Operation analysis device and operation analysis method

Abstract

To provide an operation analysis device and an operation analysis method, capable of executing highly accurate operation analysis even when achievement data has a mismatching part.SOLUTION: An operation analysis device according to the present invention, which uses a plurality of facilities to analyze an operation state of work to move baggage from a plurality of conveyance sources to a plurality of conveyance destinations, includes: an input unit that acquires information on the facilities, information on the baggage and achievement data of work executed by the facilities; a mismatching execution unit that generates a work list executed by each facility by using the information acquired by the input unit to extract a mismatching part in the work list on the basis of a predetermined rule; an achievement information correction unit that corrects the mismatching part of the work list for each facility, extracted by the mismatching extraction unit according to the predetermined rule; and an analysis unit that uses the work list for each facility, corrected by the achievement information correction unit to output information to assist in processing of identifying a delay cause for the work.SELECTED DRAWING: Figure 3

Description

[Technical field]

[0001] The present invention relates to an operation analysis device and an operation analysis method for analyzing the operational status of work in which cargo is moved from a plurality of transport sources to a plurality of transport destinations using a plurality of pieces of equipment. [Background technology]

[0002] In general, in the manufacturing industry, products manufactured in factories are packed, then moved to a storage area (product yard or warehouse) using a transport vehicle or a dolly such as an AGV, and stored in the storage area until the day of removal. On the day of removal, the products are removed from the storage area and loaded onto a transport vehicle or a transport ship by loading and unloading equipment, and transported to the customer's destination, a logistics center, etc. The movement of products within the factory until the day of removal is characterized by a shorter travel distance and travel time compared to the movement of products after removal. With regard to the movement of products after removal, the travel time to the customer's destination, a logistics center, etc. is much longer than the time required for other tasks (for example, loading and unloading onto a transport vehicle or a transport ship), so the route to be selected during the movement is a major factor in improving logistics efficiency and increasing logistics capacity, and logistics plans are often created with a focus on the method of selecting the travel route.

[0003] However, when moving products within a factory, since there is not much difference in the processing time of loading and unloading work and moving work of each piece of equipment, it is easy to encounter situations where it is difficult to determine which work is causing a delay in the entire logistics. Especially in large-scale factories, many and various pieces of equipment perform loading and unloading work in parallel at the same time, and each piece of equipment coordinates its work to move products. For this reason, for example, equipment may be put into a waiting state due to a difference in the timing of the end of each work, which may cause a chain reaction of delays in other work and lead to a large delay in the entire equipment. In order to prevent a decrease in work efficiency due to such delays, it is important to clarify the cause of the delay and take measures to prevent it. However, when many and various pieces of equipment are operating in a factory, it is often difficult to immediately find the root cause of the delay.

[0004] Here, Gantt charts are often used to visualize a work process that includes multiple steps. Specifically, Patent Document 1 proposes a technique for obtaining information on the impact of a plan change using a Gantt chart, and in the examples, a form is shown in which the work progress is displayed as a Gantt chart based on performance data. Since the transportation of goods is also a task performed by multiple machines in cooperation with each other, it can be considered as a work process that includes multiple steps. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] JP 2006-40211 A Summary of the Invention [Problem to be solved by the invention]

[0006] In baggage transportation, various devices and methods are used to collect operational results as data. To grasp the location of the baggage, systems that automatically recognize the arrival and departure of baggage are installed at storage areas, and terminals for operators to input the arrival and departure of baggage are installed, so that managers can grasp where each baggage is currently located. In general, tracking of baggage location is often performed relatively accurately, and the movement of baggage is stored as actual data. For this reason, by investigating the time when the baggage passed each storage area and the residence time at each storage area, it is expected that areas where the movement of baggage is slow can be found, the cause can be analyzed, and measures to address the delays can be considered.

[0007] However, when multiple and various types of equipment and machines work together to transport goods, each equipment and machine affects the progress of the work of the other equipment and machines, and it is difficult to fully understand the cause of the slow flow of goods by only observing the movement of the goods. In order to analyze the cause of the delay in detail, it is necessary to collect information on how each equipment and machine moved. However, the movement of each equipment and machine may not be collected with the same level of accuracy as the movement of goods. For example, even if the actual data on the movement of each equipment and machine loaded with goods is collected with the same accuracy as the movement of goods, the actual data on the movement of each equipment and machine not loaded with goods may be missing, or the time information saved in the data may be out of sync. Causes of missing or incomplete data include failure of sensors to detect the movement of equipment and machines, and failure or delay in inputting the movement of equipment and machines manually.

[0008] Since the movement of such equipment and machines that are not loaded with cargo can indirectly affect the delay in the movement of cargo, correcting such missing data and data deficiencies is an important issue in analyzing the cause of the delay. However, in the field of process control, unlike logistics operations, it is easy to deal with such missing data and data deficiencies. For example, when temperature measurements are taken continuously in an industrial process, even if one temperature measurement fails, the temperature can be estimated by calculating the average value from the values ​​at the previous and following times, and the process can be controlled, so there is no need to consider consistency. In contrast, the actual data of logistics operations must be consistent, satisfying various constraints such as the predecessor-successor relationship between tasks and capacity constraints, and corrections must be made taking this into consideration. For this reason, it is difficult to correct inconsistencies such as missing data and data deficiencies, and if there are inconsistencies in the actual data, it is not possible to perform highly accurate operation analysis.

