from shippers and warehouse managers. RCS30 records process information such as receiving, warehousing, picking, and distribution processing, as well as information on material handling equipment and workers assigned to these process details. It then retrieves shipping order information from WMS20 and creates a work plan. The consignor terminal 40 is a mobile terminal such as a mobile phone, smartphone, or tablet, or a personal computer, owned by the administrator. The terminal control unit includes a CPU, GPU, RAM, ROM, etc., the communication unit includes devices to enable communication with the computer 10, etc., and the input / output unit includes various devices to perform input and output of screens, data, etc. The shift management system 50 is a general-purpose system for managing the work status of workers. Specifically, it records information such as attendance and absence, and working hours when working, linked to identification information such as worker IDs. This information may be entered in advance by the worker or manager, or it may be automatically reflected in the shift management system 50 by reading an RFID tag given to the worker at the warehouse entrance gate.

[0026] [Work plan acquisition process] Based on Figure 3, the work plan acquisition process performed by computer 10 will be explained. This figure is a flowchart of the work plan acquisition process performed by computer 10. The work plan acquisition process is a detailed description of the flow related to the work plan acquisition process, which is part of the process (step S0) for acquiring the shipping order, work plan, and shift data mentioned above.

[0027] The work plan acquisition module implemented by computer 10 acquires the work plan for the current day from RCS30 (step S10). As mentioned above, the work plan includes the number of workers and the content of the work (e.g., working hours, workload) for each work process in warehouse operations, as well as the work process itself. RCS30 extracts the work plan for the current day from the pre-created work plans and sends it to computer 10. The work plan acquisition module acquires the work plan for the current day by receiving this work plan. Note that the source of the work plan is not limited to RCS30.

[0028] The work plan storage module implemented by computer 10 stores the acquired work plan (step S11). The work plan storage module stores the acquired work plan and associates it with its identifier (name, ID, management number, reference number, etc.).

[0029] The work plan acquisition process is performed as needed, such as each time the work plan is updated, at predetermined intervals, a predetermined time before the start of work, and after the end of work. The computer 10 uses the work plan acquired through the work plan acquisition process described above to execute the process described later. Furthermore, the computer 10 can be configured to execute the process described later without storing the work plan during the work plan acquisition process. In this case, the acquired work plan can be used as is in the process described later.

[0030] [Shipping Order Acquisition Process] Based on Figure 4, the shipping order acquisition process performed by computer 10 will be explained. This figure is a flowchart of the shipping order acquisition process performed by computer 10. The shipping order acquisition process is a detailed description of the flow related to the shipping order acquisition process, which is part of the process (step S0) for acquiring shipping orders, work plans, and shift data as described above.

[0031] The shipping order acquisition module implemented on computer 10 acquires shipping orders for the current day (step S20). As mentioned above, a shipping order is data that includes product ID, shipping quantity, deadline date and time, etc. The shipper terminal 40 receives the input of a shipping order from the shipper and transmits the received shipping order to the WMS 20. WMS20 receives this shipping order and stores it in its own memory unit. WMS20 then extracts the shipping orders for the current day from the stored orders and sends them to computer 10. The shipping order acquisition module acquires the shipping orders for the current day by receiving these shipping orders.

[0032] The shipping order storage module implemented by computer 10 stores the acquired shipping orders (step S21). The shipping order storage module stores the acquired shipping orders, associating them with their identifiers. These identifiers may include, for example, the product name, ID, management number, or reference number.

[0033] The shipping order acquisition process is performed as needed, such as each time a shipping order is updated, at predetermined intervals, or upon completion of a shipping order. Computer 10 uses the shipping orders acquired through the above-described shipping order acquisition process to execute the processes described later. Furthermore, the computer 10 can be configured to execute the processing described later without storing the shipping orders during the shipping order acquisition process. In this case, the acquired shipping orders can be used as they are in the processing described later.

[0034] [Shift data acquisition process] Based on Figure 5, the shift data acquisition process performed by computer 10 will be explained. This figure is a flowchart of the shift data acquisition process performed by computer 10. The shift data acquisition process is a detailed description of the flow related to the shift data acquisition process, which is part of the process (step S0) for acquiring the shipping order, work plan, and shift data mentioned above.

