Device and program for failure prediction

Abstract

A failure prediction device according to the present disclosure comprises a determination unit, a setting unit, and a prediction unit. In response to the input of a failure prediction instruction, the determination unit performs a determination process for an operational downtime day of an industrial machine, based on initial operational data representing the operation of the industrial machine, or secondary operational data representing the operation of a peripheral device configured to operate in conjunction with the industrial machine. The setting unit establishes an evaluation period based on the number of determined operational downtime days. The prediction unit predicts the occurrence of an industrial machine failure based on the initial operational data generated during the established evaluation period.

Description

Technical field

[0001] The present disclosure relates to a device for failure prediction and a program with a function for predicting the occurrence of a failure of an industrial machine. Background of the technology

[0002] In recent years, industrial machines such as robots have been deployed in various locations. Robots capable of repeatedly performing a predefined task with high accuracy are particularly indispensable in production facilities for manufacturing goods. A sudden robot failure is one of the most critical situations that must be avoided when using robots, as it interrupts production for an extended period. To prevent sudden robot failures, a technique for predicting robot failures is known (Patent Literature 1). List of citations from patent literature

[0003] Patent Literature 1: Japanese Unexamined Patent Application Publication No. 2021-022074 Summary of the invention; Problem to be solved by the invention

[0004] In recent years, there have been situations where factory operations were suspended for extended periods due to measures to combat infectious diseases and other reasons. Such situations can lead to a decrease in the accuracy of failure prediction. Therefore, it is desirable to propose a technique that does not compromise the accuracy of failure prediction even when a robot's operation is interrupted for a prolonged period. Solution to the problem

[0005] A failure prediction device according to the present disclosure comprises a determination unit configured to perform, in response to the input of a failure prediction instruction, a determination process of an operational downtime day of an industrial machine based on first operating data representing an operation of the industrial machine, or second operating data representing an operation of a peripheral device configured to work in cooperation with the industrial machine.

[0006] a setting unit configured to define an evaluation period based on the number of days of operational downtime identified, and a prediction unit configured to predict the occurrence of an industrial machine failure based on the initial operational data generated during the defined evaluation period. Brief description of the drawings Fig. Figure 1 is a diagram showing a failure prediction system comprising a failure prediction device according to the present embodiment. Fig. Figure 2 is a hardware configuration diagram of the failure prediction device according to the present embodiment. Fig. Figure 3 is a functional block diagram of the failure prediction device according to the present embodiment. Fig. 4 is a diagram showing an example of torque data stored in a memory unit located in Fig. 3 is shown. Fig. 5 is a supplementary diagram to complement the description of the process of a in Fig. 3 units of determination shown. Fig. 6 is a flowchart showing an example of the process of the failure prediction device according to the present embodiment. Fig. 7 is a flowchart that illustrates an example of the process of determining an evaluation period within the [system / project]. Fig. Step 6 shows S1. Fig. Figure 8 is a flowchart that shows another example 1 of the procedure for determining the valuation period from step S1. Fig. 6 shows. Fig. Figure 9 shows an example of a reception screen, which is defined by the in Fig. The process shown in step S23 is displayed. Fig. 10 is a flowchart showing another example 2 of the procedure of the evaluation period determination process from step S1, which is in Fig. 6 is shown. Fig. 11 is a flowchart showing another example 3 of the procedure of the evaluation period determination process of step S1, which is in Fig. 6 is shown. Fig. 12 is a flowchart that shows another example 4 of the evaluation period determination process in step S1 from Fig. 6 shows. Detailed description of the invention

[0007] The following describes a failure prediction device according to the present embodiment with reference to the drawings. In the following description, components with essentially the same function and configuration are designated by the same reference numerals, and repeated descriptions are given only where necessary.

[0008] Fig. Figure 1 shows a failure prediction system 1 with a failure prediction device 2 according to the present embodiment. The failure prediction system 1 comprises a plurality of robot systems 100 (100a, 100b) and the failure prediction device 2, which is connected to the plurality of robot systems 100 via a network 90, for example, a LAN. The robot system 100 (100a, 100b) comprises a robot device 110 (110a, 110b) and a conveyor device 130 (130a, 130b), which operates in cooperation with the robot device 110. The failure prediction device 2 is connected via the network 90 to a robot controller, which forms the robot device 110.

