Abnormal management method, management device and storage medium

Abstract

The abnormality management method for a transport arm includes: an acquisition step of acquiring object data, which is transition data of an abnormality scale calculated on the basis of a feature amount related to a transport operation of the transport arm; a specifying step of comparing the object data acquired in the acquiring step with reference data, which is a plurality of pieces of transition data in which an abnormality has occurred, acquired in the past and stored in advance, and specifying at least one piece of reference data similar to the object data on the basis of the comparison result; an estimation step for estimating, on the basis of the abnormality occurrence time of the at least one reference data determined in the determination step, an abnormality occurrence prediction time of the transport arm from which the target data has been acquired; and an output step for outputting information indicating the abnormality occurrence prediction time estimated in the estimation step.

Description

Technical Field

[0001] The invention relates to an abnormality management method, a management device and a storage medium. Background Art

[0002] Patent document 1 describes a method for monitoring a transport unit that transports a transported object. Specifically, Patent document 1 describes the following: based on the relational model obtained by machine learning and the plurality of state values ​​obtained for the transport unit to be monitored, the health of the transport unit to be monitored is calculated. In addition, Patent document 1 describes outputting information on the state of the transport unit to be monitored based on the calculated health.

[0003] Prior art literature

[0004] Patent Literature

[0005] Patent Document 1: Japanese Patent Application Publication No. 2020-086694 Summary of the invention

[0006] Technical problem to be solved by the invention

[0007] The present invention provides an abnormality management method, a management device and a storage medium capable of predicting the occurrence of abnormalities in a substrate conveying device with high accuracy.

[0008] Technical solutions for solving technical problems

[0009] An abnormality management method according to one aspect of the present invention is an abnormality management method for a substrate conveying device for conveying substrates, the abnormality management method comprising: an acquisition step of acquiring object data, the object data being transition (i.e., shift) data of an abnormality scale calculated based on a feature quantity related to a conveying action of the substrate conveying device; a determination step of comparing the object data acquired in the acquisition step with reference data of a plurality of transition data acquired in the past and saved in advance as abnormalities, and determining at least one reference data similar to the object data based on the comparison result; an inference step of inferring a predicted time of abnormality occurrence of the substrate conveying device that acquired the object data based on the abnormality occurrence time of at least one reference data determined in the determination step; and an output step of outputting information indicating the predicted time of abnormality occurrence calculated in the inference step.

[0010] Effects of the Invention

[0011] According to the present invention, the occurrence of abnormality related to the substrate transport device can be predicted with high accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 It is a perspective view showing a substrate processing system.

[0013] Figure 2 yes Figure 1 Cross-sectional view along line II-II.

[0014] Figure 3 yes Figure 2 Cross-sectional view along line III-III.

[0015] Figure 4 It is a diagram for explaining the structure of the transport arm.

[0016] Figure 5 It is a longitudinal sectional side view schematically showing the coating unit.

[0017] Figure 6 This is a diagram illustrating the transfer of a chip from a transport arm to a spin chuck.

[0018] Figure 7 is a schematic diagram illustrating the functional structure of a controller.

[0019] Figure 8 A diagram showing an example of an image for user visual confirmation.

[0020] Fig. 9 FIG. 2 is a diagram showing another example of an image for user visual confirmation.

[0021] Fig.10 is a schematic diagram illustrating the hardware structure of the controller.

[0022] Fig.11 is a flowchart showing the steps of the abnormality management method.

[0023] Fig.12 It is a diagram for explaining the comparison of movement trajectories and the setting of the threshold value of the abnormality scale.

[0024] Fig.13 It is a diagram for explaining the exponential degradation model.

[0025] Fig.14 It is a diagram for explaining the effects of theta and beta in the exponential degradation model.

[0026] Fig.15 This is a diagram for explaining the estimation of the remaining useful life period.

[0027] Fig.16 This is a diagram for explaining the estimation result of the remaining service life when the initial value of the parameter is close to the optimum value (optimum value).

[0028] Fig.17 This is a diagram for explaining the estimation result of the remaining service life when the initial value of the parameter deviates from the optimal value.

[0029] Fig.18This is a flowchart showing the steps of the abnormality management method according to a modified example. DETAILED DESCRIPTION

[0030] Hereinafter, the embodiments will be described in detail with reference to the drawings. In the description, the same elements or elements having the same functions are denoted by the same reference numerals, and repeated descriptions are omitted.

