JP7714130B2 - Production system, production line analysis method, learning device, inference device, trained model, and trained model generation method - Google Patents

Description

The present disclosure relates to a production system that analyzes factors that cause a production line to stop, a production line analysis method, a learning device, an inference device, a trained model, and a method for generating a trained model.

In production lines equipped with production equipment, it is common to monitor the operating status of the production line in order to improve productivity. One method for improving productivity is to identify and improve factors that are causing a decline in a Key Performance Indicator (KPI), which is a set production-related operating indicator. Hereinafter, key performance indicators related to production will be referred to as production KPIs.

Patent document 1 discloses an operation status monitoring device that selects problematic events occurring around the time period when production KPIs have declined, displays the selected problematic events on an operation status transition screen on a display, and performs operations on the displayed content.

Patent No. 6961850

However, the operation status monitoring device described in Patent Document 1 only displays the relationship between production KPIs and stoppage factors on a screen that uses a time axis as a comparison standard. As a result, the operation status monitoring device described in Patent Document 1 has the problem of being unable to quantitatively determine how much each stoppage factor has reduced the production KPI on a production line where multiple problematic events occur simultaneously. This problem reduces the efficiency of improvement activities to improve production KPIs.

The present disclosure has been made in consideration of the above and aims to provide a production system that can quantify and present multiple stoppage factors that reduce the production KPI of a production line.

To solve the above-mentioned problems and achieve the objective, the production system disclosed herein analyzes stoppage factors of a production line including multiple production devices. The production system includes: a production record acquisition unit that acquires equipment data related to the production history of each production device from the multiple production devices; an equipment status acquisition unit that acquires equipment data including the equipment status of each production device, including identification information for identifying abnormal conditions that subdivide the equipment status of the production devices, from the multiple production devices; a key performance indicator calculation unit that calculates values of key performance indicators, which are criteria for determining the operating status of the production line, based on the equipment data related to the production history of each production device; a stoppage factor calculation unit that calculates values of multiple stoppage factors that affect the decrease of the key performance indicators based on the equipment data including the equipment status of each production device; and an output unit that displays the values of the key performance indicators and the stoppage factors. The stoppage factor calculation unit calculates the values of the key performance indicators and the stoppage factors as values per unit time for the same value of the same type of time related to production on the production line, and calculates values of unaggregated stoppage factors for which stoppage factor values have not been calculated based on the equipment data including the equipment status of each production device.

This disclosure has the effect of being able to quantify and present multiple stoppage factors that reduce the production KPI of a production line.

FIG. 1 is a diagram showing a configuration of an entire system including a production system according to a first embodiment. 1 is a flowchart showing a procedure for an operation status analysis process performed by a production system according to a first embodiment. FIG. 1 is a diagram showing an example of a first table storing device data used to calculate an availability rate, which is a production KPI, in the production system according to the first embodiment; FIG. 10 is a diagram showing an example of a second table storing device data used to calculate stoppage causes in the production system according to the first embodiment; FIG. 10 is a diagram showing a first example of an analysis result of line stoppage causes displayed by the production system according to the first embodiment. FIG. 10 is a diagram showing a second example of the analysis result of the line stoppage cause displayed by the production system according to the first embodiment. 10 is a flowchart showing a processing procedure of another operation status analysis process performed by the production system according to the first embodiment. FIG. 10 is a diagram showing a configuration of a learning device according to a third embodiment; 10 is a flowchart showing a procedure of a learning process performed by a learning device according to a third embodiment. FIG. 10 is a diagram showing a configuration of an inference device according to a third embodiment. 10 is a flowchart showing a procedure for an inference process performed by an inference device according to a third embodiment. FIG. 11 is a diagram showing an example of the configuration of a processing circuit provided in a learning device according to a third embodiment when the processing circuit is realized by a processor and a memory. FIG. 10 is a diagram showing an example of a processing circuit included in a learning device according to a third embodiment, when the processing circuit is configured using dedicated hardware. FIG. 10 is a diagram showing the configuration of an entire system including a corresponding worker suggestion unit according to a fourth embodiment.

Below, the production system, production line analysis method, learning device, inference device, trained model, and trained model generation method relating to the embodiments are described in detail with reference to the drawings.

Embodiment 1.
1 is a diagram showing the configuration of an overall system 1 including a production system 10 according to the first embodiment. The overall system 1 includes a production line 20 and the production system 10.

The production system 10 functions as a production line analysis system, which analyzes line stoppage factors that cause a production line 20, which is made up of multiple connected production devices 101. The production system 10 collects device data, which is information on the operation of the production devices 101, from the production devices 101 that make up the production line 20, and analyzes the line stoppage factors that cause the production line 20 to stop based on the device data. The line stoppage factors correspond to the factors that caused the production devices 101 to stop, and are factors that affect the decline in the production KPI of the production line 20. In the following description, the line stoppage factors may be referred to as stoppage factors of the production devices 101, stoppage factors, or simply factors. The production system 10 has a data acquisition unit 200 and a data calculation unit 300.

Production line 20 is an example of a production line in which multiple connected production devices 101 sequentially process workpieces (not shown), which are objects to be processed. The multiple production devices 101 in production line 20 are connected together. Therefore, the production capacity of production line 20 decreases if even one production device 101 stops.

While FIG. 1 shows the production devices 101 lined up one by one, the production devices 101 may be arranged so that the arrangement branches off along the way. For example, the production line 20 may include a configuration in which three (N+1)th production devices 101 are arranged after the Nth (N is a natural number) production device 101, with the Nth production device 101 directly connected to each of the (N+1)th production devices 101. Note that the connection between the production devices 101 in embodiment 1 is not limited to cases in which the production devices 101 are physically connected by a free-flow conveyor or the like, but also includes cases in which the transport of work between the production devices 101 is systematically connected by a self-propelled robot, an AGV (Automatic Guided Vehicle), manual labor, or the like.

The production device 101 is a variety of production devices that process workpieces. Examples of the production device 101 include a soldering device, a coating device, a case assembly device, a screw tightening device, and an image inspection device.

Each production device 101 has a production device control unit 1011 that controls the production device 101. The production device control unit 1011 manages time information, which is information on time. For example, the production device control unit 1011 holds, as time information, time data indicating the start time and end time of workpiece processing within the production device 101. The workpiece processing start time included in the time data is information indicating the date and time when workpiece processing within the production device 101 started, i.e., information indicating the date and time when the production device 101 began processing the workpiece. The workpiece processing end time included in the time data is information indicating the date and time when workpiece processing within the production device 101 ended, i.e., information indicating the date and time when workpiece processing within the production device 101 was completed. Note that the time from the end time of processing the first workpiece to the start time of processing the second workpiece is the wait time for processing to begin on the second workpiece.

The production device 101 also individually stores various information as device data, such as sensor on/off status, sensor measurement values, drive shaft position, motor torque, captured images, signal tower display color, 2D (Dimensions) code reader reading results, device status, and device status detail number. An example of a 2D code reader reading result is a work ID (Identification) that identifies the work. The device data in embodiment 1 is associated with time information.

Network 102 is a communications network for transmitting various device data from production equipment 101 to a data acquisition unit 200 located above the production line 20 in the overall system 1. For ease of understanding, network 102 is depicted as a solid line in Figure 1, but network 102 may be connected via a wired or wireless connection.

The data acquisition unit 200 is located above the production line 20 in the overall system 1 and has the function of acquiring various equipment data from the production equipment 101. The data acquisition unit 200 is composed of a computing device such as a personal computer or a PLC (Programmable Logic Controller), and a storage medium such as a database. The data acquisition unit 200 has a production performance acquisition unit 201 and an equipment status acquisition unit 202, which are functional units realized by the computing device. The data acquisition unit 200 also has a communication unit (not shown) that communicates between the production equipment 101 and the data calculation unit 300.

The production performance acquisition unit 201 acquires and stores equipment data relating to the production history of the production equipment 101 from the production equipment 101. The equipment data relating to the production history of the production equipment 101 is data on the history of workpiece processing that is acquired when the production equipment 101 processes workpieces. In other words, the production performance acquisition unit 201 acquires various types of equipment data from the production equipment 101 as equipment data relating to the production history of the production equipment 101, such as data on the workpiece processing start time, data on the workpiece processing end time, and data on the pass/fail judgment of the workpiece processing. The equipment data relating to the production history of the production equipment 101 acquired from the production equipment 101 is used to calculate the production KPIs of the production line 20.

The production KPI is an operational indicator related to production and is an evaluation indicator that serves as a criterion for determining whether the operational status of the production line 20 is good or bad. The higher the value of the production KPI, the better the operational status of the production line 20.

The equipment status acquisition unit 202 acquires and stores equipment data relating to the equipment history of the production equipment 101 from the production equipment 101. The equipment data relating to the equipment history of the production equipment 101 is data relating to the history of the production equipment status acquired when workpieces are processed on the production equipment 101. In other words, the equipment status acquisition unit 202 acquires various equipment data from the production equipment 101 as equipment data relating to the equipment history of the production equipment 101, such as the display color of the signal tower, operation mode, and other sensor information. The equipment data relating to the equipment history of the production equipment 101 acquired from the production equipment 101 is used to calculate line stoppage factors, which are factors that cause the production line 20 to stop, or to analyze the causes of the line stoppage factors.