[0009] The failure prediction device 2 according to the present embodiment has a failure prediction function that predicts the occurrence of a failure of the robot device 110 based on data (operational data) representing the operation of the robot device 110. The operational data of the robot device 110 generated during an evaluation period are used to calculate the failure prediction of the robot device 110. A feature of the failure prediction device 2 according to the present embodiment is that the evaluation period of the operational data used to calculate the failure prediction of the robot device 110 is determined based on the fact that the operation of the robot device 110 is interrupted.This function makes it possible to predict the occurrence of a failure of the robot device 110 using a sufficient amount of operating data, even if the robot device 110 is interrupted for a long period of time, and to suppress a reduction in the accuracy of the failure prediction due to a long interruption of the robot device 110.

[0010] Typically, the device 2 for failure prediction according to the present embodiment is configured as follows.

[0011] As in Fig. As shown in Figure 2, the device 2 for failure prediction is a computer device in which hardware such as an operating device 4, a display device 5, a communication device 6 and a storage device 7 is connected to a processor 3 such as a CPU.

[0012] The operating device 4 is implemented by a keyboard, mouse, jog wheel, or similar device. The operating device 4 can also be implemented by a touch panel, which also serves as the display device 5. The user can input various types of information into the device 2 for failure prediction via the operating device 4. The display device 5 is implemented by an LCD or similar device. Under the control of the processor 3, various screens are displayed on the display device 5. The communication device 6 is implemented by a communication module that conforms to any communication standard. The communication device 6 sends and receives various types of data to and from an external device, such as the robot device 110, under the control of the processor 3. The storage device 7 is implemented by an HDD, an SSD, or similar device.A failure prediction program is stored in storage device 7.

[0013] As in Fig. As shown in Figure 3, when the failure prediction program stored in the storage device 7 is executed by the processor 3, the device 2 functions as input unit 21, display unit 22, receiving unit 23, storage unit 24, determination unit 25, setting unit 26 and failure prediction unit 27.

[0014] The input unit 21 transmits a user command 4 to the failure prediction device 2 via the operating device. In particular, the input unit 21 transmits a failure prediction instruction and the number of operational downtime days, which are used to determine the evaluation period, to the failure prediction device 2.

[0015] The display unit 22 is operated by the function of the in Fig. The display device 5 shown in Figure 2 is implemented. The display unit 22 shows various screens related to failure prediction. For example, the display unit 22 shows a screen containing a software button for receiving a failure prediction instruction, a receive screen for receiving the number of operational downtime days from the user, which is used to calculate the evaluation period, a screen for displaying the predicted failure date as a result of the failure prediction, and the like.

[0016] The receiving unit 23 is operated by the function of the in Fig. The communication device 6 shown in Figure 2 is implemented. The receiving unit 23 receives torque data from the robot device 110 as operating data of the robot device 110. For example, the torque data is repeatedly measured in a defined cycle by a torque sensor provided in a motor that drives each joint of the robot device 110. The torque data received from the robot device 110 is stored in the storage unit 24 in conjunction with date and time information generated by the robot device 110 (measured by the torque sensor) or received by the robot device 110.

[0017] The storage unit 24 is defined by the function of the in Fig. The storage device 7 shown in Figure 2 is implemented. The storage unit 24 stores the torque data received from the robot device 110. As shown, for example, in Fig. As shown in Figure 4, the torque data is organized in a text file according to the date it was generated. Along with the torque data (Tr1, Tr2, ..., Tr2784), the text file contains date and time information to identify the date and time the torque data was generated. The torque data is described according to the order in which it is received by the robot device 110. The date and time information includes a start date and time for the torque data measurement, an end date and time for the torque data measurement, a measurement cycle, and so on.

[0018] The determination unit 25 determines the downtime day of the robot device 110 based on the torque data. Typically, the determination unit 25 defines a day on which no torque data is generated as the downtime day of the robot device 110.

