JP7658235B2 - Analytical device, analytical method, and program - Google Patents

Description

This disclosure relates to an analysis device, an analysis method, and a program.

At production sites that include multiple processes, work analysis is carried out to improve work efficiency. For example, JP 2012-78896 A (Patent Document 1) discloses a technique for updating the standard work time for each process based on the work time reported by the worker. Based on the standard work time for each process, bottleneck processes are identified, personnel are assigned to each process, and other factors are considered in an attempt to improve work efficiency.

JP 2012-78896 A

Since various events can occur at production sites, standard work is not always performed. If work analysis is conducted using data from a period in which a large proportion of time was spent performing non-standard work, the accuracy of the analysis results, such as standard work hours, will decrease. For this reason, it is desirable to classify work into standard and non-standard work before conducting the analysis. However, the criteria for classifying work into standard and non-standard work may differ depending on the production site.

The present disclosure has been made in consideration of the above problems, and its purpose is to provide an analysis device, an analysis method, and a program that can obtain highly accurate analysis results according to the production site.

According to an example of the present disclosure, an analysis device that analyzes the work status of a plurality of processes includes a first acquisition unit, a division unit, a classification unit, a labeling unit, a calculation unit, and a determination unit. The first acquisition unit acquires first information indicating the implementation status of work at each time in each of the plurality of processes. The division unit divides the first information into a plurality of second information corresponding to a plurality of first periods, respectively. The classification unit extracts one or more feature amounts from each of the plurality of second information, and classifies each of the plurality of first periods into a plurality of classes based on the similarity of the one or more feature amounts between the plurality of second information. The labeling unit assigns one label selected from a plurality of labels indicating the type of work to each of the plurality of classes in response to an input to the input device. The plurality of labels include a standard label indicating that the period is a period in which standard work is being performed. The calculation unit calculates the proportion of the total time of the first period classified into the class to which the standard label is assigned for each second period longer than each of the plurality of first periods. The determination unit determines the second period in which the proportion exceeds a threshold as the period to be analyzed.

According to this disclosure, a period during which there is a high probability that standard work is being performed is determined as the period to be analyzed. This improves the accuracy of the analysis results when analyzing standard work. Furthermore, the user can select a label to be attached to each class depending on the production site and input the selection result into the input device. This allows the user to easily classify work into standard and non-standard work depending on the production site. From the above, it is possible to obtain highly accurate analysis results depending on the production site.

In the above disclosure, the labeling unit provides a screen for assisting in labeling each of the multiple classes, and labels each of the multiple classes in response to input to the screen. According to this disclosure, the user can easily perform labeling via the screen.

In the above disclosure, the labeling unit displays the Gantt chart indicated by the first information on the screen.

According to this disclosure, a user can appropriately label each of multiple classes while checking the Gantt chart.

In the above disclosure, the labeling unit receives designation of one of a plurality of classes on the screen, and makes the display format of the first period classified into one of a plurality of first periods different from the display format of the remaining periods on the Gantt chart.

According to this disclosure, a user can appropriately label a class based on the characteristics of the Gantt chart for the first period classified into that class.

In the above disclosure, the labeling unit receives designation of one of the multiple classes on the screen, and displays on the screen a Gantt chart indicated by second information corresponding to the first period classified into one of the multiple first periods.

This disclosure also allows the user to appropriately label the classes based on the characteristics of the Gantt chart for the first period classified into each class.

In the above disclosure, the analysis device further includes a second acquisition unit that acquires a video obtained by imaging a production site including multiple processes. The labeling unit accepts designation of one of the multiple classes on the screen, extracts a partial video of a first period classified into one of the multiple first periods from the video, and further provides a screen on which the partial video is played.

According to this disclosure, a user can appropriately label a class based on the partial videos of the first period classified into each class.

In the above disclosure, the division unit determines the work start time or work end time of one selected process from among the multiple processes as a division timing, and determines the period between two consecutive division timings as a first period. The labeling unit displays on the screen, for each of the multiple classes, a histogram showing the frequency distribution of the duration of the first period classified into that class.

According to this disclosure, a user can appropriately label a class by checking a histogram of the duration of the first period classified into each class.

In the above disclosure, the analysis device further includes an analysis unit that analyzes the standard work time of each of the multiple processes based on the third information of the analysis period in the first information. According to this disclosure, the standard work time of each process can be determined with high accuracy.

According to an example of the present disclosure, an analysis method for analyzing the work status of a plurality of processes includes first to sixth steps. The first step is a step of acquiring, for each time, first information indicating the implementation status of work at each time in each of a plurality of processes. The second step is a step of dividing the first information into a plurality of second information corresponding to a plurality of first periods, respectively. The third step is a step of extracting one or more feature amounts from each of the plurality of second information, and classifying each of the plurality of first periods into a plurality of classes based on the similarity of the one or more feature amounts between the plurality of second information. The fourth step is a step of attaching one label selected from a plurality of labels indicating the type of work to each of the plurality of classes in response to an input to an input device. The plurality of labels include a standard label indicating that the period is a period in which standard work is being performed. The fifth step is a step of calculating the proportion of the total time of the first period classified into the class to which the standard label is attached for each second period longer than each of the plurality of first periods. The sixth step is a step of determining the second period in which the proportion exceeds a threshold as the period to be analyzed.

According to one example of the present disclosure, a program causes a computer to execute the above-mentioned analysis method. These disclosures also make it possible to obtain highly accurate analysis results according to the production site.

This disclosure makes it possible to obtain highly accurate analysis results according to the production site.