The data calculation unit 300 is located above the data acquisition unit 200 in the overall system 1 and has the function of quantifying production KPIs and the function of quantifying line stoppage factors. The data calculation unit 300 is composed of a calculation device such as a personal computer or programmable logic controller, and a storage medium such as a database. The data calculation unit 300 has a production KPI calculation unit 301 and a stoppage factor calculation unit 302. The data calculation unit 300 also has a communication unit (not shown) that communicates between the data acquisition unit 200 and the output unit 400.

The production KPI calculation unit 301 has the function of calculating and quantifying production KPIs using various equipment data acquired and stored by the data acquisition unit 200. Specifically, the production KPI calculation unit 301 calculates the value of the production KPI using equipment data related to the production history of the production equipment 101 acquired by the production performance acquisition unit 201 of the data acquisition unit 200, and quantifies the production KPI.

The production KPI is calculated as a percentage, with 100% being the ideal state, and both the numerator and denominator in the calculation formula are converted to time. For example, if the production KPI is overall equipment effectiveness, the definition formula for the production KPI is "number of good products x line takt / operating time." Using this definition formula, the operating status of the production line 20 can be easily understood.

Overall equipment effectiveness quantifies the extent to which equipment is actually operating within its operating schedule relative to its predetermined design efficiency.

Line takt is the time it takes to produce one product of the same type. Therefore, the unit of "number of good products x line takt" in the above "number of good products x line takt / operating time" is time. Furthermore, the line takt used is the target line takt, which is the line takt targeted for production line 20.

The number of non-defective products is the value obtained by counting the information indicating that the workpiece was successfully processed in the pass/fail judgment results of the device data related to the production history of the production device 101, without duplicating workpiece IDs. For example, it is the value obtained by counting the information indicating "O" in the pass/fail judgment results in Figure 3 described below, without duplicating workpiece IDs.

Furthermore, when the production KPI is the availability rate, the definition formula for the production KPI is "number of processes x line takt / equipment load time." By using this definition formula, the operating status of the production line 20 can be easily grasped. The unit of "number of processes x line takt" in "number of processes x line takt / equipment load time" is time.

Equipment load time is the time that equipment must be operating, including equipment downtime such as breakdown time and setup time.

Availability represents the operating efficiency of equipment and quantifies the percentage of time that equipment is operating normally when needed. Availability is an indicator that should always aim for 100%. In the equation "number of processes x line takt / equipment load time," "number of processes x line takt" corresponds to the "time that equipment is operating normally."

The stoppage cause calculation unit 302 has the function of calculating the value of the line stoppage cause and quantifying the line stoppage cause using various equipment data acquired and stored by the data acquisition unit 200. Specifically, the stoppage cause calculation unit 302 calculates the value of the line stoppage cause using equipment data related to the equipment history of the production equipment 101 acquired by the equipment status acquisition unit 202 of the data acquisition unit 200, quantifies the line stoppage cause, and enables causal analysis of the line stoppage cause.

As described above, line stoppage factors correspond to the factors that caused the production equipment 101 to stop, and are factors that affect the decline in the production KPI of the production line 20. The lower the value of the line stoppage factor, the better the operating status of the production line 20. The value of the line stoppage factor is calculated as a ratio with the same denominator as the production KPI.

When the production KPI is overall equipment effectiveness, the value of the stoppage factor is defined as "duration time / operating time." In this case, the denominator used to calculate the value of the stoppage factor is the same as the denominator in the formula "number of good products x line takt / operating time" used to calculate the value of the production KPI when the above production KPI is overall equipment effectiveness.

When the production KPI is the availability rate, the value of the stoppage factor is defined as "duration time / equipment load time." In this case, the denominator used to calculate the value of the stoppage factor is the same as the denominator in the formula "number of processes x line takt / equipment load time" used to calculate the value of the production KPI when the above production KPI is the availability rate.

The duration is the time that the production device 101 remains stopped.

Operating time is the difference between the maximum processing end time and the start time of the production site to which production line 20 belongs. The maximum processing end time is the latest of the processing end times of each of the multiple production devices 101 on production line 20. For example, if the start time is 8:30, the processing end time of production device 101 (equipment A) is 17:15, the processing end time of production device 101 (equipment B) is 17:20, and the processing end time of production device 101 (equipment C) is 17:30, then 17:30 - 8:30 = 9 hours is defined as operating time.

The shutdown cause calculation unit 302 calculates the value of the unaggregated shutdown cause for the production KPI value and shutdown cause value calculated as above. The value of the unaggregated shutdown cause is defined as "100% - production KPI value (%) - total value of multiple shutdown cause values (%)." In other words, the value of the unaggregated shutdown cause is the value remaining after subtracting the production KPI value (%) and the multiple shutdown cause values (%) from 100%.

An uncounted stoppage factor is a stoppage factor that has not been counted and for which the value of the line stoppage factor has not been calculated when calculating the value of the line stoppage factor using equipment data related to the equipment history of the production equipment 101 acquired by the equipment status acquisition unit 202 of the data acquisition unit 200.

The data calculation unit 300 is realized, for example, by a personal computer. That is, the functions of the production KPI calculation unit 301 and the stoppage cause calculation unit 302 are realized, for example, by a personal computer in which calculation software or a business intelligence (BI) tool is installed.

As described above, production system 10 calculates the production KPI value, the stoppage factor value, and the unaggregated stoppage factor value as a ratio using the same value for the denominator. That is, production system 10 calculates the production KPI value, the stoppage factor value, and the unaggregated stoppage factor value as values per unit time for the same value for the same type of time related to production on production line 20. This makes it easy for production system 10 to compare the magnitude relationship between the stoppage factor value and the unaggregated stoppage factor value, as well as the degree of impact that the stoppage factor and the unaggregated stoppage factor have on a decline in the production KPI. This allows workers or managers to quantitatively understand the cause of a decline in the production KPI and the degree of impact that cause has on the decline in the production KPI.

The output unit 400 is a display unit that can visualize information and present it to workers on the production site or managers in the production management department, and is a display device that can visualize information, such as a monitor, tablet, or wearable device. The output unit 400 can display the trends or numerical values of quantified production KPIs, stoppage factors, and unaggregated stoppage factors in graphs or tables. This allows workers or managers to quantitatively grasp the operating status of the production line 20.

Next, we will explain the operation status analysis process in which the production system 10 quantitatively analyzes and visualizes the operation status of the production line 20. Figure 2 is a flowchart showing the processing steps of the operation status analysis process by the production system 10 according to the first embodiment. Here, we will explain the production KPI as the availability rate.

In step S110, the data acquisition unit 200 acquires and stores equipment data for the production line 20 when the production device 101 processes the workpiece. That is, the production performance acquisition unit 201 performs a production performance acquisition step of acquiring equipment data related to the production history of the production device 101 from the production device 101 and storing it. In addition, the equipment status acquisition unit 202 performs an equipment status acquisition step of acquiring equipment data related to the equipment history of the production device 101 from the production device 101 while the production device 101 is processing the workpiece and storing it.

Figure 3 is a diagram showing an example of a first table that stores equipment data used to calculate the availability rate, which is a production KPI, in the production system 10 according to the first embodiment. The equipment data shown in Figure 3 is an example of equipment data acquired by the production performance acquisition unit 201, and is an example of equipment data required by the production KPI calculation unit 301 to calculate the availability rate, which is a production KPI. The first table is created by the production performance acquisition unit 201 acquiring equipment data related to the production history of the production equipment 101 and storing it in a predetermined format, and is stored in the production performance acquisition unit 201.

In Figure 3, "equipment ID" is equipment identification information for identifying production equipment 101 in the production line 20, and is identification information uniquely assigned to each of the multiple production equipment 101. "Work ID" is work identification information for identifying the work to be processed by the production equipment 101, and is identification information uniquely assigned to each of the multiple workpieces. "Processing start time" is the time when processing of the workpiece by the production equipment 101 starts. "Processing end time" is the time when processing of the workpiece by the production equipment 101 ends. "Workpiece processing time" is calculated from the workpiece processing start time and workpiece processing end time. "Pass/fail judgment result" is pass/fail information indicating whether processing of the workpiece by the production equipment 101 was successful or unsuccessful.

Figure 4 is a diagram showing an example of a second table storing equipment data used to calculate stoppage causes in the production system 10 according to the first embodiment. The equipment data shown in Figure 4 is an example of equipment data acquired by the equipment status acquisition unit 202, and is an example of equipment data required for the stoppage cause calculation unit 302 to calculate stoppage causes.

In Figure 4, as in Figure 3, "equipment ID" is equipment identification information for identifying production equipment 101 on the production line 20, and is identification information uniquely assigned to each of the multiple production equipment 101. "Equipment Status" is status information indicating the state of the production equipment 101. "Detailed Number" is information indicating the state of the production equipment 101 by further subdividing the equipment status of the production equipment 101, and is detailed information that subdivides the "equipment status." For example, if the "equipment status" is "abnormal" and the detailed number is "46," this means that a problem corresponding to "error number 46" is occurring in the production equipment 101. "Status Start Time" is the time when the state of the production equipment 101 indicated in "Equipment Status" started. "Status End Time" is the time when the state of the production equipment 101 indicated in "Equipment Status" ended. The "duration" of the "equipment status" of the production equipment 101 is calculated from the state start time and state end time.