[0009] The present invention has been made in consideration of the above-mentioned problems, and an object of the present invention is to provide an operation analysis device and an operation analysis method that are capable of performing highly accurate operation analysis even when inconsistencies are present in performance data. [Means for solving the problem]

[0010] The operation analysis device of the present invention is an operation analysis device that analyzes the operational status of work to move cargo from multiple sources to multiple destinations using multiple pieces of equipment, and includes an input unit that acquires information about the equipment, information about the cargo, and performance data of the work performed by the equipment, an inconsistency extraction unit that generates a work list performed by each piece of equipment using the information acquired by the input unit and extracts inconsistencies in the work list based on predetermined rules, a performance information correction unit that corrects the inconsistencies in the work list for each piece of equipment extracted by the inconsistency extraction unit in accordance with predetermined rules, and an analysis unit that outputs information to assist in the process of identifying the cause of a work delay using the work list for each piece of equipment corrected by the performance information correction unit.

[0011] When the inconsistency is due to missing data relating to time, the performance information correction unit may estimate the missing data based on a preset probability distribution.

[0012] The analysis unit may generate a delay chain path of an operation by using a precedence relationship between the operations, and output information about the generated delay chain path of the operation.

[0013] When the inconsistency portion is a missing piece of data related to time, the performance information correction unit may target a piece of equipment from which the inconsistency portion has been extracted, define the time at which the data is missing as time τ, and calculate, as a maximum likelihood estimate, the time τ at which a product of a first probability obtained from a probability distribution for a work time of a first task performed by the equipment up until immediately before the time τ and a second probability obtained from a probability distribution for a work time of a second task performed by the equipment from immediately after the time τ is maximized.

[0014] The inconsistency extraction unit may extract the inconsistency in the task list based on missing data in the task list and insufficient data regarding time.

[0015] If the inconsistent part is missing data regarding the location of the source or destination, the performance information correction unit may be executed immediately before or after the extracted work of the inconsistent part to extract other work that has common information regarding the transported cargo of the work, and estimate the data regarding the missing location based on the data regarding the location of the other extracted work.

[0016] The operations analysis method of the present invention is an operations analysis method for analyzing the operational status of work in which cargo is moved from multiple sources to multiple destinations using multiple pieces of equipment, and includes an input step of acquiring information about the equipment, information about the cargo, and performance data on the work performed by the equipment, an inconsistency extraction step of generating a list of work performed by each piece of equipment using the information acquired in the input step and extracting inconsistencies in the work list based on predetermined rules, a performance information correction step of correcting the inconsistencies in the work list for each piece of equipment extracted in the inconsistency extraction step in accordance with the predetermined rules, and an analysis step of outputting information to assist in the process of identifying the cause of a work delay using the work list for each piece of equipment corrected in the performance information correction step.

[0017] When the inconsistency is due to missing data relating to time, the performance information correcting step may estimate the missing data based on a preset probability distribution.

[0018] The analyzing step may generate a delay chain path of an operation using a precedence relationship between operations, and output information about the generated delay chain path of the operation.

[0019] When the inconsistency part is a missing data related to time, the performance information correction step may target the equipment from which the inconsistency part is extracted, define the time at which the data is missing as time τ, and calculate, as a maximum likelihood estimate, the time τ at which the product of a first probability obtained from a probability distribution for the work time of the work performed by the equipment immediately before the time τ and a second probability obtained from a probability distribution for the work time of the work performed by the equipment immediately after the time τ is maximized.

[0020] The inconsistency extraction step may extract the inconsistency in the task list based on missing data in the task list and insufficient data regarding time.

[0021] If the inconsistent part is missing data regarding the location of the source or destination, the performance information correction step may be executed immediately before or after the extracted work of the inconsistent part, to extract other work that has common information regarding the transported cargo of the work, and to estimate the data regarding the missing location based on the data regarding the location of the other extracted work. Effect of the Invention

[0022] According to the operation analysis device and the operation analysis method of the present invention, it is possible to perform highly accurate operation analysis even when there are inconsistencies in the performance data. [Brief description of the drawings]

[0023] [Figure 1] FIG. 1 is a schematic diagram showing an example of the configuration of a physical distribution facility that is the subject of an operation analysis apparatus and an operation analysis method according to an embodiment of the present invention. [Diagram 2] FIG. 2 is a diagram illustrating an example of a configuration of a transport vehicle. [Diagram 3] FIG. 3 is a block diagram showing the configuration of an operation analysis device according to one embodiment of the present invention. [Figure 4] FIG. 4 is a diagram illustrating an example of the statistical value information. [Diagram 5] FIG. 5 is a diagram illustrating an example of the statistical value information. [Figure 6] FIG. 6 is a flowchart showing the flow of an operation analysis process according to an embodiment of the present invention. [Figure 7] FIG. 7 is a diagram illustrating an example of the work performance information. [Figure 8] FIG. 8 is a diagram showing an example of a task list including missing data portions. [Figure 9] FIG. 9 is a diagram showing an example of a modified task list. [Figure 10] FIG. 10 is a diagram showing an example of a task list including a data deficiency portion. [Figure 11] FIG. 11 is a diagram showing an example of a modified task list. [Figure 12] FIG. 12 is a diagram showing an example of a Gantt chart display. [Figure 13] FIG. 13 is a diagram showing an example of displaying a delayed chain path. [Figure 14] FIG. 14 is a diagram illustrating an example of precedence relationship information. [Figure 15] FIG. 15 is a flow chart showing the flow of the missing value extraction logic. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0024] An operation analysis apparatus and an operation analysis method according to an embodiment of the present invention will be described below with reference to the drawings.

[0025] [Configuration of logistics facilities] First, with reference to FIG. 1, a configuration example of a physical distribution facility that is a target of an operation analysis apparatus and an operation analysis method according to an embodiment of the present invention will be described.