[0031] [Structure of substrate processing system]

[0032] like Figure 1 As shown, the substrate processing system 1 includes a coating and developing device 2 and an exposure device 3. The exposure device 3 performs exposure processing on the resist film. Specifically, energy rays are irradiated to the exposure target portion of the resist film (photosensitive coating) by a method such as immersion exposure. Examples of the energy rays include ArF excimer laser, KrF excimer laser, g-rays, i-rays, and extreme ultraviolet rays (EUV).

[0033] The coating and developing device 2 is a substrate processing device that forms a resist film on the surface of the wafer W (substrate) before the exposure device 3 performs exposure processing, and performs development processing of the resist film after the exposure processing. In addition, in the present embodiment, the coating and developing device 2 acts (functions) as a management device of a substrate conveying device (details will be described later) that conveys the wafer W. In addition, in the present embodiment, the coating and developing device 2 is described as a management device of the substrate conveying device, but a server outside the substrate processing system 1 can also act as the above-mentioned management device. In the present embodiment, the wafer W is in the shape of a circular plate, but a wafer in which a part of the circle is cut off or in a shape other than a circle such as a polygon can also be used. The wafer W can also be various substrates such as a semiconductor substrate, a glass substrate, a mask substrate, an FPD (Flat Panel Display) substrate, etc.

[0034] like Figure 1 to Figure 3 As shown, the coating and developing device 2 includes a carrier block 4, a processing block 5, an interface block 6 and a controller 100 (control unit) (see Figure 3 ). The carrier block 4, the processing block 5 and the interface block 6 are arranged in a horizontal direction.

[0035] The carrier block 4 has a carrier station 12 and a loading and unloading section 13. The loading and unloading section 13 is located between the carrier station 12 and the processing block 5. The carrier station 12 supports a plurality of carriers 11. The carrier 11 stores a plurality of wafers W, for example, in a sealed state, and has an opening and closing door (not shown) on the side 11a for inserting and removing the wafers W. The carrier 11 is detachably arranged on the carrier station 12 in such a manner that the side 11a faces the loading and unloading section 13. The loading and unloading section 13 has a plurality of opening and closing doors 13a corresponding to the plurality of carriers 11 on the carrier station 12. By simultaneously opening the opening and closing doors of the side 11a and the opening and closing doors 13a, the inside of the carrier 11 is connected to the inside of the loading and unloading section 13. The loading and unloading section 13 has a built-in transfer arm A1. The transfer arm A1 takes out the wafer W from the carrier 11 and transfers it to the processing block 5, receives the wafer W from the processing block 5 and transfers it back to the carrier 11.

[0036] The processing block 5 includes a BCT module (lower film forming module) 14, a COT module (resist film forming module) 15, a TCT module (upper film forming module) 16, and a DEV module (development processing module) 17. These modules are arranged in the order of the DEV module 17, the BCT module 14, the COT module 15, and the TCT module 16 from the ground side.

[0037] The BCT module 14 is configured to form a lower film on the surface of the wafer W. The BCT module 14 has a plurality of coating units (not shown), a plurality of heat treatment units (not shown), and a transport arm A2 for transporting the wafer W to these units. The coating unit is configured to apply a coating liquid for forming the lower film to the surface of the wafer W. The heat treatment unit is configured to heat the wafer W using, for example, a hot plate, and to cool the heated wafer W using, for example, a cooling plate to perform heat treatment. As a specific example of the heat treatment performed in the BCT module 14, a heat treatment for solidifying (hardening) the coating liquid can be cited.

[0038] The COT module 15 is configured to be able to form a thermosetting and photosensitive resist film on the lower film. The COT module 15 is equipped with a plurality of coating units U1, a plurality of heat treatment units U2, and a conveying arm A3 (substrate conveying device) for conveying wafers W to these units. In addition, in the present embodiment, the conveying arm A3 is described as an example of a "substrate conveying device", but other conveying arms or transfer arms may also be an example of a "substrate conveying device". The coating unit U1 is configured to be able to apply a coating liquid (resist) for resist film formation on the lower film. The heat treatment unit U2 is configured to heat the wafer W, for example, by a hot plate, and to cool the heated wafer W, for example, by a cooling plate to perform heat treatment. As a specific example of the heat treatment performed in the COT module 15, a heat treatment for curing the coating liquid (PAB: Pre Applied Bake) can be cited.