The error number is identification information that is individually assigned to each of multiple abnormal conditions that can occur in the production device 101 and is used to identify the abnormal condition of the production device 101.

The production performance acquisition unit 201 transmits the information of the first table storing the acquired equipment data to the production KPI calculation unit 301 of the data calculation unit 300. The production KPI calculation unit 301 receives and stores the information of the first table. The equipment status acquisition unit 202 transmits the information of the second table storing the acquired equipment data to the stoppage cause calculation unit 302 of the data calculation unit 300. The stoppage cause calculation unit 302 receives and stores the information of the second table. Then, proceed to step S120.

In step S120, the data calculation unit 300 calculates the production KPI value and the stoppage factor value. Specifically, the production KPI calculation unit 301 of the data calculation unit 300 performs a key performance indicator calculation step to calculate the production KPI value. Furthermore, the stoppage factor calculation unit 302 of the data calculation unit 300 performs a stoppage factor calculation step to calculate the stoppage factor value. As described above, the availability rate, which is a production KPI, is defined as "number of processes × line takt / equipment load time." Furthermore, "line takt" is a value that can be uniquely determined before the production line 20 starts operating. Therefore, the production KPI calculation unit 301 calculates the number of workpieces processed and the equipment load time using equipment data such as that shown in FIG. 3. In other words, the production KPI calculation unit 301 acquires the number of workpieces processed and the equipment load time using equipment data acquired by the production performance acquisition unit 201, such as that shown in FIG. 3.

The "line takt" is input and stored in the data calculation unit 300 before the production line 20 starts operating. The data calculation unit 300 can receive line takt information from outside the data calculation unit 300, for example, via a communication unit. The "number of processes" is obtained by counting the number of rows in the first table in Figure 3 without duplicate work IDs. The "equipment load time" can be calculated by calculating the difference between the first "processing start time" and the last "processing end time" in the first table in Figure 3, not for each individual equipment ID, but for the entire information in the first table.

The value of the stoppage cause can be derived by calculating the "duration for each equipment state" by adding up the differences between the "state start time" and "state end time" for each row in the second table in Figure 4, and then dividing this "duration for each equipment state" by the value of the "equipment load time" used to calculate the "availability rate."

In this way, the production KPI value and the value of the stoppage cause are calculated by dividing them by the same value. In other words, the production KPI value and the value of the stoppage cause are divided by the same time value, and are calculated as values per unit time for the same value of the same type of time related to production on the production line 20. Then, proceed to step S130.

In step S130, the data calculation unit 300 calculates the value of the unaggregated stoppage cause. Specifically, in step S130, the stoppage cause calculation unit 302 calculates the value of the unaggregated stoppage cause. The value of the unaggregated stoppage cause is calculated using the formula "100% - production KPI value (%) - total value of multiple stoppage cause values (%)". Then, proceed to step S140.

In step S140, the shutdown cause calculation unit 302 determines whether the value of the unaggregated shutdown cause calculated in step S130 is a positive value. If the value of the unaggregated shutdown cause calculated in step S130 is a positive value, the answer is Yes in step S140 and the process proceeds to step S150. In this case, the production KPI calculation unit 301 transmits information on the calculated production KPI value to the output unit 400. In addition, the shutdown cause calculation unit 302 transmits information on the calculated shutdown cause value and information on the calculated unaggregated shutdown cause value to the output unit 400.

On the other hand, if the value of the uncounted stoppage cause calculated in step S130 is not a positive value, the answer is No in step S140 and the process proceeds to step S160. Note that if there is overlap in the data to be counted in Figure 4, the value of the uncounted stoppage cause may be negative. Because the production line 20 is configured by connecting production devices 101, the stoppage cause of an adjacent production device 101 may affect the counting of the stoppage cause of that production device 101.

In step S150, an output step is performed, and the output unit 400 displays the calculated production KPI value, the value of the stoppage cause, and the value of the unaggregated stoppage cause. Note that if there are no unaggregated stoppage causes, the output unit 400 does not display the value of the unaggregated stoppage cause.

Figure 5 is a diagram showing a first example of the analysis results of line stoppage factors displayed by the production system 10 according to the first embodiment. The graph in Figure 5 shows the relationship between the line stoppage factors and the values of the line stoppage factors for a specific production device 101. The graph in Figure 5 shows factor 68, factor 33, uncounted factors, factor 5, and factor 11 as line stoppage factors. The output unit 400 displays the values of the stoppage factors and the values of the uncounted stoppage factors, for example, as a bar graph.

The graph in Figure 5 displays the values of the stoppage factors and the unaggregated stoppage factor values as a percentage using the same value for the denominator. This makes it easy to compare the magnitude relationship between the stoppage factor values and the unaggregated stoppage factor values, as well as the degree of impact that the stoppage factors and the unaggregated stoppage factors have on the production KPI. This allows workers or managers to quantitatively understand what caused the decline in production KPI and the degree of impact that cause has on the production KPI.

For example, Figure 5 shows that the outage factor that has the greatest impact on the decline in production KPI is "Factor 68." However, Figure 5 also shows that the outage factor that cannot be aggregated has the third highest value, suggesting that additional investigation or collection of equipment data is necessary.

Figure 6 is a diagram showing a second example of the analysis results of line stoppage factors displayed by the production system 10 according to the first embodiment. The graph in Figure 6 displays the production KPI value, the stoppage factor value, and the unaggregated stoppage factor value on a single graph, and shows the daily changes in the production KPI value, the stoppage factor value, and the unaggregated stoppage factor value. That is, Figure 6 shows the changes in the operating status of the production line 20. In Figure 6, the horizontal axis indicates the day on which the production line 20 stopped, and the vertical axis indicates the production KPI value, the stoppage factor value, and the unaggregated stoppage factor value. The graph in Figure 6 shows factor 33, factor 5, unaggregated factor, factor 17, and factor 24 as line stoppage factors. The output unit 400 displays the changes in the production KPI value, for example, as a line graph. The output unit 400 also displays the changes in the stoppage factor value and the unaggregated stoppage factor value, for example, as a bar graph.

The graph in Figure 6 displays the production KPI value, the stoppage factor value, and the unaggregated stoppage factor value as a percentage using the same value for the denominator. This makes it easy to compare the magnitude relationship between the production KPI value, the stoppage factor value, and the unaggregated stoppage factor value, as well as the degree of impact that the stoppage factor and the unaggregated stoppage factor have on the production KPI. This allows workers or managers to quantitatively understand the cause of the decline in production KPI and the degree of impact that cause has on the production KPI.

In step S160, the shutdown cause calculation unit 302 determines that there are no uncounted shutdown causes and proceeds to step S150. In this case, the production KPI calculation unit 301 transmits information on the calculated production KPI values to the output unit 400. The shutdown cause calculation unit 302 also transmits information on the calculated shutdown cause values to the output unit 400. The shutdown cause calculation unit 302 also transmits information to the output unit 400 that there are no uncounted shutdown causes.

The data acquisition unit 200 may be provided in the production device 101. The data acquisition unit 200 and the data calculation unit 300 may also be provided in the production device 101.

The production system 10 described above realizes a production system that analyzes the causes of stoppages in a production line that includes multiple production devices. The production system includes a production history acquisition unit that acquires equipment data related to the production history of each production device from the multiple production devices; an equipment status acquisition unit that acquires equipment data including the equipment status of each production device from the multiple production devices; a key performance indicator calculation unit that calculates the value of a key performance indicator, which is a criterion for determining the operating status of the production line, based on the equipment data related to the production history of each production device; a stoppage factor calculation unit that calculates the values of multiple stoppage factors that affect the decrease of the key performance indicators based on the equipment data including the equipment status of each production device; and an output unit that displays the values of the key performance indicators and the stoppage factors. The stoppage factor calculation unit calculates the values of the key performance indicators and the stoppage factors as values per unit time for the same value of time for the same type of production related to the production on the production line.

As described above, in the production system 10 according to the first embodiment, the production KPI is calculated by the production KPI calculation unit 301 based on the equipment data relating to the production history of the production equipment 101 acquired by the production performance acquisition unit 201. The production KPI value is calculated as a percentage where 100% is the ideal state, and both the numerator and denominator of the calculation formula for calculating the production KPI value are converted to time.

In addition, in the production system 10, the value of each stoppage cause is calculated by the stoppage cause calculation unit 302 based on the equipment data related to the equipment history of the production equipment 101 acquired by the equipment status acquisition unit 202. The value of each stoppage cause is calculated by calculating the duration of each stoppage cause and dividing the duration by the denominator time used when calculating the production KPI.

The value of the unaggregated stoppage factors is calculated as "100% - production KPI value (%) - total value (%) of each stoppage factor value." The production KPI value, the value of each stoppage factor, and the value of the unaggregated stoppage factors are then displayed on the output unit 400 and presented to the worker or manager.