1 is a diagram illustrating an example of a system to which an analysis device according to an embodiment is applied; FIG. 4 is a diagram showing a process flow of the analysis device according to the embodiment. FIG. 2 is a schematic diagram illustrating an example of a hardware configuration of an information processing device. FIG. 11 is a diagram illustrating an example of flow line information. FIG. 2 is a diagram showing an example of frames included in a video captured by a camera. FIG. 2 is a schematic diagram illustrating an example of a hardware configuration of an analysis device according to an embodiment. FIG. 2 is a diagram illustrating an example of a functional configuration of an analysis device according to an embodiment. FIG. 11 illustrates an example of processing by a division unit; FIG. 1 shows cycle durations for representatives of each class. FIG. 13 is a diagram showing an example of a screen provided by a labeling unit. FIG. 13 is a diagram showing an example of analysis. FIG. 13 is a diagram showing an example of a standard operation compliance rate calculated for each period Tb. FIG. 13 is a diagram showing an example of a histogram representing a frequency distribution of the duration of cycle periods classified into each class.

The embodiment of the present invention will be described in detail with reference to the drawings. Note that the same or corresponding parts in the drawings will be given the same reference numerals and their description will not be repeated. The various modified examples described below may be combined as appropriate.

§1 Application Examples Application examples of the analysis device according to the present embodiment will be described with reference to Figures 1 and 2. Figure 1 is a diagram showing an example of a system to which the analysis device according to the embodiment is applied. As shown in Figure 1, the system 1 includes an analysis device 10, an information processing device 20, a PLC (Programmable Logic Controller) 30, a plurality of devices 40, and a camera 50.

The multiple devices 40 are arranged at the production site 2. The production site 2 includes multiple processes Pr. At the production site 2, for example, various products are produced through multiple processes Pr. The multiple processes Pr include, for example, a "painting" process, a "main work assembly" process, a "main work assembly into the main body" process, and an "inspection" process. When it is necessary to distinguish between the multiple processes Pr, they are distinguished by adding suffixes such as "(1)", "(2)", "(3)", ..., "(n)" to the reference numerals. For example, they are distinguished by notation such as "process Pr(1)", "process Pr(2)", ..., "process Pr(n)". When there is no particular need to distinguish between the multiple processes, they are simply referred to as "process Pr".

The multiple devices 40 are used in multiple processes Pr, respectively. In other words, the processes Pr and the devices 40 are associated in advance. When it is necessary to distinguish between the multiple devices 40, they are distinguished by adding suffixes such as "(1)", "(2)", ..., "(n)" to the reference numerals, and when there is no particular need to distinguish between them, they are simply referred to as "devices 40". For example, one or more devices 40(m) are used to carry out process Pr(m). In other words, one or more devices 40(1) are used to carry out process Pr(1). Similarly, one or more devices 40(2) are used to carry out process Pr(2).

The PLC 30 is a control device that controls the entire production site 2 and is communicatively connected to each of the multiple devices 40.

Various types of industrial Ethernet (registered trademark) are used as a network that connects the PLC 30 and the multiple devices 40 so that they can communicate with each other. Known examples of industrial Ethernet (registered trademark) include EtherCAT (registered trademark), Profinet IRT, MECHATROLINK (registered trademark)-III, Powerlink, SERCOS (registered trademark)-III, and CIP Motion, and any of these may be used. Furthermore, a field network other than industrial Ethernet (registered trademark) may also be used. For example, if motion control is not performed, DeviceNet, CompoNet/IP (registered trademark), etc. may be used.

The PLC 30 operates as a master in the master-slave control system, and acquires information as input data from each of the multiple devices 40 acting as input devices (measurement devices). The PLC 30 executes arithmetic processing using the acquired input data according to a pre-installed user program. The PLC 30 determines the control content for the master-slave control system in response to the execution of the arithmetic processing, and outputs control data corresponding to the control content to each of the multiple devices 40. The PLC 30 repeatedly acquires input data from each of the multiple devices 40 and acquires control data for each of the multiple devices 40 at a predetermined cycle (control cycle).

The multiple devices 40 operate as slaves in the master-slave control system. The multiple devices 40 are input devices that repeatedly transmit input data to the PLC 30 at a predetermined control period, or are output devices that repeatedly receive control data from the PLC 30 at a predetermined control period and operate according to the received control data. The multiple devices 40 may include, for example, a sensor (e.g., a photoelectric sensor) as an input device that transmits detection results to the PLC 30, a barcode reader that transmits read results, an inspection machine (tester) that transmits inspection results, etc. The multiple devices 40 may also include a PT (Programmable Terminal) to which multiple input devices are connected. Furthermore, the multiple devices 40 may also include a robot as an output device that performs screw tightening, picking, etc.

The camera 50 is installed in a position (typically on the ceiling) that provides a bird's-eye view of the entire production site 2, and generates video data (hereinafter simply referred to as "video") by capturing images of the entire production site 2. The camera 50 is, for example, a wide-angle camera or an ultra-wide-angle camera.

The information processing device 20 is communicatively connected to the PLC 30 and the camera 50. The information processing device 20 generates flow line information using at least one of the information acquired from the PLC 30 and the video acquired from the camera 50. The flow line information indicates the status of work being performed at each of the multiple processes Pr at each time. In other words, the flow line information indicates the transition of the process Pr at which work is being performed among the multiple processes Pr. In further other words, the flow line information indicates whether or not work is being performed at each of the multiple processes Pr at each time.