[0039] Figure 4 2 is a diagram for explaining the structure of the transport arm A3. Figure 4 As shown, the transport arm A3 has: a plurality of holding bodies 25 for holding the wafer W, a base 21, a lifting platform 22, and a frame 23. The holding bodies 25 are arranged on the base 21 in a manner of overlapping each other up and down, and move forward and backward in the horizontal direction on the base 21 independently of each other. The holding bodies 25 hold the wafer W by surrounding the side circumference of the wafer W and supporting the back of the wafer W. The base 21 is arranged on the lifting platform 22 and can rotate around a vertical axis. The lifting platform 22 is arranged to be surrounded by a frame 23 extending in the up and down direction. The frame 23 is connected to the shell 24 and can move along the shell 24.

[0040] A driving mechanism for moving the retaining body 25 forward and backward is provided on the base 21. A driving mechanism for rotating the base 21 is provided on the lifting platform 22. A driving mechanism for raising and lowering the lifting platform 22 is provided on the frame 23. A driving mechanism for moving the frame 23 is provided on the housing 24. Each driving mechanism is composed of a motor, a pulley, and a belt wound around these motors and pulleys. The rotational motion of each motor is converted into a linear motion by each belt, and the frame 23, the lifting platform 22, and the retaining body 25 move. In addition, the base 21 rotates by the rotation of the pulley. Hereinafter, the above-mentioned driving mechanisms are sometimes collectively recorded as "driving mechanisms 28" (refer to Figure 7 ). The driving mechanism 28 transmits torque data on driving of the conveying arm A3 to the controller 100, for example.

[0041] return Figure 1 to Figure 3 The TCT module 16 is configured to form an upper film on the resist film. The TCT module 16 is equipped with a plurality of coating units (not shown), a plurality of heat treatment units (not shown), and a transport arm A4 for transporting the wafer W to these units. The coating unit is configured to apply a coating liquid for forming an upper film to the surface of the wafer W. The heat treatment unit is configured to heat the wafer W using, for example, a hot plate, and cool the heated wafer W using, for example, a cooling plate to perform heat treatment. As a specific example of the heat treatment performed in the TCT module 16, a heat treatment for curing the coating liquid can be cited.

[0042] The DEV module 17 is configured to be able to perform a development process on the exposed resist film. The DEV module 17 is built with: a plurality of developing units (not shown), a plurality of heat treatment units (not shown), a transport arm A5 for transporting the wafer W to these units, and a direct transport arm A6 for transporting the wafer W without passing through these units. The developing unit is configured to be able to partially remove the resist film to form a resist pattern. The heat treatment unit heats the wafer W, for example, by a hot plate, and cools the heated wafer W by a cooling plate to perform heat treatment. As specific examples of the heat treatment performed in the DEV module 17, a heat treatment before the development process (PEB: Post Exposure Bake), a heat treatment after the development process (PB: Post Bake), etc. can be listed.

[0043] A shelf unit U10 is provided on the carrier block 4 side in the processing block 5. The shelf unit U10 is provided to extend from the ground to the TCT module 16 and is divided into a plurality of small chambers arranged in the vertical direction. A lifting arm A7 is provided near the shelf unit U10. The lifting arm A7 lifts the wafer W between the small chambers of the shelf unit U10.

[0044] In addition, the shelf unit U10 is provided with an inspection module 30. The inspection module 30 is an image acquisition module. The inspection module 30 includes: a mounting table for mounting the wafer W and a camera for photographing the surface of the wafer W mounted on the mounting table. The inspection module 30 photographs the surface of the wafer W after the coating process by the coating unit U1 using the camera, and sends the obtained image data to the controller 100. As described later, the controller 100 can also detect the cut width of the resist film based on the image data.

[0045] A shelf unit U11 is provided on the interface block 6 side in the processing block 5. The shelf unit U11 is provided so as to extend from the floor to the top of the DEV module 17, and is divided into a plurality of small rooms arranged in the vertical direction.

[0046] The interface block 6 has a delivery arm A8 built therein, and is connected to the exposure device 3. The delivery arm A8 is configured to be able to take out the wafer W from the shelf unit U11 and deliver it to the exposure device 3, and to receive the wafer W from the exposure device 3 and return it to the shelf unit U11.