[0019] For example, in Fig. As shown in Figure 5, the determination unit 25 assigns an operating flag "0" to the day designated as a business interruption day and an operating flag "1" to the day not designated as a business interruption day, i.e., a working day. Through the process of determining the business interruption day using the determination unit 25, each past day is distinguished as either a business interruption day or a working day.

[0020] Setting Unit 26 defines an evaluation period based on the number of operational downtime days of the robot device 110. The evaluation period is the time from the start date to the end date of the evaluation period, and the number of evaluation days corresponds to the length of the evaluation period. An initial evaluation period (hereinafter referred to as the start period) is predefined by Setting Unit 26 for the evaluation period definition process. Typically, the start period is a period where the end date is the day before the day the failure prediction instruction is entered, and the start date is the day that is a previously determined number of evaluation days from the end date (hereinafter referred to as the "start number of days"). The end date can be a user-defined date. The start number of evaluation days can be changed by the user.

[0021] If there are no business interruption days in the initial period, hiring unit 26 defines the initial period as the evaluation period. However, if there are business interruption days within the initial period, hiring unit 26 defines the length of the evaluation period as the number of days obtained by adding the number of business interruption days to the initial number of days. Specifically, hiring unit 26 defines the evaluation period as a period whose end date is the day before the day on which the outage prediction instruction is entered, and whose start date is the day representing the total number of days obtained by adding the number of business interruption days to the initial number of days from the end date. For example, if the initial period is 90 days from April 5, 2023 (start date) to April 6, 2023 (start date), the evaluation period will be 90 days from April 5, 2023 (start date) to April 6, 2023 (start date).If the end date is June 2023 and there are three days of business interruption during this period, the valuation period is set to 93 days from April 2, 2023 (start date) to June 6, 2023 (end date). Since the valuation period includes three days of business interruption, the number of operating days in the initial period, in which there are no days of business interruption, corresponds to the number of operating days in the valuation period that is set if there are days of business interruption in the initial period.

[0022] The failure prediction unit 27 predicts the occurrence of a failure of the robot device 110 based on the operating data generated during the evaluation period defined by the setting unit 26. For example, based on the torque data generated during the evaluation period, the failure prediction unit 27 generates an evaluation equation representing the change in the torque data over time and calculates the time until the torque value represented by the torque data reaches a defined threshold. A failure date for the robot device 110 is predicted based on the time until the torque value reaches the threshold and the current time.

[0023] The following refers to Fig. 6. An overview of the failure prediction processing by the device 2 for failure prediction according to the present embodiment is given. The failure prediction is executed when a user enters a failure prediction instruction. As described in Fig. As shown in Figure 6, the device 2 for failure prediction sets the evaluation period (S1), performs the failure prediction processing of the robot device 110 based on the torque data generated during the evaluation period (S2) and displays the prediction result (S3).

[0024] An example of the evaluation period determination process in step S1 of Fig. 6 is below with reference to Fig. 7 described. As in Fig. As shown in Figure 7, the failure prediction device 2 sets the initial period as the evaluation period (S12) if there are no outage days in the initial period of M days (S11; No). For example, the process in step S12 sets as the evaluation period a period whose end date is the day before the day the failure prediction instruction was entered, and whose start date is the initial number of M days before the end date. However, if there are outage days in the initial period of M days (S11; Yes), the failure prediction device 2 determines the number of outage days N in the initial period (S13) and sets the evaluation period to (M+N) days (S14).For example, the process of step S14 establishes a period in which the end date is the day before the day on which the outage prediction instruction was entered, and the start date is the day that is (M+N) days before the end date, where (M+N) days are determined by adding the number of outage days N to the initial number of days M.

[0025] If there are no days of downtime in a defined initial period, the failure prediction device 2, according to the present embodiment, can predict the occurrence of a failure of the robot device 110 based on the operating data generated during the initial period. Conversely, if there are days of downtime in the defined initial period, the occurrence of a failure of the robot device 110 can be predicted based on the operating data generated during the evaluation period, which is obtained by adding the number of days of downtime to the initial period.