The analysis device 10 is communicatively connected to the information processing device 20 and the PLC 30, and analyzes the work status of multiple processes Pr. The analysis device 10 is, for example, a general-purpose computer, and is connected to a display device and an input device.

Figure 2 is a diagram showing the process flow of the analysis device according to the embodiment. The left side of Figure 2 shows a flowchart showing the process flow of the analysis device 10, and the right side of Figure 2 shows the content of each process.

2, the analysis device 10 acquires from the information processing device 20 traffic line information 60 indicating the status of work being performed at each time in each of the multiple processes Pr (step S1). The traffic line information 60 indicates the time periods during which work is being performed in each of the multiple processes Pr (time periods indicated by hatching in FIG. 2).

The analysis device 10 divides the movement line information 60 into a plurality of pieces of information 61 that respectively correspond to a plurality of cycle periods Ta (step S2). In FIG. 2, a portion of the movement line information 60 is divided into pieces of information 61_1 to 61_4 that respectively correspond to cycle periods Ta_1 to Ta_4.

The cycle period Ta is determined, for example, based on the work start time or work end time of one selected process from among the multiple processes Pr. In the example shown in FIG. 2, the analysis device 10 determines the work end time of the last process Pr (5) as the division timing, and determines the period between two consecutive division timings as the cycle period Ta. Therefore, the length of the cycle period Ta changes depending on the content of the work, and is not constant. The cycle period Ta is, for example, a period of about several minutes.

Next, the analysis device 10 extracts one or more feature values from each of the multiple information 61, and classifies each of the multiple cycle periods Ta into multiple classes based on the similarity of the one or more feature values between the multiple information 61 (step S3). For example, when the work order, work time, etc. of multiple processes Pr in two cycle periods Ta are similar, the two cycle periods Ta are classified into the same class. In the example shown in FIG. 2, cycle period Ta_2 is classified into class "Id0", cycle periods Ta_1 and Ta_3 are classified into class "Id1", and cycle period Ta_4 is classified into class "Id2".

Next, the analysis device 10 assigns one label selected from a plurality of labels indicating the type of work to each of the plurality of classes in response to the input to the input device (step S4). The plurality of labels include the label "standard" indicating that the period is one in which standard work (standard work) is performed, and the label "non-standard" indicating that the period is one in which non-standard work (non-standard work) is performed. Each of the labels "standard" and "non-standard" may be assigned a subscript (e.g., a number) to distinguish the type of work. Alternatively, instead of the label "non-standard", a label including text indicating the type of work (e.g., the label "setup") may be used. In the example shown in FIG. 2, the labels "standard", "non-standard 1", and "non-standard 2" are assigned to the classes "Id0", "Id1", and "Id2", respectively.

Next, the analysis device 10 calculates the proportion of the total time of the cycle period Ta that is classified into a class labeled "standard" for each period Tb (e.g., one day) that is longer than each of the multiple cycle periods Ta (hereinafter referred to as the "standard work compliance rate") (step S5).

The analysis device 10 determines the period Tb during which the standard operation compliance rate exceeds the threshold as the analysis period (step S6). The threshold is determined in advance.

According to this embodiment, a period during which there is a high probability that standard work is being performed is determined as the period to be analyzed. This improves the accuracy of the analysis results when analyzing standard work. Furthermore, the user simply selects a label to be attached to each class while checking the information 61 of each class according to the production site 2, and inputs the selection result into the input device. This allows the user to easily classify work into standard and non-standard work according to the production site 2. From the above, it is possible to provide an analysis device 10 that can obtain highly accurate analysis results according to the production site.

§2 Specific example <Hardware configuration of information processing device>
Fig. 3 is a schematic diagram showing an example of a hardware configuration of an information processing device. The information processing device 20 typically has a structure according to a general-purpose computer architecture. As shown in Fig. 3, the information processing device 20 includes a processor 21 such as a CPU (Central Processing Unit) or an MPU (Micro-Processing Unit), a memory 22, a storage 23, a camera interface 24, and a communication interface 25. These components are connected to each other via a bus so as to be able to communicate data with each other.

The processor 21 implements various processes according to this embodiment by expanding various programs stored in the storage 23 into the memory 22 and executing them.

Memory 22 is typically a volatile storage device such as a dynamic random access memory (DRAM), and stores programs read from storage 23, video images received from camera 50, etc.

The camera interface 24 mediates data transmission between the processor 21 and the camera 50. More specifically, an image capture instruction is output from the processor 21 to the camera 50 via the camera interface 24. The camera interface 24 outputs a video image received from the camera 50 in response to the image capture instruction to the processor 21.

The communication interface 25 mediates data transmission between the processor 21 and an external device (e.g., the PLC 30, the analysis device 10). The communication interface 25 typically includes Ethernet (registered trademark) or USB (Universal Serial Bus).

Storage 23 is typically a non-volatile magnetic storage device such as a hard disk drive. Storage 23 stores a movement line information generating program 26 executed by processor 21 and movement line information 60 generated by execution of movement line information generating program 26.

Figure 4 is a diagram showing an example of flow line information. The flow line information 60 illustrated in Figure 4 is represented in table format. Each record in the table corresponds a process ID that identifies a process Pr to a work start time and a work end time for the process Pr identified by the process ID. At the production site 2, products are produced one by one. Therefore, when multiple products are produced sequentially, work is repeatedly performed in multiple processes Pr(1) to Pr(n). Therefore, the flow line information 60 includes multiple records that indicate the same process ID.