In this way, production system 10 calculates the production KPI value and the value of the stoppage factor as a ratio using the same value for the denominator, making it possible to quantitatively visualize how much each of multiple stoppage factors has reduced the production KPI. Furthermore, because production system 10 can quantify how much each of multiple stoppage factors has reduced the production KPI, the value of unaggregated stoppage factors that reduce the production KPI but have not been aggregated is also calculated as a ratio using the same value for the production KPI value and the value of the stoppage factor for the denominator, making it possible to quantitatively visualize how many unaggregated stoppage factors there are.

This allows the production system 10 to easily compare how much each of multiple stoppage factors affects a decline in production KPIs. In other words, the production system 10 allows easy comparison of the degree of impact of multiple stoppage factors on a decline in production KPIs. This allows production site workers or managers to address each stoppage factor efficiently and in the correct order to improve production KPIs, thereby increasing the efficiency of improvement activities for the production line 20.

Therefore, the production system 10 of embodiment 1 has the effect of being able to quantify and present multiple stoppage factors that reduce the production KPI of the production line 20.

Embodiment 2.
In the second embodiment, another operation status analysis process performed by the production system 10 according to the first embodiment will be described. Fig. 7 is a flowchart showing the processing procedure of another operation status analysis process performed by the production system 10 according to the first embodiment. The following describes the parts that differ from the flowchart in Fig. 2 described above.

In other operation status analysis processes, the data acquired by the equipment status acquisition unit 202 is corrected based on the data acquired by the production performance acquisition unit 201, and processing similar to that shown in the flowchart in Figure 2 is then performed.

In other operational status analysis processes, for example, the "equipment ID" shown in Figures 3 and 4 is used as a search key to exclude equipment states in Figure 4 that do not fall between the "processing start time" and "processing end time" in Figure 3. Because the production line 20 is composed of connected production equipment 101, the cause of a stoppage of an adjacent equipment may affect the total count for that equipment. To eliminate such effects, other operational status analysis processes count stoppage causes only during production processing.

For example, in the second table shown in Figure 4, the "Device Status" of the production device 101 whose "Device ID" is "JROC" is "Stopped." In this case, data in the device data for that production device 101 where the "Device Status" is "Stopped" is considered to be a case where production processing at that production device 101 has been stopped due to a stoppage cause of a production device 101 other than that production device 101. In this case, data corresponding to a "Device Status" of "Stopped" is excluded and the value of the stoppage cause is calculated.

In step S210, the production KPI calculation unit 301 of the data calculation unit 300 calculates the value of the production KPI.

In step S220, the stop factor calculation unit 302 of the data calculation unit 300 counts stop factors limited to data during production processing and calculates the value of the stop factor limited to data during production processing. That is, the stop factor calculation unit 302 obtains information on the "processing start time" and "processing end time" corresponding to the "equipment ID" in the first table of Figure 3, and determines the time from the "processing start time" to the "processing end time" to be the "time during production processing" for that "equipment ID."

Next, the stop cause calculation unit 302 calculates the value of the stop cause using the second table in Figure 4. At this time, the stop cause calculation unit 302 searches the second table in Figure 4 using the "equipment ID" for which the "time during production processing" was determined as a search key, and calculates the value of the stop cause using only the equipment data for which the time from the "state start time" to the "state end time" in the second table in Figure 4 is included in the "time during production processing" determined for that "equipment ID." In other words, in calculating the value of the stop cause, the stop cause calculation unit 302 excludes time outside of production processing of the production equipment 101 from the time of the stop cause, and tallies only stop causes for stoppages caused by the production equipment 101 that occurred during production processing of the production equipment 101.

In the case of the other operation status analysis process in the second embodiment described above, it is possible to prevent duplicate counting of stoppage causes for multiple production devices 101. Therefore, in the case of the other operation status analysis process in the second embodiment, it is possible to calculate the value of the stoppage cause more accurately than in the case of the operation status analysis process in the first embodiment described above. As a result, workers or managers can address the stoppage causes in a more accurate order by referring to the production KPI values calculated by the other operation status analysis process, the stoppage cause values, and the unaggregated stoppage cause values, thereby increasing the efficiency of improvement activities for the production line 20.

Embodiment 3.
Next, a third embodiment will be described with reference to Figures 8 to 11. In the third embodiment, a machine learning function suitable for cases where various indicators cannot be expressed by simple formulas, such as when the production KPI is "production volume" which is difficult to express in terms of time, or when it is difficult to perform calculations to eliminate duplicates of stoppage causes as described in the second embodiment, will be described.

<Learning phase>
8 is a diagram showing the configuration of a learning device 50 according to the third embodiment. The learning device 50 is a computer that learns the degree of influence of a plurality of stoppage factors on a decline in production KPI in an input state.

The learning device 50 comprises a data acquisition unit 51 and a model generation unit 52.

The data acquisition unit 51 acquires behavioral data and status data as learning data. The data acquisition unit 51 acquires behavioral data and status data from the production system 10 that analyzes the causes of stoppages of a production line 20 that includes multiple production devices 101.

The behavioral data acquired by the data acquisition unit 51 is behavioral data indicating the degree of impact of multiple stoppage factors on the decline in production KPI. The status data acquired by the data acquisition unit 51 is status data such as the production KPI value, equipment status data for each production device 101, and data on the processing status of each workpiece. The equipment status data for each production device 101 is equipment data related to the equipment history of the production device 101 described above, and is equipment data related to the status of each production device 101 as shown in Figure 4. The processing status data for each workpiece is equipment data related to the production history of the production device 101 described above, and is equipment data related to the processing history of each workpiece as shown in Figure 3. Note that the equipment status data for each production device 101 is not limited to the equipment data shown in Figure 4. Furthermore, the processing status data for each workpiece is not limited to the equipment data shown in Figure 3.

The model generation unit 52 learns the degree of impact of multiple stoppage factors on a decline in the production KPI in an input state based on learning data including the degree of impact of multiple stoppage factors on a decline in the production KPI, as well as data on the production KPI, the equipment status of each production device 101, and the processing status of each workpiece. In other words, the model generation unit 52 generates a trained model that infers the degree of impact of multiple stoppage factors on a decline in the production KPI when data on the production KPI, the equipment status of each production device 101, and the processing status of each workpiece in the production system 10 is input. The degree of impact of multiple stoppage factors on a decline in the production KPI in the input state is the degree of impact of the multiple stoppage factors corresponding to the input status data on a decline in the production KPI. Hereinafter, the degree of impact of multiple stoppage factors on a decline in the production KPI may be referred to as the impact on a decline in the production KPI.

The learning algorithm used by the model generation unit 52 can be any known algorithm, such as supervised learning, unsupervised learning, or reinforcement learning. As an example, we will explain the case where the model generation unit 52 applies reinforcement learning to the learning algorithm. In reinforcement learning, an agent (acting subject) in a certain environment observes the current state (environmental parameters) and decides on the action to take. The environment changes dynamically depending on the agent's actions, and the agent is given a reward in accordance with the environmental changes. The agent repeats this process and learns the course of action that will obtain the greatest reward through a series of actions. Typical reinforcement learning methods include Q-learning and TD-learning. For example, in the case of Q-learning, the general update formula for the action value function Q(s, a) is expressed as the following equation (1):

In formula (1), s _t represents the state of the environment at time t, and a _t represents the action at time t. Action a _t changes the state to s _t +1. _{r t+1} represents the reward obtained due to the change in state, γ represents the discount rate, and α represents the learning coefficient. Note that γ is in the range of 0 < γ ≦ 1, and α is in the range of 0 < α ≦ 1. The degree of influence on the decline in the production KPI becomes action a _t , and the production KPI, the equipment status of each production device 101, and the processing status of each workpiece become state s _t , and the best action a t for state s _t at time _t is learned.

The update formula expressed in equation (1) increases the action value Q if the action value Q of action a with the highest Q value at time t+1 is greater than the action value Q of action a executed at time t, and decreases the action value Q if the opposite is true. In other words, the update formula expressed in equation (1) updates the action value function Q(s, a) so that the action value Q of action a at time t approaches the best action value at time t+1. As a result, the best action value in a certain environment is sequentially propagated to the action value in the previous environment.

As described above, when generating a trained model through reinforcement learning, the model generation unit 52 includes a reward calculation unit 53 and a function update unit 54.

The reward calculation unit 53 calculates reward based on the degree of impact on the decline in the production KPI, the production KPI, the equipment status of each production device 101, and the processing status of each workpiece. The reward calculation unit 53 calculates reward r based on the difference between the theoretical value of the production KPI and the true value of the production KPI. For example, if the difference between the theoretical value of the production KPI and the true value of the production KPI is decreasing, reward r is increased (e.g., a reward of "1" is given), and on the other hand, if the difference between the theoretical value of the production KPI and the true value of the production KPI is increasing, reward r is decreased (e.g., a reward of "-1" is given).

The theoretical value of a production KPI is the value of the production KPI calculated as 100% - Σ (impact on the decline in production KPI x duration of the cause of the shutdown). Σ (impact on the decline in production KPI x duration of the cause of the shutdown) is the sum of the values of (impact on the decline in production KPI x duration of the cause of the shutdown) for each cause of the shutdown.

The true value of the production KPI is the actual value of the production KPI calculated from the equipment data shown in Figure 3, for example. The value to be adopted as the definition of the production KPI can be selected appropriately depending on the operational status of each production site.