[0026] For example, if the number of downtime days in the initial 90-day period is zero, the failure prediction device 2 performs the failure prediction using the operational data generated during the initial period. However, if the number of downtime days in the initial 90-day period is three, the failure prediction device performs the failure prediction based on the operational data generated during the evaluation period of 93 days, obtained by adding three downtime days to the initial number of days in the 90-day period. That is, if the number of downtime days in the initial 90-day period is three, the failure prediction can be performed using operational data from at least 90 days of operation.Therefore, the occurrence of a failure of the robot device 110 can be predicted using the operating data of 90 operating days, regardless of whether there are any days of downtime in the initial period, so that fluctuations in the prediction accuracy due to the presence of days of downtime can be suppressed and highly accurate predictions can be achieved.

[0027] After the in Fig. In the evaluation period determination process shown in section 6, an evaluation period longer than the initial period is automatically set if there are operational downtime days in the initial period. However, if there are unplanned, unintentional, consecutive operational downtime days, the number of operational downtime days used to determine the evaluation period can be specified by the user. Unplanned, unintentional, consecutive operational downtime days include, for example, days on which factory operations are suspended for an extended period to prevent the spread of an infectious disease.

[0028] Another example of the valuation period determination process in step S1 of Fig. 6 is below with reference to Fig. 8 and Fig. 9 described. As in Fig. As shown in Figure 8, according to the first modification of the present embodiment, the failure prediction device 2 sets the initial period as the evaluation period (S22) if there are no more consecutive days of downtime in the initial period of M days than the previously specified number of days (S21; No). For example, the process of step S22 sets as the evaluation period a period whose end date is the day before the day on which the failure prediction instruction was entered, and whose start date is the day that is M days before the end date. Conversely, if the number of consecutive days of downtime is greater than the current number of days in the initial period of M days (S21; Yes), the failure prediction device 2 displays a receive screen to receive from the user the number of days of downtime to be used for setting the evaluation period (S23). Fig. Figure 9 shows an example of the reception screen. As in Fig. As shown in Figure 9, the receiving screen 200 displays information about the consecutive downtime days included in the initial period, including the number of consecutive downtime days and the days of the week on which these days fall. The receiving screen 200 is also configured to receive the number of downtime days to be used for calculating the evaluation period. Upon pressing the Enter key, the user-defined downtime days are entered into the Failure Prediction Device 2. When the Failure Prediction Device 2 receives the number of downtime days P (S24), it sets an evaluation period of (M+P) evaluation days (S25).For example, the process of step S25 obtains a period whose end date is the day before the day on which the failure prediction instruction was entered, and whose start date is the day which has (M+P) days, by adding the number of operational disruption days P entered by the user to the initial number of days M, starting from the end date, and defining it as the evaluation period.

[0029] According to the failure prediction device 2 of the first modification of the present embodiment, if the operation of a factory is interrupted for an extended period to prevent the spread of an infectious disease, the evaluation period of the operational data used for the failure prediction of the robot device 110 installed in the factory can be set according to a user instruction. For example, if the number of consecutive days of operational interruption is 30 days, by entering a number of days greater than 30 days, the user can perform a failure prediction for the robot device 110 based on a larger amount of operational data than in normal prediction processing, thereby improving the accuracy of the failure prediction.If there are six consecutive days of operational downtime, including two days on Saturdays and Sundays that are regular holidays, the user can enter four days as operational downtime days for the evaluation period determination process, taking the regular holidays into account. This allows the outage forecast that comprehensively includes the extended period of operational downtime days to be compared with the outage forecast that does not include the extended period of operational downtime days under the same conditions.

[0030] After the in Fig. In the evaluation period determination process shown in section 6, the evaluation period is set such that the failure prediction for robot device 110 can be performed using the operating data from the initial number of days (operating days) M, regardless of whether there are any days of downtime in the initial period. However, the evaluation period can be set such that the failure prediction for robot device 110 can be performed using the operating data from more days than the initial number of days M if there are any days of downtime in the initial period.