[0047] The controller 100 is composed of one or more control computers and implements the control of the coating and developing device 2. The controller 100 has: a display unit (not shown) that displays a setting screen of control conditions, an input unit (not shown) that inputs control conditions, and a reading unit (not shown) that reads a program from a computer-readable recording medium. The recording medium records a program for causing the coating and developing device 2 to execute a process. The program can be read by the reading unit of the controller 100. As the recording medium, for example, a semiconductor memory, an optical recording disk, a magnetic recording disk, or an optical magneto-recording disk can be used. The controller 100 implements the control of the coating and developing device 2 based on the control conditions input to the input unit and the program read by the reading unit.

[0048] [Structure of coating unit]

[0049] Next, refer to Figure 5 The coating unit U1 is described below. The coating unit U1 includes, for example, two processing units 41, a plurality of resist supply nozzles 42, a solvent supply nozzle 43, and a peripheral solvent supply nozzle 57. These resist supply nozzles 42 and solvent supply nozzles 43 are shared by the two processing units 41 and can be located on the wafers W of the two processing units 41. Figure 5 Only one processing unit 41 is shown in the figure. In order to simplify the description, only the structure of one processing unit 41 will be described below.

[0050] The processing unit 41 includes: a rotating chuck 51 for holding the back side of the wafer W by suction; a cup 52 surrounding the rotating chuck 51 and having an opening on the upper side; an exhaust port 53 for exhausting the inside of the cup 52; and a drain port 54. In addition, the processing unit 41 includes a lifting pin 55 for transferring the wafer W between the rotating chuck 51 and the transport arm A3.

[0051] The processing flow of the wafer W in the coating unit U1 is described. First, the holder 25 of the transport arm A3 receives the wafer W. The transport arm A3 transports the wafer W to the rotating chuck 51. When the holder 25 moves to the vicinity of the rotating chuck 51, the base 21 rotates, as shown in FIG. Figure 6 As shown by the solid line in (a), the holding body 25 is arranged to be opposite to the front side of the rotating suction cup 51. Then, the holding body 25 moves forward on the base 21, as shown in FIG. Figure 6 As shown by the dashed line in (a), the wafer W is transported to the delivery position of the spin chuck 51. Then, the wafer W is supported by the three raised lift pins 55, and the wafer W is delivered to the spin chuck 51 by the retreat of the holding body 25 and the descent of the lift pins 55.

[0052] Then, the wafer W is rotated by the rotating chuck 51, and the diluent is released (discharged) from the solvent supply nozzle 43 toward the center of the wafer W, and the diluent is spread to the peripheral portion of the wafer W by the centrifugal force. Furthermore, the resist is supplied to the central portion of the wafer W from the resist supply nozzle 42, and a resist film is formed on the entire wafer W by spin coating. Thereafter, the peripheral solvent supply nozzle 57 moves from the standby position outside the cup 52 to the solvent processing position inside the cup 52, and releases the solvent to the peripheral portion of the rotating wafer W. The solvent spreads from the solvent release position to the periphery of the wafer W due to the centrifugal force of the wafer W, and the unnecessary portion of the peripheral portion of the wafer W is removed in a ring shape. Then, when the supply of the solvent and the rotation of the wafer W are stopped and the processing is completed, the wafer W is sent out from the coating unit U1 by the conveying arm A3.

[0053] Here, the handover position of the wafer W relative to the rotating chuck 51 is preferably as follows: Figure 6 As shown in (a) of FIG. 5 , the rotation center P1 of the rotating chuck 51 and the center P2 of the wafer W are aligned with each other. When the rotation center P1 and the center P2 of the wafer W are aligned in this way, the center of the resist film from which the unnecessary part has been removed is aligned with the center P2 of the wafer W. However, when the transport arm A3 deteriorates over time, the belts of the drive mechanisms may become loose or the belts may skip teeth. In this case, for example, Figure 6 As shown in (b), there is a situation where the rotation center P1 is inconsistent with the center P2 of the wafer W, and the handover position of the wafer W relative to the rotating chuck 51 is deviated. As a result, the center of the resist film is eccentric relative to the center P2 of the wafer W. In this embodiment, the controller 100 calculates the eccentricity amount based on the cut-off width of the resist film of the wafer W (the removal width of the resist film) based on the image data obtained from the inspection module 30 (details will be described later). Then, the controller 100 includes the eccentricity amount in the characteristic amount to calculate the health of the conveying arm A3 (the degree of deterioration over the years), and calculates the predicted time of abnormality of the conveying arm A3 (details will be described later).