[0035] The shift data acquisition module implemented by computer 10 acquires the shift data for the current day (step S30). The shift management system 50 extracts the shift data for the current day and sends it to the computer 10. The shift data acquisition module obtains the shift data for the current day by receiving this shift data.

[0036] The shift data storage module implemented by computer 10 stores the acquired shift data (step S31). The shift data storage module stores the acquired shift data, associating it with an identifier for that shift data (name, ID, management number, reference number, etc.).

[0037] The shift data acquisition process is performed as needed, such as each time the shift data is updated, a predetermined time before the start of work, and at the end of work. The computer 10 uses the shift data acquired through the shift data acquisition process described above to perform the processes described later. Furthermore, the computer 10 can be configured to perform the processing described later without storing the shift data during the shift data acquisition process. In this case, the acquired shift data can be used as is in the processing described later.

[0038] [Recovery plan generation process (personnel shortage)] Based on Figure 6, the recovery plan generation process performed by computer 10 will be described. When the determination unit 11 compares the work plan and the shift data, if it determines that there is a shortage of workers and has acquired data related to the change in the work plan, it detects the shortage of personnel in the shift data (step S40). In other words, the determination unit 11 analyzes the shift data for the day and detects workers who have had changes such as absences, tardiness, or early departures as personnel shortages that prevent them from performing some or all of the warehouse work.

[0039] Next, the determination unit 11 determines whether the work plan is impossible to achieve due to the detected shortage of personnel (step S41). If the determination unit 11 determines that the work plan is impossible to achieve (step S41 YES), the analysis unit 12 identifies the cause of the impossibility as a shortage of workers (step S42). The determination unit 11 also identifies the work in which the shortage of workers occurred. Here, if the work plan is unattainable, an alert may be generated to notify that it is unattainable.

[0040] Next, the recovery plan generation unit 13 generates multiple recovery plans that change the number of workers and / or the work content according to the analyzed factors (step S43). For example, if there are two people short for packing work, it generates a recovery plan that adds two people, or it reschedules the shipping orders corresponding to the amount of work by the missing workers to the next day's work, extends the work time, or changes the work content of workers in other processes to packing work so that packing work can be completed even with two people short, etc., according to the manager's input. When rescheduling shipping orders to the next day's work, the recovery plan generation unit designates the shipping orders to be rescheduled in order from the one with the latest cut time. On the other hand, if computer 10 determines that the task is not impossible to accomplish (step S41 NO), it terminates the process.

[0041] [Recovery plan generation process (differences in shipping orders)] Next, the recovery plan generation process performed by computer 10 will be described based on Figure 7. The determination unit 11 compares the shipping orders planned up to the previous day with the shipping orders for the current day and detects any differences (step S50). If there are differences in the shipping orders, the determination unit 11 determines whether the revised work plan is achievable (step S51). If the determination determines that the work plan is unachievable, the analysis unit 12 identifies the difference in the shipping orders as the cause of the unachievability (step S52). For example, if the determination unit 11 compares the shipping orders from the previous day with the shipping orders for the current day and detects an order number in the current day's shipping orders that is not present in the previous day's shipping orders, it determines that there are differences. Next, for each shipping order for the current day, the determination unit 11 simulates each task and determines whether it can be completed successfully. Successful completion means completion within a predetermined time. More specifically, based on the work plan, the determination unit 11 maintains data indicating that three employees can be assigned to the packing work for the day, and that six seconds are required for the packing work of one shipping order. Then, by simulating the day's shipping orders, if it is detected that one of them will not be processed within the allotted time, the work plan is determined to be unattainable. Alternatively, if the total processing time for each task is found to be insufficient to complete within the allotted time, the work plan may be determined to be unattainable. Here, if the work plan is unattainable, an alert may be generated to notify that it is unattainable.