[0031] Another example of the valuation period determination process in step S1 of Fig. 6 is below with reference to Fig. 10 described. As in Fig. As shown in Figure 10, according to the second modification of the present embodiment, the failure prediction device 2 sets the initial period as the evaluation period (S32) if there are no downtime days in the initial period of M days (S31; No). For example, the process of step S32 sets as the evaluation period a period whose end date is the day before the day on which the failure prediction instruction was entered, and whose start date is the day M days before the end date. However, if there are downtime days in the initial period of M days (S31; Yes), the failure prediction device 2 determines the number of downtime days N in the initial period (S33) and sets a number of additional days Q equal to the determined number of downtime days N (S34).Subsequently, the failure prediction device 2 establishes an evaluation period of (M+N+Q) evaluation days (S35). For example, the process in step S35 establishes a period whose end date is the day before the day the failure prediction instruction was entered, and whose start date is the day corresponding to the total number of days (M+N+Q) obtained by adding the initial number of days M, the number of downtime days N, and the number of additional days Q obtained backwards from the end date.

[0032] According to the failure prediction device 2 of the second modification of the present embodiment, if there are days with operational downtime in the initial period, the occurrence of a failure of the robot device 110 can be predicted using the operating data of a larger number of days than if there are no days with operational downtime in the initial period. For example, if the initial period is 90 days and the number of operational downtime days in the initial period is three, then according to the device 2, the failure of the robot device 110 can be predicted using the operating data of a larger number of days than if there are no days with operational downtime in the initial period. Fig. In the evaluation period determination process shown in section 6, the occurrence of a failure of robot device 110 can be predicted based on 90 days of operating data by resetting the start date of the initial period by three days. On the other hand, according to the process described in Fig. In the evaluation period determination process shown in Figure 10, the occurrence of a failure of the robot device 110 can be predicted using operating data from a total of 93 days by resetting the start date of the initial period by six days, even when taking into account the three days of downtime in the initial period. As described above, according to Device 2 for Failure Prediction of the second modification, if there is one day of downtime in the initial period, the failure prediction can be performed using operating data from more days than in Device 2 for Failure Prediction of the present embodiment, thereby further suppressing the decrease in the accuracy of the failure prediction.

[0033] In the Fig. 7, Fig. 8 and Fig. In the evaluation period determination process shown in Figure 10, it is determined, based on torque data, which are operating data of the robot device 110, whether there are any days of downtime for the robot device 110 in the initial period. However, whether there are any days of downtime for the robot device 110 in the initial period can also be determined based on operating data representing the operation of a conveyor device 130 that works in cooperation with the robot device 110.

[0034] The receiver unit 23 receives torque data from the robot device 110 as operating data, representing the operation of the robot device 110. The receiver unit 23 also receives velocity data from the conveyor device 130 as operating data, representing the operation of the conveyor device 130. The velocity data represents the conveying speed of the conveyor device 130 and is measured repeatedly by a velocity sensor attached to the conveyor device 130. The operating data of the conveyor device 130 is stored in the memory unit 24 in conjunction with date and time information generated or received by the conveyor device 130, or similar information.It should be noted that the operating data representing the operation of the robot device 110 and the operating data representing the operation of the conveyor device 130 can be received by a control unit that comprehensively controls the robot device 110 and the conveyor device 130.

[0035] The storage unit 24 stores the torque data received by the robot device 110 and the operating data of the conveyor device 130 received by the conveyor device 130.

[0036] The determination unit 25 determines the downtime of the robot device 110 based on the operating data of the conveyor device 130. Specifically, the determination unit 25 calculates a downtime of the conveyor device 130 per day based on the operating data of the conveyor device 130 and converts the calculated downtime of the conveyor device 130 into downtime days of the robot device 110. For example, a downtime of 24 hours for the conveyor device 130 is converted into 1 downtime day for the robot device 110, and a downtime of 12 hours for the conveyor device 130 is converted into 0.5 downtime days for the robot device 110. Of course, the conversion ratio of the downtime of the conveyor device 130 to the downtime days of the robot device 110 does not have to be 1:1.

[0037] The conversion method can be changed according to the maximum operating time of the robot device 110 per day. For example, if the maximum operating time of the robot device 110 per day is set to eight hours, the downtime of less than 16 hours for the conveyor device 130 is converted into 0 downtime days for the robot device 110, and the downtime of 20 hours for the conveyor device 130 is converted into 0.5 downtime days for the robot device 110.