[0054] [Function of the controller]

[0055] Next, refer to Figure 7 , the functions of the controller 100 related to the abnormal management method of the transport arm A3 are described. Figure 7 As shown, the controller 100 includes an acquisition unit 111 , a calculation unit 112 , a comparison unit 113 , a determination unit 114 , a deduction unit 115 , and an output unit 116 .

[0056] The acquisition unit 111 acquires a feature quantity related to the transport action of the transport arm A3. The acquisition unit 111 acquires the feature quantity continuously at a predetermined time interval, for example, from the start of the operation of the transport arm A3. The acquisition unit 111 acquires, for example, image data obtained by photographing the surface of the wafer W after the coating process by the coating unit U1 through a camera from the inspection module 30. The image data is image data that can determine the cut-off width of the resist film on the wafer W (the removal width of the resist film). The acquisition unit 111, for example, estimates the eccentricity of the resist film on the wafer W (information about the deviation of the wafer W transported by the transport arm A3 from the target transport position) based on the cut-off width of the resist film shown in the above image data. The acquisition unit 111 acquires the eccentricity as a feature quantity related to the transport action of the transport arm A3. In addition, the acquisition unit 111 can also acquire torque data about the drive of the transport arm A3 from the drive mechanism 28 as a feature quantity related to the transport action of the transport arm A3. In the following, both the eccentricity and torque data are described as feature quantities, but only one of them may be used as the feature quantity, or other information related to the transport operation of the transport arm A3 may be used as the feature quantity. In addition, as additional information associated with the object data described later, the acquisition unit 111 acquires information including at least one of the type of the transport arm A3, the type of the substrate processing device including the transport arm A3, and the use status of the transport arm A3. The use status of the transport arm A3 is, for example, the average number of operations of the transport arm A3 (10,000 times / day, etc.) and the average transport speed.

[0057] The calculation unit 112 calculates (acquires) object data as transition data of abnormal scale based on the feature quantity acquired by the acquisition unit 111. The abnormal scale here refers to an index of the health of the conveying arm A3 taking into account the aging degradation. The abnormal scale can be set, for example, so that the lower the health (aging degradation is progressing) and the closer to the abnormality occurrence period, the lower the value. The calculation unit 112 can also calculate the abnormal scale by, for example, inputting the feature quantity into a degradation model constructed in advance. The degradation model in this case can also be a model constructed by learning a plurality of learning data, for example. The calculation unit 112 calculates the object data as transition data of the abnormal scale by calculating the abnormal scale for each feature quantity acquired at a prescribed time interval. That is, the object data is data that specifies the abnormal scale of each time from the start of operation to the present for the conveying arm A3. In addition, the method of calculating the abnormal scale based on the feature quantity is not limited to the above method.

[0058] The comparison unit 113 compares the acquired object data with a plurality of previously acquired transition data, i.e., reference data, in which an abnormality has occurred and which has been stored in advance. The comparison unit 113 acquires a plurality of reference data accumulated in the data accumulation unit 117. The reference data is data that specifies the abnormality scale of each time from the start of operation to the occurrence of an abnormality (inability to operate) for other transport arms. In addition, the comparison unit 113 also acquires additional information associated with each reference data from the data accumulation unit 117. The additional information is information including at least one of the type of transport arm (substrate transport device), the type of substrate processing device including the transport arm, and the use status of the transport arm. The use status of the transport arm is, for example, the average number of actions of the transport arm (10,000 times / day, etc.), the average transport speed, and the number of days until the abnormality occurs.

[0059] The comparison unit 113 compares the abnormality scale of the first period from the starting point (the time point when the operation starts) to the predetermined time (for example, the present time) in the object data with the abnormality scale of the comparison object period from the starting point to the time point that has passed the same length as the first period in the plurality of reference data. In addition, the comparison unit 113 also compares the additional information associated with the object data with the additional information respectively associated with the plurality of reference data.

[0060] The determination unit 114 determines at least one reference data similar to the target data based on the comparison result of the abnormality scale comparison performed by the comparison unit 113. The determination unit 114 may further consider the comparison result of the additional information (further consider the similarity of the additional information) to determine the reference data. That is, the determination unit 114 may determine the reference data having similar changes in abnormality scales and similar types of transport arms, types of substrate processing devices, and usage conditions of transport arms.