[0042] Next, the recovery plan generation unit 13 generates multiple recovery plans that change the number of workers and / or the work content according to the analyzed factors (step S53). For example, if it is detected that the packing work will not be completed within a predetermined time, it generates a recovery plan that reallocates workers from other tasks based on the excess time. Alternatively, if the aggregate value of the processing time for each task exceeds a predetermined time, it generates a recovery plan that adds workers to each task according to the excess time. It may also generate a recovery plan that changes some shipping orders to be shipped the next day based on the cut-off time. Furthermore, it may generate a recovery plan that extends the work time. These recovery plans generate multiple plans that change the number of workers, the original tasks from which the workers were reallocated, the shipping date, etc. In addition, the recovery plan generation unit 13 may generate a recovery plan that supplements the shortage of personnel in a task with personnel from other tasks, according to the analyzed factors and the task in which the factors occurred. On the other hand, if computer 10 determines that the task is not impossible to accomplish (step S51 NO), it terminates the process.

[0043] [Recovery plan determination process] Based on Figure 8, the recovery plan determination process performed by computer 10 will be explained. This figure is a flowchart of the recovery plan determination process performed by computer 10. The recovery plan determination process is the process of determining one recovery plan from the multiple recovery plans generated in the recovery plan generation process described above.

[0044] The simulation module implemented by computer 10 performs a simulation of each of the generated recovery plans (step S60). The simulation module performs a simulation of the recovery plan generated by the recovery plan generation process described above. The simulation performed by the simulation module involves the computer 10 simulating each work process in each recovery plan with the assigned personnel. The simulation module may simulate multiple recovery plans simultaneously.

[0045] Next, the profit calculation module implemented on computer 10 calculates the overall profit for the operation in the simulated recovery plan (step S61). The profit calculation module calculates revenue, expenses, and profit for each work process in the recovery plan. Here, revenue for each work process is the distribution of delivery fees paid by the shipper to each work process. Expenses are the labor costs of the personnel assigned to each work process. The profit calculation module calculates revenue, expenses, and profit for each work process, such as receiving, warehousing, picking, distribution processing, packaging, and shipping. The profit calculation module calculates the sum of revenue, expenses, and profit for each of these work processes. The profit calculation module divides the sum of profits by the sum of revenues and expresses the resulting value as a percentage to calculate the profit margin.

[0046] Next, the work efficiency calculation module implemented on computer 10 calculates the overall work efficiency for the day based on the recovery plan (step S62). Work efficiency is measured by one of the following values: total time required, cut-off time allowance, allowance rate, or excess man-hours. The work efficiency calculation module calculates the workload for each work process, such as receiving, warehousing, picking, distribution processing, packing, and shipping. The work efficiency calculation module calculates the total required time based on the time it would take to perform warehouse operations according to this recovery plan. Cut-off time is the deadline set for each shipping order. The work efficiency calculation module calculates the cut-off time by subtracting the current time from the truck departure time. The work efficiency calculation module calculates the cut-off time buffer by subtracting the total required time from the cut-off time. The work efficiency calculation module also calculates the buffer rate as the ratio of the cut-off time buffer to the cut-off time. A positive cut-off time buffer means there is a buffer in the work time, and a negative cut-off time buffer means there is not enough work time. The work efficiency calculation module calculates excess man-hours as the number of people required per unit time when the cut-off time buffer is negative. If the cut-off time buffer is positive, it is set to 0 man-hours.

[0047] In addition, the simulation result generation module implemented on computer 10 generates the simulation results (step S63). The simulation result generation module generates simulation results that summarize the profit margin and work efficiency for each recovery plan calculated by the processing in steps S61 to S63 described above. The simulation result generation module generates a summary of simulation content that includes the identifier of each recovery plan, the changes in the number of workers and / or work content in this recovery plan, the daily income, expenses, profit and profit margin (including income, expenses and profit in each work process) in this recovery plan, the calculated total required time for the day, cut-time slack time, slack rate, excess man-hours in this recovery plan, and the simulation content that summarizes the simulation content of each recovery plan, such as overall progress (work volume in each work process, work completion rate, etc.), processing capacity, and personnel allocation, visualized in graphs, tables, etc. (see Figure 9). Furthermore, the simulation results generated by the simulation result generation module are not limited to those described above; they may include other content, or any combination of any or more of these.