<Method of generating flow line information>
The processor 21 of the information processing device 20 may generate the flow line information 60 based on information acquired from the PLC 30 , or may generate the flow line information 60 based on video acquired from the camera 50 .

(First Generation Method)
A first generation method for generating the flow line information 60 based on the information acquired from the PLC 30 will be described.

The PLC 30 receives operation results indicating the content and results of the operations performed by the devices 40 from each of the multiple devices 40, for example, repeatedly at a predetermined control period. That is, the devices 40 repeatedly transmit operation results indicating the content and results of the operations actually performed when carrying out the process Pr to the PLC 30 at a predetermined period.

For example, the PLC 30 acquires the measurement result, which is the result of the measurement operation performed by the equipment 40 as an input device (measuring device), as the operation result of the equipment 40. Furthermore, if the equipment 40 is an inspection machine, the PLC 30 acquires the result of the inspection operation performed by the equipment 40 as the operation result of the equipment 40. Furthermore, the PLC 30 acquires, for example, the result of the output operation performed by the equipment 40 as an output device as the operation result of the equipment 40. If the equipment 40 is a robot that performs screw tightening, picking, etc., the PLC 30 acquires the number of screw tightenings, the picking result (picking success or picking error), etc. as the operation result of the equipment 40.

The PLC 30 generates process information regarding the operation results that are repeatedly received from the device 40 at a predetermined interval, and transmits the generated process information to the information processing device 20.

The information processing device 20 identifies the operation start time and operation completion time of the operation performed by the equipment 40 when carrying out the process Pr based on the process information received from the PLC 30. The operation start time is the time when the equipment 40 used in the process Pr starts to perform the operation when carrying out the process Pr. The operation completion time is the time when the equipment 40 completes the operation. The information processing device 20 creates a record that includes a process ID that identifies the process Pr to which the equipment 40 belongs, and has the operation start time and operation completion time as the work start time and work end time, and generates traffic line information 60 that includes the created record.

(Second Generation Method)
A second generation method for generating the flow line information 60 based on the video captured by the camera 50 will be described.

Figure 5 is a diagram showing an example of a frame included in a video captured by a camera. Figure 5 shows frames of a video obtained by capturing an image of a production site 2 that includes five processes Pr. As shown in Figure 5, each frame of the video shows equipment (including devices 40) located at the production site 2 and a worker Pe working at the production site 2.

A monitoring area Ar is set in advance for each of the five processes Pr. Specifically, monitoring areas Ar(1) to Ar(5) are set for processes Pr(1) to Pr(5), respectively. Monitoring areas Ar(1) to Ar(5) are areas within the video frame. Monitoring areas Ar(1) to Ar(5) are, for example, rectangular, and are defined by the coordinates of four vertices.

The processor 21 of the information processing device 20 detects the position in the frame where the worker Pe appears, using a known object recognition technique. Specifically, the processor 21 detects one or more pixels where the worker Pe appears, using a known object recognition technique. The processor 21 identifies a rectangular area Ap that includes the detected one or more pixels, and determines the center of the rectangular area Ap as the position Pp of the worker Pe. In the example shown in FIG. 5, the positions Pp(1) and Pp(2) of the workers Pe(1) and Pe(2) are detected, respectively.

At the imaging time of each frame, the processor 21 determines whether or not the worker Pe is present in the monitoring area Ar set for each process Pr. Specifically, the processor 21 determines that the worker Pe is present in the monitoring area Ar in response to the position Pp of the worker Pe being included within the monitoring area Ar. The time period during which the worker Pe is present in the monitoring area Ar is regarded as the period during which work is being carried out in the process Pr corresponding to that monitoring area Ar.

Therefore, for each process Pr, processor 21 identifies multiple consecutive frames from the video in which it is determined that worker Pe is present in the monitoring area Ar corresponding to that process Pr. Processor 21 creates a record for the identified multiple frames, including a process ID that identifies that process Pr. Processor 21 determines the image capture time of the first frame of the identified multiple frames as the work start time of that record, and determines the image capture time of the last frame of the identified frames as the work end time of that record. Processor 21 generates traffic line information 60 that includes the records created in this way.

<Hardware configuration of the analysis device>
6 is a schematic diagram showing an example of a hardware configuration of an analysis device according to an embodiment of the present invention. As shown in FIG. 6, the analysis device 10 typically has a structure conforming to a general-purpose computer architecture.

Specifically, the analysis device 10 includes a processor 11 such as a CPU or an MPU, a memory 12, a storage 13, a display controller 14, an input interface 15, and a communication interface 16. These components are connected to each other via a bus so that they can communicate data with each other.

The processor 11 deploys various programs stored in the storage 13 in the memory 12 and executes them to realize various processes according to this embodiment.

Memory 12 is typically a volatile storage device such as a DRAM, and stores programs and other data read from storage 13.

Storage 13 is typically a non-volatile magnetic storage device such as a hard disk drive. Storage 13 stores analysis program 17 executed by processor 11 and traffic line information 60 received from information processing device 20. Analysis program 17 installed in storage 13 is distributed in a state stored on a memory card or the like.

The display controller 14 is connected to the display device 70 and outputs signals to the display device 70 to display various information in accordance with internal commands from the processor 11.

The input interface 15 mediates data transmission between the processor 11 and an input device 75 such as a keyboard, mouse, touch panel, or dedicated console. In other words, the input interface 15 accepts operation commands given by the user operating the input device 75.

The communication interface 16 mediates data transmission between the processor 11 and an external device (e.g., the information processing device 20, the PLC 30). The communication interface 16 typically includes Ethernet (registered trademark) or USB (Universal Serial Bus). The analysis program 17 may be downloaded from a distribution server or the like via the communication interface 16.