[0038] Since the actual operating time of the robot device 110 per unit of time, for example per cycle or per hour, differs from the operating time of the conveyor device 130 (peripheral device), the conversion procedure can be changed based on the actual operating time per unit of time of the robot device 110 and the conveyor device 130 (peripheral device).

[0039] Another example of the valuation period determination process in step S1 of Fig. 6 is below with reference to Fig. 11 described. As in Fig. As shown in Figure 11, according to the third modification, the failure prediction device 2 sets the initial period as the evaluation period (S42) if there is no downtime day of conveyor device 130 in the initial period of M days (S41; No). For example, the process in step S42 sets as the evaluation period a period whose end date is the day before the day on which the failure prediction instruction was entered, and whose start date is the day that is M days before the end date. However, if there is a downtime of conveyor device 130 in the initial period of M initial days (S41; Yes), the failure prediction device 2 converts the downtime of conveyor device 130 per day into R downtime days of robot device 110 (S43) and sets an evaluation period of (M+R) evaluation days (S44).For example, the process of step S44 defines a period as the evaluation period, the end date of which is the day before the day on which the failure prediction instruction was entered, and the start date of which is the day corresponding to the number of days (M+R) obtained by adding the number of downtime days R of the robot device 110, into which the downtime of the conveyor device 130 was converted, to the initial number of days M before the end date.

[0040] The device 2 for failure prediction according to the third modification of the present embodiment has the same effect as the device 2 for failure prediction according to the present embodiment and can determine the operational downtime day of the robot device 110 on the basis of the operating data of the peripheral device that cooperates with the robot device 110, so that the degree of freedom of the processed operating data can be improved.

[0041] In the Fig. 7, Fig. 8, Fig. 10 and Fig. In the evaluation period determination process shown in Figure 11, the evaluation period, which corresponds to the number of operational downtime days of robot device 110, is determined by focusing attention on the operational downtime days of robot device 110. This has the effect of suppressing a reduction in the accuracy of the failure prediction due to a reduction in the number of days (quantity) of operational data used for failure prediction. As another method for determining the evaluation period with the same effect, there is a method for determining an evaluation period according to the number of operational days of robot device 110 by focusing attention on the operational days of robot device 110.

[0042] Another example of the valuation period determination process in step S1 of Fig. 6 is below with reference to Fig. 12 described. As in Fig.As shown in Figure 12, according to the fourth modification, the Failure Prediction Device 2 initializes the number of operating days a to the numerical value 0 (S51) and initializes the number of elapsed days b to the numerical value 1 (S52). The Failure Prediction Device 2 determines whether the day b days prior to the day on which the failure prediction instruction was received is an operating day (S53). If the day b days prior to the day on which the failure prediction instruction was received is an operational interruption day (S53; No), the value of the number of elapsed days b is incremented (S54), and the processing returns to the determination process of step S53.If the day b days prior to the day the failure prediction instruction was received is an operating day (S53; Yes), Failure Prediction Device 2 increments the value of the number of operating days a (S55) and determines whether the number of operating days a has reached a predetermined number of days (S56). If the number of operating days a has not reached the predetermined number of days (S56; No), the value of the number of days past b is incremented (S54), and the process returns to the determination process of step S53. The processes of steps S53 through S56 are repeated until the number of operating days a reaches the predetermined number of days.When the number of operating days a reaches the specified number of days (S56; Yes), the failure prediction device 2 sets as the evaluation period a period whose end date is the day before the day on which the failure prediction instruction was entered, and whose start date is the day b days before the end date (S57).

[0043] A feature of the failure prediction device 2 according to the present embodiment is that the evaluation period of the operating data used to calculate the failure prediction of the robot device 110 is determined based on the fact that the operation of the robot device 110 is interrupted. Accordingly, the target of the failure prediction is not limited to the robot device 110. The target of the failure prediction can be, for example, industrial machines such as a machine tool, a lathe, a conveyor, and a welding machine.