[0048] [Simulation Results] Based on Figure 9, the simulation results generated by the simulation result generation module will be explained. This figure schematically shows an example of the simulation results generated by the simulation result generation module. The simulation result generation module generates simulation result 60, which includes the name of each recovery plan, the changes made, and the overall work efficiency and profit margin for the day's work in that recovery plan. Note that the simulation result 60 generated by the simulation result generation module is not limited to what is shown in Figure 9; it may also include other content, or any combination of any or more of these.

[0049] Returning to Figure 8, we will continue to explain the recovery plan determination process. The decision unit 14 determines a recovery plan that satisfies predetermined conditions based on the simulation results, based on input from an administrator or the like (step S64). Here, the predetermined conditions may be, for example, that the cut-time buffer time is greater than a predetermined value, or that the total required time is within a predetermined time (within the time without overtime). Here, the decision unit 14 may make a decision using weight values ​​set by the weighting unit, rather than input from an administrator or the like, or it may set thresholds for profit margin and work efficiency in advance and decide on a recovery plan that meets those thresholds.

[0050] [Weighting of decisions] The weighting unit performs weighting according to the balance of predetermined viewpoints. Here, the predetermined viewpoints may be, for example, profit margin and work efficiency. In this embodiment, weighting is performed according to profit margin and work efficiency (excess man-hours). For example, the weighting unit performs weighting by multiplying profit margin and work efficiency (excess man-hours) by coefficients that sum to 10. Specifically, to maximize the weighting of profit margin, the ratio of profit margin is set to 10 and the other ratios are set to zero. Furthermore, the weighting unit may also weight profit margin and work efficiency separately. For example, if the weighting unit weights profit margin and work efficiency (overtime) at a ratio of 2:1, the coefficient for profit margin would be 2 and the coefficient for overtime would be 1. The weighting unit may also perform weighting in a similar manner when using perspectives other than those described above.

[0051] Computer 10 pre-sets these weighting values ​​in response to input from the administrator, and the decision unit 14 calculates an evaluation value based on the weighting values ​​to determine the recovery plan. The evaluation value is calculated as: profit rate of recovery plan 1 × weighting value / (profit rate of recovery plan 1 + profit rate of recovery plan 2 + profit rate of recovery plan 3) - excess man-hours of recovery plan 1 × weighting value / (excess man-hours of recovery plan 1 + excess man-hours of recovery plan 2 + excess man-hours of recovery plan 3). For example, in the example in Figure 9, the calculation is performed when profit rate:excess man-hours = 2:1. Note that a higher profit rate is desirable, but a lower excess man-hours is desirable, so the excess man-hours are treated as a negative value, and these are added together to determine the recovery plan with the largest value. [Recovery Plan 1] 8.8%×2 / (8.8%+13.2%+12.1%)-13 person hours×1 / (13 person hours+6 person hours+0 person hours)=-0.168 [Recovery Plan 2] 13.2%×2 / (8.8%+13.2%+12.1%)-6 people hours×1 / (13 people hours+6 people hours+0 people hours)=0.46 [Recovery Plan 3] 12.1%×2 / (8.8%+13.2%+12.1%)-0 person hours×1 / (13 persons hours+6 persons hours+0 persons hours)=0.71 The decision unit 14 then determines the recovery plan 3 that has the highest evaluation value based on the weighted values.

[0052] Next, the learning unit 15 associates the determined recovery plan with the factors that generated the recovery plan, or the factors with the work that caused those factors (step S65). The factors are, as mentioned above, for example, a shortage of workers due to an increase in shipping orders, or a shortage of workers due to an increase in shipping orders and packing work.

[0053] The learning unit 15 then executes the learning process shown in Figure 10. The learning unit 15 learns data relating the determined recovery plan to its factors, or the factors to the work that caused those factors (step S70). That is, the learning unit 15 uses the recovery plan associated with the factors (e.g., insufficient workforce or discrepancies in shipping orders), or the factors to the work that caused those factors (e.g., packaging work), as training data to perform machine learning and generate a learning model (step S71), and stores the generated trained model (step S72).