When using a computer having a structure conforming to the above-mentioned general-purpose computer architecture, an OS (Operating System) for providing basic computer functions may be installed in addition to an application for providing the functions according to this embodiment. In this case, the program according to this embodiment may execute processing by calling necessary modules from among program modules provided as part of the OS in a predetermined order and timing. In other words, the program according to this embodiment itself does not include the above-mentioned modules, and processing may be executed in cooperation with the OS.

Alternatively, some or all of the functions provided by execution of analysis program 17 may be implemented as dedicated hardware circuits.

<Functional configuration of the analysis device>
7 is a diagram showing an example of a functional configuration of an analysis device according to an embodiment. As shown in FIG. 7, the analysis device 10 includes a first acquisition unit 101, a second acquisition unit 102, a division unit 103, a classification unit 104, a labeling unit 105, a calculation unit 106, a determination unit 107, an analysis unit 108, and a storage unit 110. The first acquisition unit 101 and the second acquisition unit 102 are realized by the communication interface 16 and the processor 11 that executes the analysis program 17. The division unit 103, the classification unit 104, the calculation unit 106, the determination unit 107, and the analysis unit 108 are realized by the processor 11 executing the analysis program 17. The labeling unit 105 is realized by the display controller 14, the input interface 15, and the processor 11 that executes the analysis program 17.

The first acquisition unit 101 acquires movement line information 60 indicating the status of work being performed at each time in each of the multiple processes Pr, and stores the acquired movement line information 60 in the memory unit 110.

The second acquisition unit 102 acquires a video 65 obtained by capturing an image of the production site 2, and stores the acquired video 65 in the storage unit 110. The second acquisition unit 102 may acquire the video 65 from the information processing device 20, or may acquire the video 65 from the camera 50.

The division unit 103 divides the movement line information 60 into multiple pieces of information 61 each corresponding to a multiple cycle period.

FIG. 8 is a diagram showing an example of processing by the division unit. The division unit 103 determines the work start time or work end time of one selected process Pr from among the multiple processes Pr as the division timing. The division unit 103 determines the period between two consecutive division timings as the cycle period Ta. In the example shown in FIG. 8, the work end time of the last process Pr (9) of the nine processes Pr is determined as the division timing, and the flow line information 60 is divided into information 61_1, 61_2, ... corresponding to cycle periods Ta_1, Ta_2, ... respectively. The division timing is determined according to the production site 2. For example, the work start time of the first process Pr (1) may be determined as the division timing.

The classification unit 104 shown in FIG. 7 extracts one or more feature values from each of the multiple pieces of information 61, and classifies each of the multiple cycle periods Ta into multiple classes based on the similarity of the one or more feature values between the multiple pieces of information 61. The one or more feature values include, for example, a parameter that characterizes the work order of the multiple processes Pr in the cycle period Ta, the work time of each process Pr, etc.

The classification unit 104 classifies each of the multiple cycle periods Ta into multiple classes using a known clustering technique. A technique using a dendrogram can be adopted as a known clustering technique. A dendrogram is a diagram in which clusters are arranged in a tree diagram shape for a certain sample group. Therefore, by using distance as a threshold for the dendrogram, the sample group can be classified (clustered) into several small unit sample groups. In other words, the classification unit 104 classifies each of the multiple cycle periods Ta into multiple classes using a certain threshold for the distance. The classification unit 104 assigns a class ID to each of the multiple classes to identify the class.

Figure 9 shows the representative cycle period for each class. As shown in Figure 9, the work time and work order of each of the multiple processes Pr differ between classes.

The labeling unit 105 shown in FIG. 7 assigns one label selected from a plurality of labels indicating the type of work to each of the plurality of classes in response to input to the input device 75. Specifically, the labeling unit 105 provides a screen for assisting in labeling each of the plurality of classes, and assigns a label to each of the plurality of classes in response to input to the screen.

FIG. 10 is a diagram showing an example of a screen provided by the labeling unit. As shown in FIG. 10, a screen 80 includes areas 81 and 82, a group of radio buttons 83, a video playback button 84, and a save button 85.

Area 81 displays a Gantt chart showing the time periods during which work is being performed in each of the multiple processes Pr. The Gantt chart displayed in area 81 is generated based on the flow line information 60.

Area 82 displays a table for setting a label for each of a number of classes. The table displayed in area 82 includes a record for each class. Each record has a field 82a in which a class ID that identifies the class is written, a field 82b in which the number (frequency) of pieces of information 61 classified into the class is written, and a field 82c in which a label set for the class is written. Fields 82a and 82b each describe a class ID and frequency according to the classification result by classification unit 104.

Field 82c describes the label according to the user's input. That is, the user enters the label selected from the drop-down menu into field 82c.

The example shown in FIG. 10 shows the labels that are assigned when the standard work involves producing products one at a time. That is, labels that can be entered in field 82c include "Standard 1," "Standard 2," "Batch," "Setup," "Different organization," and "Continuous work."

"Standard 1" is assigned to a class that indicates the work sequence and work time of multiple processes Pr when a first variety of products is produced one by one. "Standard 2" is assigned to a class that indicates the work sequence and work time of multiple processes Pr when a second variety of products is produced one by one. "Batch" is assigned to a class that indicates the work sequence and work time of multiple processes Pr when multiple products are produced at once. "Setup" is assigned to a class that indicates the work sequence and work time of multiple processes Pr when a product is changed. "Different organization" is assigned to a class that indicates the work sequence and work time of multiple processes Pr when the movement of workers between processes differs from the standard. "Continuous work" is assigned to a class that indicates the work sequence and work time of multiple processes Pr when products are backed up in some processes.