[0044] In the present embodiment, days on which no torque data were generated are defined as downtime days. However, days on which the operating time per day based on the torque data is short can also be defined as downtime days. For example, days on which the operating time is short include a day on which the operating time was intentionally short, a day on which maintenance work was carried out on the robot device 110, a day on which the robot device 110 was tested, days on which the robot device 110 was not operated normally, and a day on which the operation was unstable and the robot device 110 was operated only for a short time.

[0045] For example, let's assume the initial period is 90 days, the maximum operating time per day is 12 hours, and there is one day in the initial period when operations only lasted two hours. In this case, a day with only two hours of operation is counted as an operational downtime day, and the evaluation period is set to 91 days. Since operational data within the evaluation period is used for failure prediction, the operational data from a day with only two hours of operation can also be used for failure prediction. The operational data from a day with only two hours of operation can have a significant impact on the failure prediction.According to the present embodiment, the failure prediction can be carried out using the operating data of 90 operating days and also the operating data of the days of operational downtime on which the operating time is equal to or less than the predetermined time, so that the accuracy of the failure prediction can be improved compared to the failure prediction using the operating data of only the operating days on which the operating time is longer than the predetermined time.

[0046] In the present embodiment, the torque data relating to the torque value are used as data for calculating the failure prediction of the robot device 110, but the data are not limited to torque data as long as the failure prediction calculation of the robot device 110 can be performed. For example, all data representing the operation of the robot device 110 can be used, such as data relating to the value of the current flowing in the motor and data relating to the disturbance torque.

[0047] The various types of data, such as a failure prediction program, stored in the storage device can be distributed by recording them to a removable storage medium or by downloading them to device 2 for failure prediction over a network.

[0048] The following annexes also contain information on the present embodiment and on modifications. (Annex 1)

[0049] A failure prediction device 2 comprises a determination unit 25 configured to perform, in response to a failure prediction instruction, a determination process for an operational downtime day of an industrial machine based on first operational data representing operation of the industrial machine, or second operational data representing operation of a peripheral device configured to operate in cooperation with the industrial machine; a setting unit 26 configured to establish an evaluation period based on a number of specified operational downtime days; and a prediction unit 27 configured to predict the occurrence of an industrial machine failure based on the first operational data generated during the specified evaluation period. (Annex 2)

[0050] The assessment period described in Annex 1 is set to a number of days obtained by adding the number of days of business interruption to a fixed initial number of days. (Annex 3)

[0051] Setting unit 26 described in Annex 2 defines an evaluation period whose end date is one day before the day on which the failure prediction instruction was entered, and whose start date is a day which is a number of days obtained by adding the number of business interruption days to the previously defined initial number of days before the end date. (Annex 4)

[0052] Setting unit 26 described in Annex 1 sets the length of the evaluation period to a number of days obtained by adding a user-specified number of days to a previously specified initial number of days if the number of consecutive days of operational downtime exceeds a previously specified threshold, and sets the length of the evaluation period to the initial number of days if the number of consecutive days of operational downtime does not exceed the specified threshold. (Annex 5)

[0053] Setting unit 26 described in Annex 1 defines the evaluation period such that the number of operating days of the industrial machine in the evaluation period corresponds to a specified number of days. (Annex 6)

[0054] The determination unit 25 described in one of Annexes 1 to 5 converts the downtime per day of the peripheral device into the number of downtime days of the industrial machine based on the second operating data. (Annex 7)

[0055] The determination unit 25 described in one of Annexes 1 to 5 determines a day on which the first operating data are not generated, or a day on which the time during which the first operating data are generated does not reach a predetermined time, as the operating interruption day of the industrial machine. (Annex 8)

[0056] The industrial machine according to one of Annexes 1 to 7 is a robot device 110, the first operating data are data relating to a load torque in relation to a drive shaft of a motor for driving a joint part of the robot device 110, and the prediction unit 27 predicts the occurrence of a failure of the drive shaft based on the data relating to the load torque generated during the specified evaluation period. (Annex 9)

[0057] A program that causes a computer to function as: Means of executing, in response to an input of a failure prediction instruction, a determination process of an operational downtime day of an industrial machine based on first operational data representing an operation of the industrial machine, or second operational data representing an operation of a peripheral device configured to operate in cooperation with the industrial machine; Means of determining an evaluation period based on a specified number of business interruption days; and Means of predicting the occurrence of an industrial machine failure based on initial operational data generated during the defined evaluation period.