[0054] Here, if the learned factors, or the work that caused those factors and the same work plan changes occur again, the computer 10 uses this learned model to determine the associated recovery plan as the work plan for that day. Here, the associated factors may include alerts that occur when the work plan is unattainable.

[0055] [Recovery plan modification process] Based on Figure 11, the correction process performed by computer 10 will be explained. This figure is a flowchart of the correction process performed by computer 10. The correction process is the process of modifying the recovery plan determined in the recovery plan determination process described above.

[0056] The output module implemented by computer 10 outputs the determined recovery plan to a designated shipper terminal 40 or the like (step S80). The output module transmits the recovery plan determined by the recovery plan determination process described above to the shipper terminal 40, and the shipper terminal 40 outputs and displays the received recovery plan on its display unit.

[0057] Then, the correction reception unit implemented on computer 10 receives correction input from the administrator for the displayed recovery plan (step S81). A modification is, for example, a modification that changes the number of workers and / or the content of the work. For example, a modification that increases the number of workers for a given task by a predetermined number, or decreases the number of shipments for the day by a predetermined number. If there are any corrections, the shipper terminal 40 sends the corrections to the computer 10. The correction request module accepts the correction to the recovery plan upon receiving this correction.

[0058] The computer 10 then modifies the recovery plan with the entered modifications (step S82). The computer 10 may then run a simulation of the modified recovery plan, calculate the profit, calculate the work efficiency, and generate the simulation results. Then, the learning process described above is performed by associating this revised recovery plan with the factors that necessitated this recovery plan, or with those factors and the work that led to those factors. This allows the learning model to be generated using the recovery plan that has been fine-tuned by the administrator, rather than the recovery plan itself.

[0059] The means and functions of the computer 10 described above are realized when the computer 10 (including the CPU, information processing unit, and various terminals) reads and executes a predetermined program. The program may be provided, for example, in the form of a software as a service (SaaS) delivered via a network from the computer, or as a cloud service. Alternatively, the program may be provided in the form of a recording medium that can be read by the computer. In this case, the computer reads the program from the recording medium, transfers it to an internal or external recording device, records it, and executes it. Alternatively, the program may be pre-recorded on a recording device (recording medium) and provided to the computer from that recording device via a communication line.

[0060] Although embodiments of the present invention have been described above, the present invention is not limited to these embodiments. Furthermore, the effects described in the embodiments of the present invention are merely a list of the most preferred effects arising from the present invention, and the effects of the present invention are not limited to those described in the embodiments.

[0061] A first aspect disclosed in this embodiment is a warehouse work plan learning system 1 for generating work plans in a warehouse, comprising: a determination unit 11 that determines whether the modified work plan is achievable when it acquires data relating to a change in the work plan; an analysis unit 12 that analyzes the factors causing the unachievable work plan if the determination unit 11 determines it is unachievable; a recovery plan generation unit 13 that generates a plurality of recovery plans that change the number of workers and / or the work content based on the factors analyzed by the analysis unit 12; a decision unit 14 that determines a recovery plan that satisfies predetermined conditions from among the plurality of recovery plans generated by the recovery plan generation unit 13; and a learning unit 15 that learns the relationship between the recovery plan and the factors analyzed by the analysis unit 12 when the work is completed as planned using the recovery plan determined by the decision unit 14.

[0062] A second aspect disclosed in this embodiment further comprises a modification acceptance unit that accepts modifications to the determined recovery plan, The warehouse work plan learning system according to the first embodiment, wherein when the correction receiving unit receives a correction, the learning unit learns the corrected recovery plan and the factors in association.

[0063] A third aspect disclosed in this embodiment is a warehouse work planning learning system according to the first or second embodiment, wherein the predetermined condition is at least one of profit margin and work efficiency, and the determination unit determines a recovery plan that satisfies the condition to a preset value.