The group of radio buttons 83 is used to specify one of a number of classes. In the Gantt chart displayed in the area 81, it is preferable that the cycle periods Ta classified into the class specified by the group of radio buttons 83 are displayed in a different display format from the remaining cycle periods Ta. For example, only the cycle periods Ta classified into the class specified by the group of radio buttons 83 are displayed in a different color or highlighted. This allows the user to easily grasp the work sequence and work time of the multiple processes Pr for each class. As a result, the user can set an appropriate label for each class.

In response to clicking the video playback button 84, the labeling unit 105 provides a window in which a video of the cycle period Ta classified into the class specified by the radio button group 83 is played. The labeling unit 105 extracts a partial video of the cycle period Ta classified into the class specified by the radio button group 83 from the video 65 stored in the storage unit 110, and plays the extracted partial video. This allows the user to check the state of the production site 2 for each class and set an appropriate label. For example, if the user confirms that a worker is changing the product type in the video of the selected class, the user can set the label "setup" for that class. Alternatively, if the user confirms that the movement of workers between processes is different from the standard in the video of the selected class, the user can set the label "different organization" for that class.

In response to clicking the save button 85, the labeling unit 105 stores in the memory unit 110 the labeling information 66 that associates the class ID and label described in each record of the table displayed in the area 82.

The calculation unit 106 shown in FIG. 7 calculates the proportion of the total time of the cycle period Ta classified into a class with a standard label (standard work compliance rate) for each period Tb (e.g., one day) longer than each of the multiple cycle periods Ta. For each period Tb, the calculation unit 106 extracts two or more cycle periods Ta included in the period Tb. For each of the two or more extracted cycle periods Ta, the calculation unit 106 identifies the class classified by the classification unit 104 and reads out the label associated with the class from the labeling information 66. The calculation unit 106 calculates the total time of the cycle periods Ta classified into a class with a standard label (in the example shown in FIG. 10, the labels "standard 1" and "standard 2"). The calculation unit 106 calculates the standard work compliance rate by dividing the total time by the period Tb.

The determination unit 107 determines the period Tb during which the standard work compliance rate exceeds the threshold as the analysis period. The analysis unit 108 uses the data from the analysis period to analyze the work situation at the production site 2.

FIG. 11 is a diagram showing an example of analysis. The analysis unit 108 extracts information from the movement line information 60 in which the work start time and work end time are included in the analysis target period. Based on the extracted information, the analysis unit 108 creates a histogram of the work time from the work start time to the work end time for each of the multiple processes Pr. The analysis unit 108 calculates the most frequent value of the histogram created for each process Pr as the standard work time for that process Pr. Note that the analysis unit 108 can employ various known analysis methods in addition to the analysis method shown in FIG. 11.

<Processing example>
FIG. 12 is a diagram showing an example of the standard work compliance rate calculated for each period Tb. FIG. 12 shows the standard work compliance rate calculated when the length of the period Tb is one day. In FIG. 12, a table 90 shows the standard work compliance rate and the ratio of the total time of the cycle period of the classes labeled "Non-standard 1", "Non-standard 2", and "Non-standard 3" for each day from April 27 to May 25. A table 91 shows the standard work time calculated based on all information of each process Pr in the flow line information 60 for each day from April 27 to May 25. A table 92 shows the standard work time of each process Pr calculated using only information of the cycle period of the class labeled "standard" in the flow line information 60 for each day from April 27 to May 25.

In the example shown in FIG. 12, the frequency of batch production was high in the "intermediate inspection" process on April 28 and May 3, resulting in a low compliance rate with standard work. As a result, as can be seen from a comparison of tables 91 and 92, the standard work time for the "intermediate inspection" process calculated using all the information for both days is longer than the standard work time calculated using only the information on the cycle period of the class labeled "standard."

In addition, between May 4th and 7th and May 10th and 12th, the frequency of worker movements between processes deviating from the standard was high, resulting in a low compliance rate with standard work. As a result, as can be seen from a comparison of tables 91 and 92, the standard work times for the "soldering", "assembly" and "intermediate inspection" processes (especially the "intermediate inspection" process) calculated using all the information for both days are longer than the standard work times calculated using only information on the cycle period of the class labeled "standard".

In this way, the accuracy of the standard working time is low when data on days with a low standard working compliance rate is used. Therefore, the determination unit 107 determines the analysis period to be days on which the standard working compliance rate exceeds 80%, for example, so that this data is excluded. This improves the accuracy of the analysis results using data from the analysis period.

<Modification>
The labeling unit 105 may display, on a screen 80 shown in FIG. 10, a histogram representing the frequency distribution of the time length of the cycle periods Ta classified into each of a plurality of classes.

Figure 13 shows an example of a histogram showing the frequency distribution of the duration of cycle periods classified into each class. Figure 13 shows a stacked histogram 95 showing the variation in cycle period Ta for each class.

As shown in FIG. 13, the cycle periods Ta classified into the class labeled "standard" are relatively short. In contrast, the cycle periods Ta classified into the classes labeled "batch" and "setup," which indicate non-standard work, are relatively long. Therefore, the user can set an appropriate label for each class by checking the distribution of the length of the cycle periods Ta for each class.