[0026] FIG. 1 is a schematic diagram showing an example of the configuration of a logistics facility that is the target of an operation analysis device and an operation analysis method according to an embodiment of the present invention. As shown in FIG. 1, the cargo handled by the logistics facility 1 that is the target of the operation analysis device and the operation analysis method according to an embodiment of the present invention is coils C, which are steel products, and the cargo moving equipment is roughly classified into loading and unloading equipment and transport equipment. The loading and unloading equipment is an overhead crane 2 and a ship crane 3, and the transport equipment is a transport vehicle 4. In this logistics facility 1, the coils C manufactured in the factory 5 are carried into the warehouse 6 by the transport vehicle 4. The warehouse 6 plays the role of a buffer for temporarily storing the coils C until shipment, and is arranged along the shipping berth where the transport ship is moored. The warehouse 6 is configured in a form in which a plurality of buildings (buildings) 6a are adjacent to each other, and each building is provided with an overhead crane 2 and a vehicle entrance 7 for the transport vehicle 4 to enter. The overhead crane 2 is used to move the coils C carried by the transport vehicle 4 into the building, and to load the coils C that are close to being shipped onto the transport vehicle 4. When the coil C is close to being shipped, the coil C is moved by the transport vehicle 4 to the shipping berth where the transport ship is anchored.

[0027] 2(a) and (b) are diagrams showing an example of the configuration of the transport vehicle 4. As shown in FIG. 2(a) and (b), the transport vehicle 4 has a structure in which the towing unit V and the pallet (loading unit) P can be separated. The towing unit V loads the pallet P on which the coils C are placed at the factory 5, and transports the pallet P to the designated warehouse. When the towing unit V arrives at the warehouse 6, it enters the vehicle entrance 7 and lowers the pallet P into the vehicle entrance 7. After lowering the pallet P, the towing unit V loads another pallet P placed at the same vehicle entrance 7, or exits the vehicle entrance 7 and moves to the vehicle entrance 7 where the other pallet P is placed in order to transport the other pallet P. The coils C placed on the pallet P placed in the warehouse 6 are moved into the warehouse 6 by the ceiling crane 2 (this is called warehousing). When all the coils C on the pallet P have been moved into the warehouse 6, the towing unit V carries the empty pallet P and carries it out through the vehicle entrance 7. This towing unit V may be different from the towing unit V that carried the pallet P to the vehicle entrance 7.

[0028] In this way, since the transport vehicle 4 has a structure in which the towing unit V and the pallet P can be separated, the moving operation (traveling) of the towing unit V and the loading operation on the pallet P can be performed in parallel. In a vehicle in which the towing unit V and the pallet P cannot be separated, such as a general truck, the vehicle must be stopped during loading and unloading operations, but a vehicle in which the towing unit V and the pallet P can be separated can transport more efficiently than such a vehicle. On the other hand, the transport operation of a vehicle in which the towing unit V and the pallet P can be separated is also characterized by being easily affected by the operation of other transport vehicles. This is because the same pallet P is used by multiple towing units V, and for example, the towing unit V1 transports the pallet P1 carrying a certain product from a factory to a warehouse 6, and then the towing unit V2 transports the pallet P1, which has finished unloading the product at the warehouse 6, to the factory to have another product loaded on it. In this case, if the work of the towing unit V1 is delayed and the transport of the pallet P1 is delayed, the delay will cause a chain reaction of loading and unloading operations at the warehouse and subsequent delays in the work of the towing unit V2. In addition, since multiple towing units V and pallets P enter the vehicle entrance 7 of the same factory 5 or warehouse 6, the logistics facility 1 shown in Figure 1 operates in a way that makes it easy for work delays at each facility to affect the others. In this type of logistics facility, if a certain facility causes a work delay, it is difficult to know the root cause because the delays affect each other in a chain reaction. For this reason, by supporting the analysis of the operation status with an operation analysis device that visualizes the operation status, it becomes easy or possible to identify the cause, and it becomes easier to take measures to improve the efficiency of operations.

[0029] [Configuration of the operation analysis device] Next, the configuration of an operation analysis device according to one embodiment of the present invention will be described with reference to FIG.

[0030] Fig. 3 is a block diagram showing the configuration of an operation analysis device according to one embodiment of the present invention. As shown in Fig. 3, an operation analysis device 10 according to one embodiment of the present invention is configured by an information processing device such as a personal computer or a workstation, and includes an input unit 11, an inconsistency extraction unit 12, a performance information correction unit 13, an analysis unit 14, a display unit 15, a user setting unit 16, and an operation information database (operation information DB) 17.

[0031] The input unit 11, the inconsistency extraction unit 12, the performance information correction unit 13, and the analysis unit 14 are functional blocks that are realized by the arithmetic processing unit in the information processing device executing a computer program. The functions of each functional block will be described later.

[0032] The display unit 15 is composed of a display device such as a liquid crystal display, and displays and outputs various information in accordance with control signals from the analysis unit 14. The user setting unit 16 is composed of input devices such as a keyboard, a mouse pointer, a touch panel, and the like, and inputs user operation input information to the analysis unit 14.

[0033] The operation information DB 17 is configured by a non-volatile storage device, and stores various information necessary for the operation of the operation analysis device 10. In this embodiment, the operation information DB 17 includes an equipment information storage unit 17a, a work performance information storage unit 17b, and a transported baggage information storage unit 17c.

[0034] The equipment information storage unit 17a stores information on a storage location for storing the coil C and on loading and unloading equipment for transferring the coil C. Specifically, the equipment information storage unit 17a stores the following information.