[0064] A fourth aspect disclosed in this embodiment is a warehouse work plan learning system according to the first embodiment, further comprising a weighting unit that sets weighting values ​​for each of the predetermined conditions, wherein the determination unit uses the weighting values ​​to determine one of the plurality of recovery plans. [Explanation of Symbols]

[0065] 1. Warehouse Work Planning Learning System 9 Network 10 Computers 20 WMS 30 RCS 40. Shipper terminal 50 Shift Management System 60 Simulation Results

Claims

1. A warehouse work plan learning system that generates a daily work plan for a warehouse, When data relating to the number of workers and / or the content of work is acquired as data related to the change in the work plan, a determination unit determines whether or not the work plan can be achieved after the change. If the determination unit determines that the achievement is impossible, the analysis unit analyzes the differences in the number of workers and / or work content extracted from the comparison of the work plan and the acquired data for each work process and identifies the factors that make the achievement impossible. Based on the factors identified by the analysis unit, a generation unit generates multiple recovery plans by modifying the number of workers and / or the work content in the work plan using the acquired data, A determination unit determines the daily recovery plan that satisfies predetermined conditions regarding profitability and / or work efficiency from among the multiple recovery plans generated by the generation unit, If the day's work is completed as planned according to the recovery plan determined by the decision unit, the learning unit learns by associating the recovery plan with the factors identified by the analysis unit or the work in which those factors occurred. A warehouse work planning learning system equipped with the following features.

2. The system further includes a modification acceptance unit that accepts modifications to the recovery plan determined by the decision unit, The warehouse work plan learning system according to claim 1, wherein when the correction receiving unit receives a correction, the learning unit learns the corrected recovery plan and the factors in association.

3. A simulation unit that performs simulations of each of the multiple recovery plans, The system further comprises a weighting unit that sets weighting values ​​for each predetermined viewpoint of the simulation results, The aforementioned predetermined viewpoint is at least one of the profit margin and / or the work efficiency, The warehouse work plan learning system according to claim 1, wherein the determination unit calculates an evaluation value for a plurality of recovery plans generated by the generation unit based on the predetermined viewpoint and the weighting value, and determines one recovery plan based on the calculated evaluation value.

4. The warehouse work planning learning system according to claim 3, wherein the weighting unit sets weight values ​​for each predetermined viewpoint in accordance with the balance of each predetermined viewpoint.

5. A warehouse work plan learning method in which a computer generates a daily work plan for a warehouse, When data relating to the number of workers and / or the content of work is obtained as data related to the change in the work plan, the step of determining whether the work plan is achievable after the change is made, If it is determined that the plan is unattainable, the differences in the number of workers and / or the content of the work extracted from the comparison between the work plan and the acquired data are analyzed for each work process to identify the factors that make the plan unattainable. Based on the factors identified through analysis, the steps include generating multiple recovery plans by modifying the number of workers and / or the work content in the work plan using the acquired data, The steps include determining the daily recovery plan from among the multiple recovery plans generated that satisfies predetermined conditions regarding profitability and / or work efficiency, If the day's work is completed as planned according to the determined recovery plan, the process involves learning by associating the recovery plan with the factors identified through analysis or the work in which those factors occurred. A warehouse work planning learning method that includes the following features.

6. A computer that generates a daily work plan in a warehouse, When data relating to the number of workers and / or the content of work is obtained as data related to the change in the work plan, the step of determining whether the work plan is achievable after the change is made, If it is determined that the achievement is impossible, the differences in the number of workers and / or the content of the work extracted from the comparison of the work plan and the acquired data are analyzed for each work process, and the factors causing the achievement to be impossible are identified. A step of generating multiple recovery plans by modifying the number of workers and / or the work content in the work plan using the acquired data, based on the factors identified through the analysis. A step of determining the daily recovery plan from among the multiple recovery plans generated that satisfies predetermined conditions regarding profitability and / or work efficiency, If the day's work is completed as planned according to the determined recovery plan, the next step is to learn by associating the recovery plan with the factors identified through analysis or the work in which those factors occurred. A computer-readable program for executing a command.

Images