The function update unit 54 updates the function for determining the degree of influence on a decline in the production KPI in the input state in accordance with the reward calculated by the reward calculation unit 53, and outputs the updated function as a learned model to the learned model storage unit 70. For example, in the case of Q-learning, the function update unit 54 uses the action value function Q(s _t , a _t ) expressed by equation (1) as a function for calculating the degree of influence on a decline in the production KPI in the input state.

The model generation unit 52 repeatedly executes the above-described learning. The learned model storage unit 70 stores the action value function Q(s _t , a _t ) updated by the function update unit 54, i.e., the learned model.

Next, using Figure 9, we will explain the process by which the learning device 50 learns the degree of impact on a decline in production KPI. Figure 9 is a flowchart showing the processing steps of the learning process by the learning device 50 in embodiment 3.

In step S310, the data acquisition unit 51 acquires the degree of impact on the decline in production KPI, the production KPI, the equipment status of each production device 101, and the processing status of each workpiece as learning data.

In step S320, the model generation unit 52 calculates the reward based on the degree of impact on the decline in the production KPI, the production KPI, the equipment status of each production device 101, and the processing status of each workpiece. Specifically, the reward calculation unit 53 acquires the degree of impact on the decline in the production KPI, the production KPI, the equipment status of each production device 101, and the processing status of each workpiece, and determines whether to increase or decrease the reward based on the difference between the predetermined theoretical value of the production KPI and the true value of the production KPI.

When the remuneration calculation unit 53 determines that the remuneration should be increased (step S320, the difference between the theoretical value of the production KPI and the true value of the production KPI has decreased), the remuneration calculation unit 53 increases the remuneration in step S330. In other words, the remuneration calculation unit 53 increases the remuneration when the remuneration increase criteria are met due to the decrease in the difference between the theoretical value of the production KPI and the true value of the production KPI.

On the other hand, if the remuneration calculation unit 53 determines that the remuneration should be reduced (step S320, the difference between the theoretical value of the production KPI and the true value of the production KPI has increased), it reduces the remuneration in step S340. In other words, the remuneration calculation unit 53 reduces the remuneration when the remuneration reduction criteria are met due to an increase in the difference between the theoretical value of the production KPI and the true value of the production KPI.

In step S350, the function update unit 54 updates the action value function Q(s _t , a _t ) represented by equation (1) stored in the trained model storage unit 70 based on the reward calculated by the reward calculation unit 53.

The learning device 50 repeatedly performs the above steps S310 to S350, and stores the generated action value function Q(s _t , a _t ) as a learned model in the learned model storage unit 70.

Step S310 can be considered a data acquisition step. Steps S320 to S350 can be considered model generation steps. The learning device 50 then repeatedly performs the above data acquisition step and model generation step, updates the action value function Q(s _t , a _t ) represented by equation (1) stored in the trained model storage unit 70, and stores it in the trained model storage unit 70.

In the learning device 50 of embodiment 3, the learned model is stored in a learned model storage unit 70 located outside the learning device 50, but the learned model storage unit 70 may also be located inside the learning device 50.

According to the learning device 50 of the third embodiment, a learning device is realized that includes: a data acquisition unit that acquires learning data from a production system that analyzes factors that cause a production line that includes multiple production devices, including key performance indicators of the production line, the equipment status of the multiple production devices, the processing status of workpieces processed by the production devices, and the degree of impact of stoppage factors on a decline in key performance indicators in terms of key performance indicators, the equipment status of the multiple production devices, and the processing status of workpieces; and a model generation unit that uses the learning data to generate a trained model for inferring the degree of impact of stoppage factors on a decline in key performance indicators from the key performance indicators of the production line, the equipment status of the multiple production devices, and the processing status of workpieces processed by the production devices.

According to the learning device 50 of the third embodiment, a method for generating a trained model is implemented, which includes a data acquisition step of acquiring training data from a production system that analyzes factors that cause a production line including multiple production devices, including key performance indicators of the production line, the equipment status of the multiple production devices, the processing status of workpieces processed by the production devices, and the degree of impact of the stoppage factors on a decline in the key performance indicators in terms of the key performance indicators, the equipment status of the multiple production devices, and the processing status of the workpieces; and a model generation step of using the training data to generate a trained model for inferring the degree of impact of the stoppage factors on a decline in the key performance indicators from the key performance indicators of the production line, the equipment status of the multiple production devices, and the processing status of workpieces processed by the production devices.

According to the learning device 50 of embodiment 3, a method for generating a trained model is implemented in which a new data acquisition step is performed to acquire new training data, and a new model generation step is performed using the new training data, thereby updating the trained model.

According to the learning device 50 of the third embodiment, for a production system that analyzes factors that cause a production line that includes multiple production machines, machine learning is performed using learning data that includes key performance indicators of the production line, the equipment status of the multiple production machines, the processing status of the workpieces processed by the production machines, and the degree of impact of factors that cause a stoppage on a decline in the key performance indicators in the key performance indicators, the equipment status of the multiple production machines, and the processing status of the workpieces, and a learned model is generated that outputs the degree of impact of factors that cause a stoppage on a decline in the key performance indicators from the key performance indicators, the equipment status of the multiple production machines, and the processing status of the workpieces.

<Utilization phase>
10 is a diagram showing a configuration of an inference device 60 according to the third embodiment. The inference device 60 is a computer that uses a trained model to infer the degree of influence on a decline in a production KPI.

The inference device 60 comprises a data acquisition unit 61 and an inference unit 62.

The data acquisition unit 61 acquires production KPIs, the equipment status of each production device 101, and the processing status of each workpiece from the production system 10, which analyzes the causes of stoppages in a production line 20 that includes multiple production devices 101.

The inference unit 62 uses the trained model stored in the trained model storage unit 70 to infer the degree of impact on a decline in the production KPI in the input state. In other words, by inputting the production KPI acquired by the data acquisition unit 61, the equipment status of each production device 101, and the processing status of each workpiece into this trained model, the inference unit 62 can infer the degree of impact on a decline in the production KPI that is appropriate for the input production KPI, the equipment status of each production device 101, and the processing status of each workpiece.

In the third embodiment, the inference device 60 is described as outputting the degree of impact on a decline in the production KPI using a trained model trained by the model generation unit 52 for the production system 10 connected to the production line 20, but the inference device 60 may also acquire trained models trained for other production systems connected to other production lines. In this case, the inference device 60 outputs the degree of impact on a decline in the production KPI using trained models trained for other production systems connected to other production lines.

Next, the process by which the inference device 60 infers the degree of impact on a decline in production KPI will be explained using Figure 11. Figure 11 is a flowchart showing the processing steps of the inference process by the inference device 60 according to the third embodiment.

In step S410, the data acquisition unit 61 acquires data on production KPIs, the equipment status of each production device 101, and the processing status of each workpiece as inference data.

In step S420, the inference unit 62 inputs data on the production KPI, the equipment status of each production device 101, and the processing status of each workpiece into the trained model stored in the trained model storage unit 70, and obtains the degree of impact on the decline in the production KPI corresponding to the input information. The degree of impact on the decline in the production KPI obtained here is output as the degree of impact on the decline in the production KPI for each of multiple stoppage factors. For example, the degree of impact on the decline in the production KPI for each of multiple stoppage factors is output as follows: Stoppage factor A reduced the production KPI by P%, and stoppage factor B reduced the production KPI by Q%. Furthermore, the unit of the degree of impact on the decline in the production KPI is not limited to "%". For example, the number of production units is used as the production KPI, and the degree of impact on the decline in the production KPI may be expressed in units of units, such as stoppage factor A reducing the production number by R units and stoppage factor B reducing the production number by S units. In step S430, the inference unit 62 outputs the obtained degree of influence on the decline in the production KPI to the output unit 400 of the production system 10.

In step S430, the output unit 400 of the production system 10 outputs the degree of impact on the decline in the production KPI sent from the inference unit 62, i.e., displays the degree of impact on the decline in the production KPI, thereby notifying the workers at the production site of the degree of impact on the decline in the production KPI.

In step S440, each production device 101 processes the workpiece under operating conditions of the production device 101 that improve the production KPI of the production line 20. That is, the worker at the production site checks the degree of impact on the decline in the production KPI displayed on the output unit 400, carries out improvement activities to improve the production KPI of the production line 20, determines improved operating conditions that are the operating conditions of the production device 101 that improve the production KPI of the production line 20, and sets the determined improved operating conditions to each production device 101. Each production device 101 processes the workpiece under the newly set improved operating conditions of the production device 101 that improve the production KPI of the production line 20.

According to the inference device 60 of the third embodiment, an inference device is realized that includes: a data acquisition unit that acquires, from a production system that analyzes factors that cause a production line that includes multiple production devices, key performance indicators of the production line, the equipment status of the multiple production devices, and the processing status of workpieces processed by the production devices; and an inference unit that uses a trained model for inferring the degree of impact of factors that cause a stoppage on a decline in key performance indicators based on the key performance indicators of the production line, the equipment status of the multiple production devices, and the processing status of workpieces processed by the production devices, and outputs the degree of impact of factors that cause a stoppage on a decline in key performance indicators from the key performance indicators, the equipment status of the multiple production devices, and the processing status of the workpieces acquired by the data acquisition unit.