[0035] Equipment ID: Identifier information assigned to each piece of equipment Equipment type: Equipment type information (such as storage area, overhead crane, loading crane, and type of transport vehicle) - Storage area information (for cranes, a list of storage areas and entrances for loading and unloading, for transport vehicles, a list of entrances that can be used for loading and unloading) - Available pallet types (pallet types that can be brought into the entrance of a storage facility, pallet types that can be transported by a transport vehicle) Statistical information (distribution and average value of waiting time at the storage facility, distribution and average value of travel time from the source to the destination for transport vehicles, etc.)

[0036] 4 and 5 show an example of the statistical value information stored in the facility information storage unit 17a. FIG. 4 shows information summarizing the statistical values ​​of the travel time of the transport vehicle 4. The facility ID is an identifier given to the target transport vehicle, and the statistical value information shows the distribution of travel time when the target transport vehicle travels from the departure ID to the arrival ID. The item Ave. is the average travel time (minutes), and is 15.0 (minutes) in the first record. The values ​​stored in 5% increments of 5 to 100% indicate what percentage of the total number of results is the number of results in which the "travel time" is this time or less. For example, in the first record, the value of the 10% item is 13.0 (minutes), which means that the number of results in which the travel time was 13.0 (minutes) or less was 10% of the total number of results. The value of the 100% item is 20.0 (minutes), which means that the number of results in which the travel time was 20.0 (minutes) or less is the same as the total number of results, and the longest travel time is stored in this item. Similarly, Fig. 5 shows the distribution of waiting times for the transport vehicle 4 at each storage yard (frontage). When the transport vehicle 4 goes to each storage yard, there are cases where the loading of the coils C onto the pallet P or the unloading of the coils C from the pallet P is not completed and the transport vehicle 4 has to wait. An example of the distribution of such waiting times is shown in Fig. 5.

[0037] Returning to Fig. 3, the work result information storage unit 17b stores the result information of the work performed by each piece of equipment. Specifically, the work result information storage unit 17b stores the following information:

[0038] Equipment ID: Identifier information assigned to each piece of equipment - Achievement details: Details of the actions taken by each piece of equipment (loading pallets, starting travel, waiting, etc.) Location: The location where each facility performed the action. Time: The time when each facility performed an action Pallet ID: Pallet ID of each equipment action target Lot ID: Transport lot identifier of coil C loaded on the pallet targeted for the action of each facility

[0039] The transported luggage information storage unit 17c stores information on the transported coil C. Specifically, the transported luggage information storage unit 17c stores the following information.

[0040] Product ID: Identifier given to coil C Lot ID: Transport lot identifier assigned to coil C Product size: weight, width, outer diameter, etc. · Variety: hot rolled, pickled, cold rolled, galvanized, etc. - Passing process list: Storage location information (storage location ID, entry time, exit time)

[0041] The information stored in the operation information DB 17 is updated when work is completed at each facility. When the transport vehicle 4 starts or finishes its transport work or passes a set point, it automatically or manually transmits information to the operation information DB 17 from a communication device installed on the transport vehicle 4. The operation information DB 17 updates each information storage unit according to the transmitted information. Similar communication devices are also installed on the overhead crane 2 and the loading crane 3, and transmit information to the operation information DB 17 when the crane starts or finishes moving, or when the coil C is loaded or unloaded.

[0042] [Operation analysis processing] Next, the flow of the operation analysis process executed by the operation analysis device 10 will be described with reference to FIG.

[0043] Fig. 6 is a flowchart showing the flow of an operation analysis process according to one embodiment of the present invention. The flowchart shown in Fig. 6 starts when an execution command for the operation analysis process is input to the operation analysis device 10, and the operation analysis process proceeds to step S1.

[0044] In the process of step S1, the input unit 11 extracts operation information of the logistics facility 1 required for the operation analysis process from the operation information DB 17. This completes the process of step S1, and the operation analysis process proceeds to the process of step S2.

[0045] In the process of step S2, the inconsistency extraction unit 12 generates a work list for each facility based on the work performance information included in the operation information of the logistics facility 1 extracted in the process of step S1, and extracts inconsistencies such as missing data and incomplete data included in the generated work list. An example of work performance information is shown in FIG. 7. The work performance information shown in FIG. 7 indicates in chronological order what work the transport vehicle (towing unit) VH0001 performed. FIG. 8 shows a work list for the transport vehicle VH0001 created by the inconsistency extraction unit 12 based on this work performance information. In the work list shown in FIG. 8, the transport vehicle VH0001 departs from the source (transport source) ("Departure" in FIG. 7) and arrives at the next destination (transport destination) ("Arrival" in FIG. 7) in the work performance information shown in FIG. 7 as one record of transport work. In addition, if waiting occurs for some reason before arriving at the destination ("Waiting" in FIG. 7), the transport work until the waiting occurs is one record. The work list shown in FIG. 8 is composed of the travel route (information about location) of the transport vehicle VH0001, the start and end times of travel (data about time), and the transported pallets and lots (information about cargo). In the example shown in FIG. 8, the data of the destination and end time of travel in the second record of the work list is missing (destination and end time are blank). The reason for this missing data is as follows. That is, the seventh record of the work performance information (FIG. 7) stores the result of the transport vehicle VH0001 departing from the yard G03 (origin) with pallet A00003 that does not carry coils (lot ID is blank). However, the eighth record of the work performance information stores the result of arriving at the yard G01 (destination) without loading a pallet (pallet ID and lot ID are blank), and there is an inconsistency between the information (pallet ID) about the cargo of the source and destination. For this reason, the eighth record cannot be identified as the work performance information of the destination that is paired with the seventh record, and the data of the destination and end time are missing.