[0061] The estimating unit 115 estimates the predicted time of abnormality occurrence of the transport arm A3 that has acquired the target data based on the abnormality occurrence time in at least one reference data determined by the determining unit 114. The estimating unit 115 may directly estimate the abnormality occurrence time of the reference data as the predicted time of abnormality occurrence of the target data. In addition, the estimating unit 115 may consider the difference in the change of the abnormality scale between these data and estimate the predicted time of abnormality occurrence of the target data from the abnormality occurrence time of the reference data.

[0062] The output unit 116 outputs information indicating the calculated predicted time of abnormal occurrence. The output unit 116 outputs (displays) an image for user visual confirmation. Specifically, the output unit 116 outputs an image for user visual confirmation in which the trajectory of the object data and the predicted time of abnormal occurrence are associated on the graph. In addition, the output unit 116 may also output an image for user visual confirmation in which the end point of the trajectory of the object data on the graph is connected to the predicted time of abnormal occurrence.

[0063] Figure 8 FIG. 4 is a diagram showing an example of an image for user visual confirmation. Figure 8 In the figure, the horizontal axis represents time and the vertical axis represents the abnormality scale. The abnormality scale is set to "1" at the start of operation, and the lower the health level (the more advanced the deterioration over the years), the lower the value is as the abnormality occurs. Figure 8 In the user visual confirmation image shown, the trajectory RT of a plurality of reference data is shown. In the trajectory RT of each reference data, the abnormality occurrence time AT2 is indicated by an × mark. Figure 8 In the user visual confirmation image shown, the trajectory T1 of the object data is indicated by a thick solid line in parallel with the trajectory RT of each reference data. In addition, the abnormality occurrence prediction time AT1 estimated for the object data is indicated by an × mark. And the prediction trajectory T2 connecting the end point EP of the trajectory T1 of the object data and the abnormality occurrence prediction time AT1 is indicated by a thick dotted line.

[0064] Fig. 9 FIG. 4 is a diagram showing another example of an image for user visual confirmation. Fig. 9 In , the horizontal axis represents time and the vertical axis represents the anomaly scale. Fig. 9 In the user visual confirmation image shown in FIG. 1 , the trajectory T3 of the target data is represented by a thick solid line. Fig. 9 In the example shown, for example, due to the fact that there are multiple reference data that have been determined, the abnormality occurrence prediction time AT3 is expressed not as a point but as an area. In addition, the prediction trajectory T4 connecting the end point EP of the trajectory T3 of the object data and the area of ​​the abnormality occurrence prediction time AT3 is expressed by a thin solid line. In this way, the abnormality occurrence prediction time AT3 is expressed not as a fixed point (pinpoint) but as an area with a certain degree of area, thereby improving the recognition of abnormality prediction by the user, for example, like the expected forward route of a typhoon.

[0065] Fig.10 1 is a schematic diagram illustrating the hardware structure of the controller 100. The controller 100 is composed of one or more control computers. Fig.10As shown, the controller 100 has a circuit 190 , which includes at least one processor 191 , a memory 192 , a storage 193 , an input / output port 194 , an input device 195 , and a display device 196 .

[0066] The memory 193 has a computer-readable storage medium such as a hard disk. The memory 193 stores a program for causing the controller 100 to execute a method for managing abnormalities of the conveying arm A3 of the coating and developing device 2. For example, the memory 193 stores a program for causing the controller 100 to configure the above-mentioned functional blocks (functional modules).

[0067] The memory 192 temporarily stores the program loaded from the storage medium of the storage 193 and the calculation results of the processor 191. The processor 191 executes the above-mentioned program by cooperating with the memory 192, thereby forming the above-mentioned various functional modules. The input / output port 194 inputs and outputs electrical signals with the inspection module 30 and the drive mechanism 28 according to the instructions from the processor 191.

[0068] The input device 195 and the display device 196 function as a user interface of the controller 100. The input device 195 is, for example, a keyboard, etc., and acquires information input by a user. The display device 196 includes, for example, a liquid crystal monitor, etc., and is used to display information to the user (for example, the display of the above-mentioned user visual confirmation image). The display device 196 is used, for example, to display the above-mentioned main cause information. The input device 195 and the display device 196 can be integrated into a so-called touch panel.

[0069] [Abnormal management method of transport arm]

[0070] Next, refer to Fig.11 , the processing steps of the abnormality management method of the conveying arm A3 executed by the controller 100 are explained. Fig.11 is a flowchart showing the steps of the abnormality management method.