By applying the above-described learning device 50 and inference device 60 to the production system 10, the inference device 60 can automatically present to production site workers or managers the degree of impact of the cause of the shutdown on a decline in production KPI, in accordance with the shutdown status of each production device 101. Then, by having production site workers refer to the degree of impact of the cause of the shutdown of the production device 101 on a decline in production KPI, the efficiency of production line improvement activities is increased.

In the third embodiment, we have described a case where reinforcement learning is applied to the learning algorithm used by the inference unit 62, but the learning algorithm is not limited to reinforcement learning. As for the learning algorithm used by the inference unit 62, it is also possible to apply supervised learning, unsupervised learning, semi-supervised learning, etc. in addition to reinforcement learning.

The learning algorithm used in the model generation unit 52 can also be deep learning, which learns to extract features themselves. The model generation unit 52 may also perform machine learning according to other known methods, such as neural networks, genetic programming, functional logic programming, and support vector machines.

The learning device 50 and the inference device 60 may be devices separate from the production system 10, connected to the production system 10 via a network such as the Internet. The learning device 50 and the inference device 60 may also be built into the production system 10. Furthermore, the learning device 50 and the inference device 60 may reside on a cloud server.

The model generation unit 52 may also use learning data acquired from multiple production systems 10 to learn the degree of impact on a decline in the production KPI in the input state. The model generation unit 52 may acquire learning data from multiple production systems 10 used in the same area, or may use learning data collected from multiple production systems 10 operating independently in different areas to learn the degree of impact on a decline in the production KPI in the input state. A production system 10 that collects learning data may be added to or removed from the target at any time. Furthermore, a learning device 50 that has learned the degree of impact on a decline in the production KPI in the input state for a certain production system 10 may be applied to another production system 10, and the degree of impact on a decline in the production KPI in the input state for the other production system 10 may be re-learned and updated.

Here, we will explain the hardware configuration of the learning device 50 and the inference device 60. Since the learning device 50 and the inference device 60 have similar hardware configurations, we will explain the hardware configuration of the learning device 50 here.

The learning device 50 is realized by a processing circuit. The processing circuit may be a processor and memory that executes a program stored in memory, or it may be dedicated hardware such as a dedicated circuit. The processing circuit is also called a control circuit.

12 is a diagram showing an example of the configuration of a processing circuit provided in the learning device 50 according to embodiment 3, when the processing circuit is realized by a processor and memory. The processing circuit 90 shown in FIG. 12 is a control circuit and includes a processor 91 and a memory 92. When the processing circuit 90 is configured with the processor 91 and the memory 92, each function of the processing circuit 90 is realized by software, firmware, or a combination of software and firmware. The software or firmware is written as a program and stored in the memory 92. In the processing circuit 90, each function is realized by the processor 91 reading and executing the program stored in the memory 92. That is, the processing circuit 90 includes a memory 92 for storing a program that results in the processing of the learning device 50 being executed. This program can also be said to be a program that causes the learning device 50 to execute each function realized by the processing circuit 90. This program can be provided by a storage medium on which the program is stored, or by other means such as a communication medium. The above program can also be said to cause the learning device 50 to execute the learning process.

Here, the processor 91 is, for example, a CPU (Central Processing Unit), processing device, arithmetic unit, microprocessor, microcomputer, or DSP (Digital Signal Processor). The processor 91 is included in a PC (Personal Computer) or a PLC. A PLC is also called a sequencer.

In addition, memory 92 may be, for example, a non-volatile or volatile semiconductor memory such as RAM (Random Access Memory), ROM (Read Only Memory), flash memory, EPROM (Erasable Programmable ROM), EEPROM (registered trademark) (Electrically EPROM), a magnetic disk, a flexible disk, an optical disk, a compact disk, a minidisk, or a DVD (Digital Versatile Disc).

Figure 13 is a diagram showing an example of a processing circuit provided in the learning device 50 according to embodiment 3, configured with dedicated hardware. The processing circuit 93 shown in Figure 13 corresponds to, for example, a single circuit, a composite circuit, a programmed processor, a parallel programmed processor, an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or a combination of these. The processing circuit 93 may be partially implemented with dedicated hardware and partially implemented with software or firmware. In this way, the processing circuit 93 can realize each of the above-mentioned functions by dedicated hardware, software, firmware, or a combination of these.

In addition, the production performance acquisition unit 201, the equipment status acquisition unit 202, the production KPI calculation unit 301, and the stoppage cause calculation unit 302 also have the hardware configuration described in Figures 12 and 13.

Embodiment 4.
In the fourth embodiment, a case will be described in which the overall system 1 according to the first embodiment is equipped with a corresponding worker suggestion unit 500. Fig. 14 is a diagram showing the configuration of the overall system 1 including the corresponding worker suggestion unit 500 according to the fourth embodiment.

The corresponding worker suggestion unit 500 is located below the data calculation unit 300 in the overall system 1, and has the function of determining that a line stoppage factor whose rate of change in the value of the line stoppage factor calculated by the stoppage factor calculation unit 302 of the data calculation unit 300 exceeds a predetermined threshold is a stoppage factor to be removed, and automatically determining the corresponding worker from among multiple workers at the production site who is the worker responsible for removing the stoppage factor to be removed.

The corresponding worker suggestion unit 500 includes a current stop cause grasp unit 501, a worker position information acquisition unit 502, and a corresponding worker determination unit 503. The corresponding worker suggestion unit 500 is composed of a computing device such as a personal computer or programmable logic controller, and a storage medium such as a database. The corresponding worker suggestion unit 500 also includes information acquisition devices such as an overhead camera and a beacon receiver. The corresponding worker suggestion unit 500 also includes a communication unit (not shown) that communicates with the data acquisition unit 200 and the output unit 400.

The current stoppage cause grasping unit 501 detects changes in the degree of impact of line stoppage causes on the decline of key performance indicators at any time. The current stoppage cause grasping unit 501 obtains line stoppage cause values from the stoppage cause calculation unit 302 of the data calculation unit 300 for multiple production devices 101, and determines that a line stoppage cause whose rate of change in the value of the line stoppage cause exceeds a preset threshold is a stoppage cause to be removed. The threshold is a threshold that the current stoppage cause grasping unit 501 uses to determine a stoppage cause to be removed by comparing it with the rate of change in the value of the line stoppage cause.

In other words, the current stoppage cause grasping unit 501 has the function of acquiring the value (%) of each line stoppage cause calculated by the stoppage cause calculation unit 302 of the data calculation unit 300 for multiple production devices 101 on the production line 20 from the stoppage cause calculation unit 302 at predetermined intervals, and monitoring the rate of change in the acquired line stoppage cause value.In other words, the current stoppage cause grasping unit 501 has the function of constantly grasping the current line stoppage cause of each production device 101 at the production site.

Then, if the rate of change in the value of the line stop cause exceeds a preset threshold, the current stop cause grasping unit 501 determines that the line stop cause needs to be removed. In other words, if the rate of change in the value of the line stop cause exceeds a preset threshold, the current stop cause grasping unit 501 determines that the line stop cause is a line stop cause to be removed, that is, a line stop cause that needs to be removed.

The rate of change is a value calculated by using a given time as the denominator and the fluctuation value (%) of the line stoppage factor within that given time as the numerator. If the rate of change of a certain line stoppage factor is relatively large compared to the rates of change of other line stoppage factors, it can be said that that line stoppage factor is currently occurring and that the impact of that line stoppage factor on production KPIs is relatively large.

The worker position information acquisition unit 502 determines the positions of multiple workers at a production site where multiple production devices 101 are installed. That is, the worker position information acquisition unit 502 has the function of determining and storing the position of each worker at a production site where a production line 20 is installed. The worker position information acquisition unit 502 has equipment such as an overhead camera mounted on the ceiling of the production site and a beacon receiver for obtaining information on the position of beacons carried by workers. The worker position information acquisition unit 502 determines and stores the position of each worker at the production site, for example, based on images from the overhead camera mounted on the ceiling of the production site. The worker position information acquisition unit 502 also determines and stores the position of each worker at the production site, for example, based on information received by a beacon receiver.

The corresponding worker determination unit 503 determines which of the multiple workers at a production site where multiple production devices 101 are installed is the worker responsible for removing the line stoppage cause to be removed. In other words, the corresponding worker determination unit 503 has the function of determining and suggesting which of the workers at the production site should be assigned to remove the line stoppage cause that is currently detected by the stoppage cause grasping unit 501 and is causing a decrease in production KPI. The corresponding worker determination unit 503 holds location information of the production devices 101 and stoppage cause attribute information.

The location information of the production device 101 is used to calculate the distance between the production device 101 and each worker by combining it with the location information of the worker acquired by the worker location information acquisition unit 502. That is, the corresponding worker determination unit 503 acquires the location information of each worker from the worker location information acquisition unit 502. The corresponding worker determination unit 503 then calculates the distance between the production device 101 and each worker based on the location information of each worker and the location information of the production device 101 that it holds.

The stoppage cause attribute information stores data indicating which production device 101 each line stoppage cause belongs to. Each line stoppage cause is also linked to the hat color photographed by the overhead camera and the worker number set on the beacon. In other words, the stoppage cause attribute information contains correspondence information between hat color and line stoppage cause, indicating which worker with which hat color is responsible for each line stoppage cause. The stoppage cause attribute information also contains correspondence information between worker numbers set on the beacon and line stoppage cause, indicating which worker is responsible for each line stoppage cause.