[0046] The cause of this is considered to be that after the transport vehicle VH0001 departs from the storage site G03 with the pallet A00003 without coils loaded (the result content of the seventh record in FIG. 7), the pallet A00003 is actually unloaded at another storage site and then departs for the storage site G01 (the location of the eighth record in FIG. 7), but the result input of the arrival at the storage site where the pallet A00003 was unloaded, the completion of unloading the pallet, and the departure to the storage site G01 is omitted. In particular, when the result input is performed manually, such input omissions are likely to occur frequently. Even when the input is performed automatically, a similar input omission may occur if communication does not work well or if the current position of the transport vehicle 4 is not recognized well. In this way, the inconsistency extraction unit 12 generates data on the transport work from the movement origin to the movement destination of each transport vehicle 4 from the work result information for each transport vehicle 4 (FIG. 7), and creates a work list (FIG. 8) from the generated data, thereby extracting the missing data as an inconsistency. Furthermore, even if there is no missing data in the task list, if the actual input of "waiting" is missing in the task performance information, a data deficiency occurs in the information regarding time, and an inconsistency occurs, such as the time required from the start to the end of the transportation task (travel time) exceeding a threshold value determined based on the statistical value information. For this reason, the inconsistency extraction unit 12 also extracts such data deficiencies in the information regarding time as inconsistencies. This completes the processing of step S2, and the operation analysis processing proceeds to the processing of step S3.

[0047] In the process of step S3, the performance information correction unit 13 corrects the inconsistent part of the work list of each facility extracted in the process of step S2 according to the missing value extraction logic created in advance according to the type of missing value (missing data, incomplete data, etc.). First, a method of correcting the inconsistent part of the missing data will be described. FIG. 15 is a flowchart showing the flow of the extraction logic of the missing data value. As shown in FIG. 15, in order to correct the inconsistent part of the missing data, the performance information correction unit 13 estimates missing information about the location of the travel route (origin or destination) (step S31) and estimates missing data about time (start time or end time) (step S32). The contents of steps S31 and S32 will be specifically described below.

[0048] <Estimation of missing information regarding locations on the travel route (Step S31)> To correct the missing data in the second record of the work list in Fig. 8, it is necessary to estimate which location (destination) pallet A00003 was transported to and the end time of the trip. Therefore, when an inconsistent part of the work list is due to missing data related to the source or destination location, performance information correction unit 13 extracts other work that is executed immediately before or after the extracted work of the inconsistent part and that shares information (pallet ID, lot ID) related to the transported cargo of that work. Then, performance information correction unit 13 identifies the missing data related to the location (source or destination) based on the data related to the location of the other extracted work (source or destination).

[0049] Specifically, as shown in FIG. 15, the performance information correction unit 13 performs the steps of [1] extracting the trace target (step S31a), [2] extracting the storage location information of the trace target (step S31b), and [3] estimating the travel route of the transport vehicle (step S31c) to identify the storage location information of the destination. Step S31a is a step of extracting, as a trace target, transported goods information (pallet ID or lot ID) having an inconsistency in the movement route for the work before and after the inconsistent part of the work list of each facility. The processing of step S31a will be specifically described using the work list shown in FIG. 8. The performance information correction unit 13 extracts pallet A00003, which is the transported goods information (pallet ID) of the second record having missing data, as a candidate for the trace target. Then, the performance information correction unit 13 checks whether there is an inconsistency in the movement route of pallet A00003 for the work before and after the inconsistent part of the work list. 8, after pallet A00003 is transported out from yard G03 (origin) at start time 2022 / 12 / 1 7:20, there is no information about pallet A00003 and there is an inconsistency in the movement path of pallet A00003 from when it is transported out again from yard G02 (origin) at 2022 / 12 / 1 9:00. Therefore, performance information correction unit 13 extracts pallet A00003 as a tracing target.

[0050] Step S31b is executed immediately before or after the extracted work of the inconsistent part of the work list, and is a step of extracting other work from the work list that has the same transported luggage information (pallet ID or lot ID) as the tracing target extracted in step S31a, and extracting data on the location of the other extracted work (origin or destination) as the storage location information of the tracing target. Specifically, the performance information correction unit 13 searches the records of the work list, including records of transport vehicles other than transport vehicle VH0001, and extracts records related to pallet A00003 (information on transported luggage) as the tracing target at the earliest time after the start time of 7:20 on December 1, 2022 (no missing data related to time). In the work list shown in Figure 8, there is a record (the fifth line of the work list in Figure 8) for moving pallet A00003 from location G02 (the source) at 9:00 on December 1st, 2022. If this is considered to be the record related to pallet A00003 that occurs earliest after 7:20 on December 1st, 2022, the performance information correction unit 13 extracts location G02, which is the source of this work, as the location information to be traced.

[0051] Step S31c is a step of generating a travel route of the transport vehicle that resolves the inconsistency in the travel route of the transported luggage information (pallet ID or lot ID) to be traced based on the storage yard information to be traced extracted in step S31b, and identifying the data (origin or destination) related to the missing location. Specifically, the performance information correction unit 13 estimates the travel route of the transport vehicle VH0001 from storage yard G03 (origin of the second record in FIG. 8) to storage yard G01 (origin of the third record in FIG. 8) via storage yard G02 extracted in step S31b during the time from 2022 / 12 / 1 7:20 to 2022 / 12 / 1 7:45). If the transport vehicle VH0001 transfers pallet A0003 to another pallet (pallet A00002) at the storage site G02, the inconsistency in the movement path of pallet A00003 in step S31a is resolved. Therefore, the performance information correction unit 13 identifies the storage site G02 as the missing location-related data (movement destination).