In the above description, the labeling unit 105 makes the display format of the cycle period Ta classified into the selected class different from the display format of the remaining periods in the Gantt chart displayed in the area 81 of the screen 80 shown in FIG. 10. However, the labeling unit 105 may display only the Gantt chart indicated by the information 61 corresponding to the cycle period Ta classified into the selected class in the area 81 of the screen 80.

§3 Supplementary Note As described above, the present embodiment includes the following disclosure.

(Configuration 1)
An analysis device (10) for analyzing the working status of a plurality of processes (Pr),
A first acquisition unit (101, 11) that acquires first information (60) indicating an implementation status of work at each time in each of the plurality of processes (Pr);
A division unit (103, 11) that divides the first information (60) into a plurality of second information (61) corresponding to a plurality of first periods (Ta),
a classification unit (104, 11) that extracts one or more feature amounts from each of the plurality of pieces of second information (61) and classifies each of the plurality of first periods (Ta) into a plurality of classes based on a similarity between the plurality of pieces of second information (61) of the one or more feature amounts;
a labeling unit (105, 11) that assigns one label selected from a plurality of labels indicating types of work to each of the plurality of classes in response to an input to an input device (75), the plurality of labels including a standard label indicating that the period is one in which standard work is being performed, and the analysis device (10) further includes:
a calculation unit (106, 11) that calculates, for each second period longer than each of the plurality of first periods (Ta), a proportion of a total time of the first period (Ta) classified into a class to which the standard label is attached;
and a determination unit (107, 11) that determines the second period in which the ratio exceeds a threshold as a period to be analyzed.

(Configuration 2)
The labeling unit (105, 11)
providing a screen (80) for assisting in labelling each of said plurality of classes;
2. The analysis device (10) of configuration 1, further comprising: a display (80) for displaying a label for each of the plurality of classes in response to an input to the display (80).

(Configuration 3)
The analysis device (10) according to configuration 2, wherein the labeling unit (105, 11) displays a Gantt chart indicated by the first information on the screen (80).

(Configuration 4)
The labeling unit (105, 11)
On the screen (80), a designation of one of the plurality of classes is received,
The analysis device (10) according to configuration 3, wherein in the Gantt chart, a display format of a first period (Ta) classified into the one class among the plurality of first periods (Ta) is made different from a display format of the remaining periods.

(Configuration 5)
The labeling unit (105, 11)
On the screen (80), a designation of one of the plurality of classes is received,
The analysis device according to configuration 2, wherein the screen (80) displays a Gantt chart indicated by second information (61) corresponding to a first period (Ta) classified into the one class among the plurality of first periods (Ta).

(Configuration 6)
A second acquisition unit (102, 11) for acquiring a video (65) obtained by imaging a production site (2) including the plurality of processes (Pr),
The labeling unit (105, 11)
On the screen (80), a designation of one of the plurality of classes is accepted,
Extracting a partial moving image of a first period (Ta) classified into the one class among the plurality of first periods (Ta) from the moving image (65);
The analysis device (10) according to any one of configurations 2 to 5, further providing a screen on which the partial moving image is played.

(Configuration 7)
The dividing unit (103, 11)
determining a work start time or a work end time of a selected one of the plurality of processes (Pr) as a division timing;
A period between two consecutive separation timings is determined as the first period (Ta);
The labeling unit (105, 11)
The analysis device (10) of any one of configurations 2 to 6, wherein the screen (80) displays, for each of the plurality of classes, a histogram (95) representing a frequency distribution of the duration of the first period (Ta) classified into that class.

(Configuration 8)
The analysis device (11) according to any one of configurations 1 to 7, further comprising an analysis unit (108, 11) configured to analyze a standard operation time for each of the plurality of processes (Pr) based on third information for the analysis period included in the first information (60).

(Configuration 9)
An analysis method for analyzing work statuses of a plurality of processes (Pr), comprising the steps of:
acquiring first information (60) indicating a status of work performed at each time in each of the plurality of processes (Pr);
Dividing the first information (60) into a plurality of second information (61) corresponding to a plurality of first periods (Ta), respectively;
extracting one or more feature amounts from each of the plurality of pieces of second information (61), and classifying each of the plurality of first periods (Ta) into a plurality of classes based on a similarity between the plurality of pieces of second information (61) of the one or more feature amounts;
and labeling each of the plurality of classes with one label selected from a plurality of labels indicating a type of work in response to an input to an input device (75), the plurality of labels including a standard label indicating that the period is one in which standard work is being performed, the analysis method further comprising:
Calculating a proportion of the total time of the first period (Ta) classified into the class to which the standard label is attached for each second period longer than each of the plurality of first periods (Ta);
determining, as an analysis target period, the second period in which the ratio exceeds a threshold value.

(Configuration 10)
A program for causing a computer to execute an analysis method for analyzing work statuses of a plurality of processes,
The analysis method includes:
acquiring first information (60) indicating a status of work performed at each time in each of the plurality of processes (Pr);
Dividing the first information (60) into a plurality of second information (61) corresponding to a plurality of first periods (Ta), respectively;
extracting one or more feature amounts from each of the plurality of pieces of second information (61), and classifying each of the plurality of first periods (Ta) into a plurality of classes based on a similarity between the plurality of pieces of second information (61) of the one or more feature amounts;
and labeling each of the plurality of classes with one label selected from a plurality of labels indicating a type of work in response to an input to an input device (75), the plurality of labels including a standard label indicating that the period is one in which standard work is being performed, the analysis method further comprising:
Calculating a proportion of the total time of the first period (Ta) classified into the class to which the standard label is attached for each second period longer than each of the plurality of first periods (Ta);
determining, as an analysis target period, the second period in which the ratio exceeds a threshold value.