Next, we will explain the operation of the corresponding worker suggestion unit 500.

While the production line 20 is operating, the current stoppage cause grasping unit 501 acquires the value of each line stoppage cause from the data calculation unit 300 and constantly monitors the rate of change of the value of each line stoppage cause. While monitoring the rate of change of the value of a line stoppage cause, if the rate of change of the value of the line stoppage cause exceeds a preset threshold, the current stoppage cause grasping unit 501 determines that the line stoppage cause in question needs to be removed. In other words, if the rate of change of the value of a line stoppage cause exceeds a preset threshold, the current stoppage cause grasping unit 501 determines that the line stoppage cause in question is a line stoppage cause to be removed.

The corresponding worker determination unit 503 identifies which production equipment 101 is causing the line stoppage cause that has been determined to need to be removed, i.e., the line stoppage cause to be removed. In other words, the corresponding worker determination unit 503 identifies which production equipment 101 is causing the line stoppage cause to be removed, based on the stoppage cause attribute information it holds.

The corresponding worker determination unit 503 also identifies a worker who can respond to removing the cause of line stoppage based on the stored stoppage cause attribute information, and calculates the distance between the identified worker and the production equipment 101 for each identified worker. To calculate the distance between the worker and the production equipment 101, the corresponding worker determination unit 503 obtains information on the worker's location at the time of calculation from the worker location information acquisition unit 502 and uses this information. The corresponding worker determination unit 503 determines the worker with the shortest calculated distance as the corresponding worker who should respond to removing the cause of line stoppage.

The corresponding worker suggestion unit 500 displays on the output unit 400 the name of the worker who should be responsible for removing the cause of the line stoppage. For example, if the output unit 400 is composed of a large monitor and a wearable device, the corresponding worker suggestion unit 500 displays the cause of the line stoppage that is occurring and the name of the worker who should be responsible for removing that cause of the line stoppage side by side. The corresponding worker suggestion unit 500 also notifies the worker that he or she will be instructed to move to the production equipment 101 where the cause of the stoppage is occurring. Methods for notifying the worker include, for example, vibrating the worker's wearable device in a predetermined pattern, and playing a predetermined sound on the worker's wearable device.

The cause of the line stoppage and the worker name displayed on the large monitor of the output unit 400 can be deleted at any time, such as when the cause of the line stoppage is resolved or when the worker's wearable device notifies the corresponding worker suggestion unit 500 that they are unable to respond to the removal of the cause of the line stoppage. If the cause of the line stoppage continues and only the worker name has been removed from the display on the large monitor, the corresponding worker determination unit 503 will perform a re-determination and determine an alternative corresponding worker, excluding the worker from the workers who should respond to the removal of the cause of the line stoppage.

The overall system 1 equipped with such a corresponding worker suggestion unit 500 can automatically suggest the workers who should respond to each line stoppage factor that worsens production KPI at any time, making it possible to reduce the duration of each line stoppage factor.

The configurations shown in the above embodiments are merely examples, and may be combined with other known technologies, or different embodiments may be combined with each other. Parts of the configuration may also be omitted or modified without departing from the spirit of the invention.

The various aspects of this disclosure are summarized below as appendices.

(Appendix 1)
A production line analysis system for analyzing stoppage causes of a production line including a plurality of production devices,
a production performance acquisition unit that acquires equipment data relating to the production history of each of the production equipments from the plurality of production equipments;
a device status acquisition unit that acquires device data including a device status for each of the production devices from the plurality of production devices;
a key performance indicator calculation unit that calculates a value of a key performance indicator, which is a criterion for determining whether the operating status of the production line is good or bad, based on equipment data related to the production history of each of the production equipment;
a stoppage factor calculation unit that calculates values of a plurality of stoppage factors that affect a decrease in the key performance indicator based on equipment data including the equipment status of each of the production equipment;
an output unit that displays the value of the key performance indicator and the value of the stoppage cause;
Equipped with
the stoppage factor calculation unit calculates the value of the key performance indicator and the value of the stoppage factor as values per unit time for the same value of the same type of time related to production on the production line;
A production line analysis system characterized by:
(Appendix 2)
the stoppage cause calculation unit calculates the value of an unaggregated stoppage cause for which the value of the stoppage cause has not been calculated based on the equipment data including the equipment status of each of the production equipment, using a calculation formula of "100 - key performance indicator - total value of stoppage causes";
2. The production line analysis system according to claim 1,
(Appendix 3)
the output unit displays the value of the stop cause and the value of the unaggregated stop cause on the same graph;
3. The production line analysis system according to claim 2,
(Appendix 4)
the output unit displays the values of the key performance indicators, the values of the stoppage factors, and the unaggregated values of the stoppage factors on the same graph;
3. The production line analysis system according to claim 2,
(Appendix 5)
The equipment data relating to the production history of each production equipment is
Equipment identification information for identifying the production equipment;
workpiece identification information for identifying a workpiece to be machined by the production device;
A start time of processing the workpiece by the production device;
A time when the processing of the workpiece by the production device is completed;
Pass/fail information indicating whether the processing of the workpiece by the production device was successful or unsuccessful; and
Including,
The equipment data including the equipment status for each production equipment is
Equipment identification information for identifying the production equipment;
Status information indicating the status of the production equipment;
Detailed information on the device status;
a state start time, which is the time when the state of the production equipment indicated in the equipment state is started;
a state end time, which is the time when the state of the production equipment indicated in the equipment state ends;
containing,
5. The production line analysis system according to claim 1, wherein:
(Appendix 6)
The definition formula for the key performance indicator is line takt x number of processes / equipment load time,
the line tact time is a target tact time that is a target for the production line,
the number of processed products is a value obtained by counting the workpiece identification information without duplication in the device data including the device status for each of the production devices,
The equipment load time is the difference between the first processing start time and the last processing end time for all of the equipment data related to the production history of each of the production equipment.
6. The production line analysis system according to claim 5,
(Appendix 7)
The definition formula of the key performance indicator is the number of non-defective products × line takt/operation time,
the line tact time is a target tact time that is a target for the production line,
The number of non-defective products is a value obtained by counting information indicating that the workpieces have been successfully processed in the pass/fail information of the device data relating to the production history of each of the production devices, without duplication of the workpiece identification information,
The operating time is the difference between the maximum processing end time and the start time of the production site to which the production line belongs;
6. The production line analysis system according to claim 5,
(Appendix 8)
the stoppage cause calculation unit excludes time outside of production processing of the production device in calculating the value of the stoppage cause;
8. The production line analysis system according to any one of claims 1 to 7,
(Appendix 9)
A production line analysis method for analyzing stoppage causes of a production line including a plurality of production devices, comprising:
a production performance acquisition step of acquiring device data relating to the production history of each of the production devices from the plurality of production devices;
an apparatus status acquisition step of acquiring apparatus data including an apparatus status for each of the production apparatuses from the plurality of production apparatuses;
a key performance indicator calculation step of calculating a value of a key performance indicator, which is a criterion for determining whether the operating status of the production line is good or bad, based on the equipment data relating to the production history of each of the production equipment;
a stoppage factor calculation step of calculating values of a plurality of stoppage factors that affect a decrease in the key performance indicator based on equipment data including the equipment status of each of the production equipment;
an output step of displaying the values of the key performance indicators and the values of the stoppage factors;
Including,
the stoppage factor calculation step calculates the value of the key performance indicator and the value of the stoppage factor as values per unit time for the same value of the same type of time related to production on the production line;
A production line analysis method characterized by:
(Appendix 10)
a data acquisition unit that acquires learning data from a production line analysis system that analyzes factors that cause a production line that includes a plurality of production devices, including key performance indicators of the production line, the equipment status of the plurality of production devices, the processing status of workpieces processed by the production devices, and the degree of influence of the factors that cause a production line to stop on a decrease in the key performance indicators in the equipment status of the plurality of production devices and the processing status of the workpieces;
a model generation unit that generates a trained model for inferring the degree of influence of the stoppage factors on a decrease in key performance indicators from key performance indicators of the production line, the device states of the plurality of production devices, and the processing status of workpieces processed by the production devices, using the training data; and
A learning device comprising:
(Appendix 11)
a data acquisition unit that acquires, from a production line analysis system that analyzes factors that cause a production line including a plurality of production devices to stop, key performance indicators of the production line, the device states of the plurality of production devices, and the processing status of workpieces processed by the production devices;
an inference unit that uses a trained model for inferring the degree of influence of the stoppage factor on a decline in a key performance indicator of the production line, the equipment states of the plurality of production devices, and the processing status of the workpieces processed by the production devices, and outputs the degree of influence of the stoppage factor on a decline in a key performance indicator from the key performance indicator, the equipment states of the plurality of production devices, and the processing status of the workpieces acquired by the data acquisition unit;
An inference device comprising:

1 Overall system, 10 Production system, 20 Production line, 50 Learning device, 51, 61, 200 Data acquisition unit, 52 Model generation unit, 53 Reward calculation unit, 54 Function update unit, 60 Inference device, 62 Inference unit, 70 Learned model storage unit, 90, 93 Processing circuit, 91 Processor, 92 Memory, 101 Production equipment, 102 Network, 201 Production performance acquisition unit, 202 Equipment status acquisition unit, 300 Data calculation unit, 301 Production KPI calculation unit, 302 Stoppage cause calculation unit, 400 Output unit, 500 Responding worker suggestion unit, 501 Current stoppage cause grasp unit, 502 Worker position information acquisition unit, 503 Responding worker determination unit, 1011 Production equipment control unit.