[0052] In the above example, only one piece of trace target storage location information is extracted in step S31b, but if multiple pieces of trace target storage location information are extracted in step S31b, the performance information correction unit 13 generates multiple travel routes for the transport vehicle in which the order of each extracted storage location is changed in various ways, and selects a travel route that eliminates the inconsistency in the travel route of the transported goods to be traced. As a result, the corrected work list becomes as shown in FIG. 9(a). In addition, if the missing information on the location of the travel route is the "origin of movement," a record related to information on the transported goods such as a pallet ID or lot ID at the closest time (latest time) before the time-free data (end time) may be extracted, and the destination of the extracted record may be estimated as the origin of movement, which is the missing information on the location of the travel route. In addition, the above example was a case where information regarding transported cargo was present among records containing inconsistent portions of missing data, but when the transport vehicle VH0001 travels without carrying any cargo, it is possible to extract from the records of the transport vehicle VH0001 the records around the time closest to the missing data regarding time (start time or end time), and infer that the origin or destination of the extracted record is missing information regarding the location on the travel route.

[0053] <Estimation of missing data regarding time (Step S32)> According to step S31, the transport vehicle VH0001 moves from the yard G03 to the yard G02 estimated in step S31 to the yard G01 during the period from 7:20 on 12 / 1 / 2022 to 7:45 on 12 / 1 / 2022. Therefore, in step S32, the performance information correction unit 13 calculates the arrival time τ that maximizes the value of the following formula (1), with the arrival time at the yard G02 (departure time from the yard G02) being τ. Thus, in step S32, the performance information correction unit 13 uses statistical data on the travel time and waiting time to associate the missing data with the work before and after the missing data regarding time, and probabilistically estimates the arrival time τ at the yard G02 as missing data regarding time. That is, the performance information correction unit 13 calculates the time τ at which the product of the probability (first probability) obtained from the probability distribution of the work time of one work (first work) performed by the transport vehicle VH0001 until just before the time τ and the probability (second probability) obtained from the probability distribution of the work time of another work (second work) performed by the transport vehicle VH0001 immediately after the time τ is maximized as the maximum likelihood estimate of the arrival time τ. Here, the first work performed by the transport vehicle VH0001 until just before the time τ is the work of transporting pallet A00003 from yard G03 (starting point ID) to yard G02 (arrival point ID). Also, the second work performed by the transport vehicle VH0001 immediately after the time τ is the work of moving from yard G02 (starting point ID) to yard G01 (arrival point ID) without loading a pallet. The probability distribution of the first task and the second task can be calculated based on the approximated probability density function by approximating a probability density function based on the statistical value information of each task, which is a compilation of the statistical values ​​of the travel time of the transport vehicle VH0001 shown in FIG. 4. The task times of the first task and the second task performed immediately before and after time τ are linked to time τ, and the first probability and the second probability are also linked to time τ. The likelihood of time τ can be calculated by the product of the first probability and the second probability, and the time τ at which the product of the first probability and the second probability is maximum can be identified as the maximum likelihood estimate of the arrival time τ. This completes the process of step S3, and the operation analysis process proceeds to step S4.

[0054]

number

[0055] Here, t G01 is the time when the transport vehicle VH0001 arrived at the storage site G01 (7:45 a.m. on December 1, 2022), and t G03 is the time when the transport vehicle VH0001 departs from the storage site G03 (2022 / 12 / 1 7:20), P(VH0001, G02, G03, τ-t G03 ) is the time taken for the transport vehicle VH0001 to move between the storage sites G02 and G03, and the travel time is τ-t G03 -T~τ-t G03 +T (T is a fixed parameter) interval, P(VH0001,G01,G02,t G01 -τ) is the travel time of the transport vehicle VH0001 when it moves between the storage sites G01 and G02. G01 -τ-T~t G01 It indicates the probability that the probability falls within the interval -τ+T (T is a fixed parameter). P(*) is a function that approximates the probability density function based on the statistical value information shown in Figure 4 and returns a probability value using that function. The maximum likelihood estimate is calculated by calculating the time τ at which formula (1) is maximized. In this example, it shows the case where the maximum value of formula (1) is maximized at time τ = 7:30 on December 1, 2022, and the task list revised accordingly will be as shown in Figure 9(b).

[0056] Next, an example of a task list including inconsistent data regarding time is shown. The task list shown in FIG. 10 is consistent with no missing information, but the time required from the start to the end of the fourth record is very long, and this is detected as a data defect. The detection of the data defect itself can be determined based on a time threshold determined based on statistical information (travel time distribution). The cause of the long time is considered to be that the transport vehicle VH0005 waited (in a waiting state) until the coil C was placed on the pallet K00003 at the destination Q01. The inconsistent data regarding time is considered to have occurred due to a lack of actual value input in the waiting state. For a task list including inconsistent data regarding time, the waiting start time τ can be estimated by the same method as step S32 shown in FIG. 15. As a method of calculating the waiting start time, similar to the method of maximizing the time τ shown in the above-mentioned formula (1), the time τ when the transport vehicle VH0005 arrives at the destination Q01 is set as τ, and the time τ that maximizes the following formula (2) is calculated. That is, the time τ at which the product of the probability (first probability) obtained from the probability distribution of the work time of one work (first work) performed by the transport vehicle VH0005 until just before the time τ and the probability (second probability) obtained from the probability distribution of the work time of another work (second work) performed by the transport vehicle VH0005 immediately after the time τ is maximized is calculated as the maximum likelihood estimate of the waiting start time τ. Here, the first work performed by the transport vehicle VH0005 until just before the time τ is the work of transporting the pallet K00001 from the yard G04 (starting point ID) to the yard G01 (arrival point ID). Also, the second work performed by the transport vehicle VH0005 immediately after the time τ is the work of waiting at the yard G01 (opening). The probability distribution of the former work can be calculated based on the approximated probability density function by approximating a probability density function based on statistical value information summarizing the statistical values ​​of the travel time of the transport vehicle VH0005 shown in FIG. 4. The probability distribution of the latter operation can be calculated using a probability density function based on statistical information summarizing the statistical values ​​of the waiting time of pallet K00001 in yard G01 shown in Fig. 5. Then, the time τ at which the product of the first probability and the second probability is maximum is identified as the maximum likelihood estimate of the waiting start time.