Although the embodiments of the present invention have been described, the embodiments disclosed herein should be considered to be illustrative and not restrictive in all respects. The scope of the present invention is defined by the claims, and it is intended to include all modifications within the meaning and scope of the claims.

1 System, 2 Production site, 10 Analysis device, 11, 21 Processor, 12, 22 Memory, 13, 23 Storage, 14 Display controller, 15 Input interface, 16, 25 Communication interface, 17 Analysis program, 20 Information processing device, 24 Camera interface, 26 Flow line information generation program, 30 PLC, 40 Equipment, 50 Camera, 60 Flow line information, 61 Information, 65 Video, 66 Labeling information, 70 Display device, 75 Input device, 80 Screen, 81, 82 Area, 82a, 82b, 82c Field, 83 Radio button group, 84 Video playback button, 85 Save button, 90, 91, 92 Table, 95 Stacked histogram, 101 First acquisition unit, 102 Second acquisition unit, 103 Division unit, 104 Classification unit, 105 Labeling unit, 106 Calculation unit, 107 determination unit, 108 analysis unit, 110 storage unit, Ap rectangular area, Ar monitoring area, Pe worker, Pp position, Pr process, Ta cycle period.

Claims (10)

An analysis device for analyzing work statuses of a plurality of processes,
A first acquisition unit that acquires first information indicating an implementation status of work at each time in each of the plurality of processes;
a division unit that divides the first information into a plurality of second information pieces that respectively correspond to a plurality of first periods;
a classification unit that extracts one or more feature amounts from each of the plurality of second information, and classifies each of the plurality of first periods into a plurality of classes based on a similarity between the plurality of second information of the one or more feature amounts;
a labeling unit that, in response to an input to an input device, attaches one label selected from a plurality of labels indicating types of work to each of the plurality of classes, wherein the plurality of labels includes a standard label indicating that the period is one in which standard work is being performed, and the analysis device further comprises:
a calculation unit that calculates, for each second period longer than each of the plurality of first periods, a proportion of a total time of the first period classified into a class to which the standard label is attached;
A determination unit that determines the second period in which the ratio exceeds a threshold as a period to be analyzed. The labeling unit includes:
providing a screen for assisting in labeling each of the plurality of classes;
The analysis device according to claim 1 , further comprising: a label for each of the plurality of classes in response to an input to the screen. The analysis device according to claim 2, wherein the labeling unit displays a Gantt chart indicated by the first information on the screen. The labeling unit includes:
On the screen, a class among the plurality of classes is designated,
The analysis device according to claim 3 , wherein a display format of the first period classified into the one class among the plurality of first periods in the Gantt chart is made different from a display format of the remaining periods. The labeling unit includes:
On the screen, a class among the plurality of classes is designated,
The analysis device according to claim 2 , wherein the screen displays a Gantt chart indicated by second information corresponding to a first period classified into the one class among the plurality of first periods. A second acquisition unit that acquires a video obtained by capturing an image of a production site including the plurality of processes,
The labeling unit includes:
accepting, on the screen, a designation of one of the plurality of classes;
extracting a partial moving image of a first period classified into the one class among the plurality of first periods from the moving images;
The analysis device according to claim 2 , further comprising a screen on which the partial moving image is played. The division unit is
determining a work start time or a work end time of a selected one of the plurality of processes as a division timing;
determining a period between two successive separation timings as the first period;
The labeling unit includes:
The analysis device according to claim 2 , wherein a histogram is displayed on the screen, for each of the plurality of classes, that represents a frequency distribution of the duration of the first periods classified into that class. The analysis device according to any one of claims 1 to 7, further comprising an analysis unit that analyzes the standard work time of each of the plurality of processes based on third information of the analysis target period in the first information. An analysis method for analyzing work statuses of a plurality of processes, comprising the steps of:
acquiring, for each time, first information indicating an implementation status of work at each time in each of the plurality of processes;
Dividing the first information into a plurality of second information pieces each corresponding to a plurality of first time periods;
extracting one or more feature amounts from each of the plurality of second information, and classifying each of the plurality of first periods into a plurality of classes based on a similarity between the plurality of second information of the one or more feature amounts;
and labeling each of the plurality of classes with one label selected from a plurality of labels indicating a type of work in response to an input to an input device, the plurality of labels including a standard label indicating that the period is one in which standard work is being performed, the analysis method further comprising:
calculating, for each second period longer than each of the plurality of first periods, a proportion of the total time of the first period classified into the class to which the standard label is attached;
determining, as an analysis target period, the second period in which the ratio exceeds a threshold value. A program for causing a computer to execute an analysis method for analyzing work statuses of a plurality of processes,
The analysis method includes:
acquiring, for each time, first information indicating an implementation status of work at each time in each of the plurality of processes;
Dividing the first information into a plurality of second information pieces each corresponding to a plurality of first time periods;
extracting one or more feature amounts from each of the plurality of second information, and classifying each of the plurality of first periods into a plurality of classes based on a similarity between the plurality of second information of the one or more feature amounts;
and labeling each of the plurality of classes with one label selected from a plurality of labels indicating a type of work in response to an input to an input device, the plurality of labels including a standard label indicating that the period is one in which standard work is being performed, the analysis method further comprising:
calculating, for each second period longer than each of the plurality of first periods, a proportion of the total time of the first period classified into the class to which the standard label is attached;
determining, as an analysis target period, the second period in which the ratio exceeds a threshold value.

Images