Claims (16)

A production system that analyzes causes of stoppages of a production line including a plurality of production devices,
a production performance acquisition unit that acquires equipment data relating to the production history of each of the production equipments from the plurality of production equipments;
a device status acquisition unit that acquires, from the plurality of production devices, device data including device statuses of the production devices, the device statuses including identification information for identifying abnormal states obtained by subdividing the device statuses of the production devices;
a key performance indicator calculation unit that calculates a value of a key performance indicator, which is a criterion for determining whether the operating status of the production line is good or bad, based on equipment data related to the production history of each of the production equipment;
a stoppage factor calculation unit that calculates values of a plurality of stoppage factors that affect a decrease in the key performance indicator based on equipment data including the equipment status of each of the production equipment;
an output unit that displays the value of the key performance indicator and the value of the stoppage cause;
Equipped with
The stop cause calculation unit
Calculating the value of the key performance indicator and the value of the stoppage factor as values per unit time for the same value of the same type of time related to production on the production line;
calculating the value of the stoppage cause that has not yet been calculated based on the equipment data including the equipment status of each of the production equipment;
A production system characterized by: the stop factor calculation unit calculates the value of the unaggregated stop factor using a calculation formula of "100 - key performance indicator - total value of stop factors";
The production system according to claim 1 . the output unit displays the value of the stop cause and the value of the unaggregated stop cause on the same graph;
The production system according to claim 2 . the output unit displays the values of the key performance indicators, the values of the stoppage factors, and the unaggregated values of the stoppage factors on the same graph;
The production system according to claim 2 . The equipment data relating to the production history of each production equipment is
Equipment identification information for identifying the production equipment;
workpiece identification information for identifying a workpiece to be machined by the production device;
A start time of processing the workpiece by the production device;
A time when the processing of the workpiece by the production device is completed;
Pass/fail information indicating whether the processing of the workpiece by the production device was successful or unsuccessful; and
Including,
The equipment data including the equipment status for each production equipment is
Equipment identification information for identifying the production equipment;
Status information indicating the status of the production equipment;
Detailed information on subdivided device states including identification information for identifying abnormal states obtained by subdividing the device states of the production devices;
a state start time, which is the time when the state of the production equipment indicated in the equipment state is started;
a state end time, which is the time when the state of the production equipment indicated in the equipment state ends;
containing,
The production system according to claim 1 . The definition formula for the key performance indicator is line takt x number of processes / equipment load time,
the line tact is a target line tact that is a target for the production line,
the number of processed products is a value obtained by counting the workpiece identification information in the device data relating to the production history of each of the production devices without duplication,
The equipment load time is the difference between the first processing start time and the last processing end time for all of the equipment data related to the production history of each of the production equipment.
The production system according to claim 5 . The definition formula of the key performance indicator is the number of non-defective products × line takt/operation time,
the line tact is a target line tact that is a target for the production line,
The number of non-defective products is a value obtained by counting information indicating that the workpieces have been successfully processed in the pass/fail information of the device data relating to the production history of each of the production devices, without duplication of the workpiece identification information,
The operating time is the difference between the maximum processing end time and the start time of the production site to which the production line belongs;
The production system according to claim 5, the stoppage cause calculation unit excludes time outside of production processing of the production device in calculating the value of the stoppage cause;
The production system according to claim 1 . a current stoppage cause grasping unit that acquires values of the stoppage cause for the plurality of production devices from the stoppage cause calculation unit and determines that a stoppage cause whose change rate exceeds a predetermined threshold is a stoppage cause to be removed;
a worker position information acquisition unit that determines the positions of a plurality of workers present at a production site where a plurality of the production devices are installed;
a corresponding worker determination unit that determines a corresponding worker who is a worker responsible for removing the stop cause to be removed, among the plurality of workers;
9. The production system according to claim 1, further comprising: calculating distances between the production equipment and the plurality of workers based on stoppage cause attribute information indicating which production equipment the stoppage cause belongs to and location information of the production equipment, and determining the worker with the shortest calculated distance as the corresponding worker;
The production system according to claim 9, A production line analysis method for analyzing stoppage causes of a production line including a plurality of production devices, comprising:
a production performance acquisition step of acquiring device data relating to the production history of each of the production devices from the plurality of production devices;
an apparatus status acquisition step of acquiring apparatus data from the plurality of production apparatuses, the apparatus data including the apparatus status of each of the production apparatuses, the apparatus status including identification information for identifying abnormal states obtained by subdividing the apparatus status of the production apparatuses;
a key performance indicator calculation step of calculating a value of a key performance indicator, which is a criterion for determining whether the operating status of the production line is good or bad, based on the equipment data relating to the production history of each of the production equipment;
a stoppage factor calculation step of calculating values of a plurality of stoppage factors that affect a decrease in the key performance indicator based on equipment data including the equipment status of each of the production equipment;
an output step of displaying the values of the key performance indicators and the values of the stoppage factors;
Including,
In the stop cause calculation step,
Calculating the value of the key performance indicator and the value of the stoppage factor as values per unit time for the same value of the same type of time related to production on the production line;
calculating the value of the stoppage cause that has not yet been calculated based on the equipment data including the equipment status of each of the production equipment;
A production line analysis method characterized by: a data acquisition unit that acquires learning data from a production system that analyzes factors that cause a production line that includes a plurality of production devices, the learning data including key performance indicators that are criteria for determining whether the operating status of the production line is good or bad , the equipment status of the plurality of production devices including identification information for identifying abnormal conditions that subdivide the equipment status of the production devices, the processing status of workpieces processed by the production devices, and the degree of influence of the factors that cause the line to stop on a decrease in the key performance indicators in the equipment status of the plurality of production devices and the processing status of the workpieces;
a model generation unit that generates a trained model for inferring the degree of influence of the stoppage factors on a decrease in key performance indicators from key performance indicators of the production line, the device states of the plurality of production devices, and the processing status of workpieces processed by the production devices, using the training data; and
Equipped with
The trained model outputs the degree of impact of the stop factor on a decline in a key performance indicator for each of the plurality of stop factors;
A learning device characterized by: a data acquisition unit that acquires, from a production system that analyzes factors that cause a production line including a plurality of production devices to stop, key performance indicators that are criteria for determining whether the operating status of the production line is good or bad , the device statuses of the plurality of production devices including identification information for identifying abnormal states that subdivide the device statuses of the production devices, and the processing statuses of workpieces processed by the production devices;
an inference unit that uses a trained model for inferring the degree of influence of the stoppage factor on a decline in a key performance indicator of the production line, the equipment states of the plurality of production devices, and the processing status of the workpieces processed by the production devices, and outputs the degree of influence of the stoppage factor on a decline in a key performance indicator from the key performance indicator, the equipment states of the plurality of production devices, and the processing status of the workpieces acquired by the data acquisition unit;
Equipped with
The trained model outputs the degree of impact of the stop factor on a decline in a key performance indicator for each of the plurality of stop factors;
An inference device characterized by: The data is generated by performing machine learning using learning data including key performance indicators, which are criteria for determining whether the operating status of a production line including a plurality of production devices is good or bad , device states of the plurality of production devices including identification information for identifying abnormal states obtained by subdividing the device states of the production devices, processing states of workpieces processed by the production devices, and the degree of influence of the stoppage factors on a decrease in the key performance indicators in the device states of the plurality of production devices and the processing states of the workpieces, for a production system that analyzes the stoppage factors of the production line including a plurality of production devices,
outputting the degree of influence of the stoppage factor on the decline of the key performance indicators for each of the plurality of stoppage factors based on the key performance indicators, the equipment status of the plurality of production devices, and the processing status of the workpiece;
A trained model characterized by: a data acquisition step of acquiring learning data from a production system that analyzes factors that cause a production line including a plurality of production devices to stop, the learning data including key performance indicators that are criteria for determining whether the operating status of the production line is good or bad , the equipment status of the plurality of production devices including identification information for identifying abnormal conditions that subdivide the equipment status of the production devices, the processing status of workpieces processed by the production devices, and the degree of influence of the factors that cause the stoppage on a decrease in the key performance indicators in the equipment status of the plurality of production devices and the processing status of the workpieces;
a model generation step of generating a trained model for inferring the degree of influence of the stoppage factors on a decrease in key performance indicators from key performance indicators of the production line, the device states of the plurality of production devices, and the processing status of workpieces processed by the production devices, using the training data;
Including,
The trained model outputs the degree of impact of the stop factor on a decline in a key performance indicator for each of the plurality of stop factors;
A method for generating a trained model characterized by: performing the data acquisition step of acquiring new learning data and performing the model generation step using the new learning data, thereby updating the trained model;
The method for generating a trained model according to claim 15,