[0057]

number

[0058] Here, t Q01 is the end time (2022 / 12 / 1 14:10), t Q04 is the start time (2022 / 12 / 1 13:44), P'(Q01,t Q01 -τ) is the waiting time for loading and unloading pallet type A at the storage area Q01. Q01 -τ-T~t Q01 -τ+T (T is a fixed parameter) interval, P(VH0005,Q01,Q04,τ-t Q04 ) is the time taken for the transport vehicle VH0005 to move between the storage sites Q01 and Q04. Q04 -T~τ-t Q04 +T (T is a fixed parameter). P'(*) is a function that approximates the probability density function based on the statistical value information shown in Figure 5 and returns a probability value using that function. The maximum likelihood estimate is calculated by calculating the time τ at which formula (2) is maximized. In this embodiment, the maximum value of formula (2) is maximized at time τ = 2022 / 12 / 1 14:03, and the work list corrected accordingly is as shown in Figure 11. This completes the processing of step S3, and the operation analysis processing proceeds to step S4.

[0059] In the process of step S4, the analysis unit 14 generates information on which the Gantt chart display is based using the work list corrected in the process of step S3, and displays the Gantt chart by outputting the generated information to the display unit 15. An example of the Gantt chart displayed on the display unit 15 is shown in FIG. 12. In the Gantt chart shown in FIG. 12, the work performed by the transport vehicle VH0007 and the overhead crane 2 of the storage site G05 is displayed. The rectangles in each row represent traveling work for the transport vehicle VH0007, and loading and unloading work of the overhead crane 2 for the storage site G05. There are two types of work in the storage site G05: storing (the work of putting the coil C on the pallet P into the storage site) and unloading (the work of taking the coil C out of the storage site and placing it on the pallet P), but both are performed by one overhead crane 2. Work A highlighted in the work of the transport vehicle VH0007 represents work that is delayed by more than a set time from the original plan.

[0060] When the user selects an operation on the Gantt chart by operating the user setting unit 16, the analysis unit 14 displays the preceding operations that prevent the selected operation from being executed at an earlier time (hereinafter, the chain of preceding operations is referred to as a "delay chain path"). FIG. 13 shows a state in which the delay chain path is displayed after selecting operation A of the transport vehicle VH0007 highlighted in FIG. 12. A path made of operations with a thick rectangular frame indicating the operation is a delay chain path. By looking at the delay chain path, it is possible to recognize the operation group that delays the execution time of the selected operation. In the example shown in FIG. 13, it can be seen that the selected operation (the operation of transporting the pallet P loaded with the coil C at the storage site G05) is delayed by the preceding unloading operation at the storage site G05 (the operation of placing the coil C transported by the selected operation on the pallet), and further, the unloading operation cannot be executed earlier due to the numerous storage operations and unloading operations executed before it. Therefore, it can be considered that one of the causes of the delay of operation A is that the load on the overhead crane 2 at the storage site G05 is large during this time period. Therefore, in order to eliminate work delays, it may be possible to reduce delays by taking measures such as changing the work plan creation policy and leveling out the workload at site G05.

[0061] Here, a method for generating a delay chain path will be described. When a user of the present device selects one operation, the analysis unit 14 generates a preceding operation path from the selected operation in the following procedure.

[0062] (a) Generate precedence relationship information for selected operations (b) Generate a path from the selected operation by tracing back the precedence relationships that satisfy the following conditions: (Condition) In the precedence relationship, the length of "successor operation start time" - "predecessor operation completion time" is less than or equal to the set value

[0063] An example of the precedence relationship information generated in the above process (a) is shown in FIG. 14. In the example shown in FIG. 14, what is indicated by the arrow is the precedence relationship of the work, and the analysis unit 14 arranges the work of each piece of equipment (the overhead crane 2 in the storage site G03 and the transport vehicle VH0008) in the order in which they were performed, and generates the precedence relationship in order from the work performed at the latest time to the first. In addition, the analysis unit 14 generates a precedence relationship for the work using the same pallet (pallet A00010 in the example shown in FIG. 12) between the work of the overhead crane in the storage site G03 and the transport vehicle VH0008. Then, the analysis unit 14 generates a path by tracing the precedence relationship from the clicked work in process (b) among the precedence relationships generated here. Process (b) is a situation in which the subsequent work is started soon after the completion of the preceding work, so there is no leeway time between the works. In such a case, it is considered that the preceding work is pushing back the start time of the subsequent work, and it is interpreted that the preceding work is causing the delay. This process is repeated for each preceding task, until the process finally reaches the task that is the initial trigger of the delay. The delay chain path information obtained here is sent to the display unit 15 and displayed as shown in Figure 13.