[0058] Although embodiments of the present disclosure have been described in detail, the present disclosure is not limited to the individual embodiments described above. These embodiments may be subjected to various additions, substitutions, modifications, partial deletions, etc., without departing from the essence of the invention or the idea and spirit of the present invention as derived from the content of the claims and their equivalents. For example, the embodiments described above show the sequence of operations and the sequence of processes as examples, and the sequences are not limited to these. The same applies if numerical values ​​or formulas are used in the description of the embodiments described above. Explanation of reference symbols

[0059] 1: Failure prediction system, 2: Failure prediction device, 3: Processor, 4: Operating device, 5: Display device, 6: Communication device, 7: Storage device, 21: Input unit, 22: Display unit, 23: Receiving unit, 24: Storage unit, 25: Determination unit, 26: Setting unit, 27: Failure prediction unit, 90: Network, 100: Robot system, 110: Robot device, 130: Conveyor device. QUOTES INCLUDED IN THE DESCRIPTION

[0000] This list of documents cited by the applicant was automatically generated and is included solely for the reader's convenience. The list is not part of the German patent or utility model application. The DPMA accepts no liability for any errors or omissions. Cited patent literature

[0000] JP 2021-022074

[0003]

Claims

[1] Failure prediction device comprising: a determination unit configured to perform a determination process of an operational downtime day of an industrial machine in response to an input of a failure prediction instruction, based on first operating data representing an operation of the industrial machine, or second operating data representing an operation of a peripheral device configured to operate in cooperation with the industrial machine; a hiring unit configured to establish a fixed evaluation period based on a number of identified business interruption days; and a prediction unit configured to predict the occurrence of an industrial machine failure based on the initial operational data generated during the specified evaluation period. [2] The failure prediction device according to claim 1, wherein the evaluation period is set to a length of a number of days, is obtained by adding the number of days of operational downtime to a preset initial number of evaluation days. [3] The failure prediction device according to claim 2, wherein the setting unit defines an evaluation period whose end date is one day before the day on which the failure prediction instruction was entered and whose start date is a day, is obtained by adding the number of days of operational downtime to the previously defined number of evaluation days from the end date. [4] The failure prediction device according to claim 1, wherein the setting unit sets the length of the evaluation period to a number of days, is obtained by adding a user-specified number of days to a specified initial number of evaluation days when a number of consecutive days of downtime exceeds a specified threshold, and sets the length of the evaluation period to the initial number of evaluation days when the number of consecutive days of downtime does not exceed the specified threshold. [5] The device for failure prediction according to claim 1, wherein the setting unit determines the evaluation period such that a number of operating days of the industrial machine in the evaluation period corresponds to a specified number of days. [6] The device for failure prediction according to any one of claims 1 to 5, wherein the determining unit converts an operational downtime per day of the peripheral device into the number of operational downtime days of the industrial machine based on the second operating data. [7] The device for failure prediction according to any one of claims 1 to 5, wherein the determining unit determines a day on which the first operating data are not generated or a day on which a time in which the first operating data are generated does not reach a predetermined time, as the operational interruption day of the industrial machine. [8] The device for failure prediction according to any one of claims 1 to 7, wherein the industrial machine is a robot device The first operating data are data relating to a load torque in relation to a drive shaft of a motor for driving a joint section of the robot device, and The prediction unit predicts the occurrence of a drive shaft failure based on data relating to the load torque generated during the specified evaluation period. [9] A program that causes a computer to function as: Means of executing, in response to the input of a failure prediction instruction, a determination process of an operational downtime day of an industrial machine based on first operational data representing an operation of the industrial machine, or second operational data representing an operation of a peripheral device configured to operate in cooperation with the industrial machine; Means of determining an evaluation period based on a specified number of business interruption days; and Means of predicting the occurrence of an industrial machine failure based on initial operational data generated during the defined evaluation period.

Images