[0064] As is clear from the above description, in the operation analysis device 10 which is one embodiment of the present invention, first, the inconsistency extraction unit 12 generates a list of tasks performed by equipment and extracts inconsistencies in the task list based on predetermined rules. Then, the performance information correction unit 13 corrects the inconsistencies in the task list for each piece of equipment extracted by the inconsistency extraction unit 12 in accordance with predetermined rules, and the analysis unit 14 outputs information that supports the process of identifying the cause of a task delay using the task list for each piece of equipment corrected by the performance information correction unit 13. This allows the performance data to be appropriately corrected even if there are inconsistencies in the performance data, making it possible to perform highly accurate operation analysis.

[0065] Although the embodiments of the present invention have been described above, the present invention is not limited by the descriptions and drawings that form part of the disclosure of the present invention according to the present embodiments. In other words, other embodiments, examples, and operation techniques, etc., made by those skilled in the art based on the present embodiments are all included in the scope of the present invention. [Explanation of symbols]

[0066] 1 Logistics equipment 2. Overhead crane 3. Loading Crane 4 Transport vehicles 5. Factory 6 Warehouse 7 Vehicle entrance / exit 10 Operational Analysis Equipment 11 Input section 12 Inconsistency extraction part 13. Performance Information Correction Department 14 Analysis Department 15 Display section 16 User Settings 17 Operation Information Database (Operation Information DB) 17a Equipment information storage department 17b Work performance information storage section 17c Luggage Information Storage Department A Work C coil P Pallet (loading area) V Traction section

Claims

1. An operation analysis device that analyzes an operation state of a task of moving cargo from a plurality of transport sources to a plurality of transport destinations using a plurality of pieces of equipment, an input unit for acquiring information about the facility, information about the cargo, and performance data of the work performed by the facility; an inconsistency extraction unit that generates a list of operations performed by each piece of equipment using the information acquired by the input unit, and extracts inconsistencies in the operation list based on a predetermined rule; a performance information correction unit that corrects inconsistencies in the work list for each piece of equipment extracted by the inconsistency extraction unit in accordance with a predetermined rule; an analysis unit that outputs information for supporting a process of identifying a cause of a delay in an operation using the operation list for each piece of equipment corrected by the performance information correction unit; An operation analysis device comprising:

2. 2 . The operation analysis device according to claim 1 , wherein when the inconsistency is a missing piece of data relating to time, the performance information correcting unit estimates the missing data based on a preset probability distribution.

3. 3 . The operation analysis device according to claim 1 , wherein the analysis unit generates a delay chain path of an operation by using a precedence relationship between the operations, and outputs information about the generated delay chain path of the operation.

4. 3. The operation analysis device according to claim 2, wherein, when the inconsistency portion is a missing data related to time, the performance information correction unit targets a piece of equipment from which the inconsistency portion has been extracted, defines a time τ as a time when the data is missing, and calculates, as a maximum likelihood estimate, the time τ at which a product of a first probability calculated from a probability distribution for a work time of a first operation performed by the equipment up until immediately before the time τ and a second probability calculated from a probability distribution for a work time of a second operation performed by the equipment from immediately after the time τ is maximized.

5. The operation analysis device according to claim 1 , wherein the inconsistency extraction unit extracts the inconsistency in the task list based on missing data in the task list and insufficient data regarding time.

6. 2. The operation analysis device according to claim 1, wherein, when the inconsistency is a missing data regarding the location of the source or destination of the transport, the performance information correction unit is executed immediately before or after the extracted work of the inconsistency, extracts other work having common information regarding the transported cargo of the work, and estimates the missing location data based on the extracted data regarding the location of the other work.

7. An operation analysis method for analyzing an operation status of a task of moving cargo from a plurality of transport sources to a plurality of transport destinations using a plurality of pieces of equipment, comprising: an input step of acquiring information about the equipment, information about the cargo, and performance data of the work performed by the equipment; an inconsistency extraction step of generating a list of operations performed by each piece of equipment using the information acquired in the input step, and extracting inconsistencies in the operation list based on a predetermined rule; a performance information correction step of correcting inconsistencies in the work list for each piece of equipment extracted in the inconsistency extraction step according to a predetermined rule; an analysis step of outputting information for supporting a process of identifying a cause of a delay in work using the work list for each piece of equipment corrected in the performance information correction step; Operational analysis methods, including:

8. 8. The operations analysis method according to claim 7, wherein, when the inconsistency is a missing data related to time, the performance information correcting step estimates the missing data based on a preset probability distribution.

9. 9. The operation analysis method according to claim 7, wherein the analyzing step generates a delay chain path of an operation by using a precedence relationship between the operations, and outputs information about the generated delay chain path of the operation.

10. 9. The operations analysis method according to claim 8, wherein, when the inconsistency portion is a missing data related to time, the performance information correction step targets equipment from which the inconsistency portion is extracted, defines a time at which the data is missing as time τ, and calculates, as a maximum likelihood estimate, the time τ at which a product of a first probability obtained from a probability distribution for an operation time of an operation performed by the equipment immediately before the time τ and a second probability obtained from a probability distribution for an operation time of an operation performed by the equipment immediately after the time τ is maximized.

11. The operations analysis method according to claim 7 , wherein the inconsistency extraction step extracts the inconsistency in the task list based on missing data in the task list and insufficient data regarding time.

12. 8. The operations analysis method according to claim 7, wherein, when the inconsistent portion is a missing data regarding a location of the source or destination of the transport, the performance information correction step is executed immediately before or after the extracted work of the inconsistent portion, to extract other work having common information regarding the transported cargo of the work, and to estimate the data regarding the missing location based on the data regarding the location of the other extracted work.

Images