[0071] like Fig.11 As shown, first, object data, which is transition data of an abnormality scale calculated based on a feature amount related to the transport operation of the transport arm A3, is acquired (step S1, acquisition step).

[0072] Next, the object data acquired in step S1 is compared with the reference data acquired in the past and saved in advance, and one or more reference data similar to the object data is determined based on the comparison result (step S2, determination step). In the determination step, the abnormal scale of the first period from the starting point to the end point of the object data can also be compared with the abnormal scale of the comparison object period from the starting point to the time point of the same length as the first period of the multiple reference data. In addition, at least one reference data similar to the object data can also be determined. In addition, in the determination step, the additional information associated with the object data and the additional information respectively associated with the multiple reference data can also be compared, and the reference data can be determined by further considering the similarity of the additional information.

[0073] Next, based on the abnormality occurrence time of at least one reference data item determined in step S2, the predicted abnormality occurrence time of the transport arm A3 that has acquired the target data is estimated (step S3, estimating step).

[0074] Next, information indicating the predicted time of abnormality occurrence calculated in step S3 is output (step S4, output step). In the output step, an image for user visual confirmation that makes the trajectory of the object data correspond to the predicted time of abnormality occurrence on the graph may be output. Furthermore, in the output step, an image for user visual confirmation that includes a predicted trajectory that connects the end point of the trajectory of the object data with the predicted time of abnormality occurrence on the graph may be output.

[0075] [Effects]

[0076] Next, the effects of the abnormality management method of the transport arm A3 according to the present embodiment will be described.

[0077] The abnormality management method of the transport arm A3 of the present embodiment is an abnormality management method of the transport arm A3 that transports the wafer W. The abnormality management method includes: an acquisition step of acquiring object data, which is transition data of an abnormality scale calculated based on a feature value related to the transport action of the transport arm A3. The abnormality management method compares the object data acquired in the acquisition step with a plurality of transition data, i.e., reference data, that have been acquired and stored in the past and in which an abnormality has occurred. Moreover, the abnormality management method includes: a determination step of determining at least one reference data similar to the object data based on the comparison result. In addition, the abnormality management method includes a calculation step, in which the predicted time of abnormality occurrence of the transport arm A3 that acquired the object data is calculated based on the abnormality occurrence time of at least one reference data determined in the determination step. Furthermore, the abnormality management method includes an output step, in which information indicating the predicted time of abnormality occurrence calculated in the calculation step is output.

[0078] In the abnormality management method of the transport arm A3 of the present embodiment, the change data of the abnormality scale calculated based on the characteristic quantity related to the transport action, that is, the object data, is compared with the change data of the abnormality that occurred in the past, that is, a plurality of reference data. Then, the reference data similar to the object data is determined, and the predicted time of the abnormality occurrence of the transport arm A3 that has acquired the object data is estimated based on the abnormality occurrence time of the reference data, and the predicted time of the abnormality occurrence is output. According to such a structure, the predicted time of the abnormality occurrence of the object data can be estimated based on the abnormality occurrence time of the reference data similar to the change of the abnormality scale, so the predicted time of the abnormality occurrence of the transport arm A3 that has acquired the object data can be predicted with high accuracy. In addition, in the abnormality management method, the change data of the abnormality scale calculated based on the characteristic quantity related to the transport action is used for comparison. Therefore, for example, even between different substrate processing devices and substrate transport devices, the above-mentioned comparison can be performed, and the abnormality prediction (prediction) and diagnosis can be performed using a unified abnormality index.

[0079] In the determination step, the abnormality scale of the first period from the starting point to the specified time point of the object data is compared with the abnormality scale of the comparison target period from the starting point to the time point of the same length as the first period of the plurality of reference data. In addition, at least one reference data similar to the object data can be determined. According to such a structure, by comparing the periods of the same length from the starting point to the specified time point, reference data with a higher degree of similarity to the object data can be determined. As a result, the predicted time of occurrence of the abnormality of the transport arm A3 that has acquired the object data can be predicted with higher accuracy.

[0080] The object data and the plurality of reference data are associated with additional information, and the additional information includes at least one of the type of substrate conveying device, the type of substrate processing device including the substrate conveying device, and the use conditions of the substrate conveying device. Moreover, in the determination step, the additional information associated with the object data and the additional information respectively associated with the plurality of reference data may be further compared, and the reference data may be determined by further considering the similarity of the additional information. In this way, by further considering the similarity of the additional information such as the use conditions of the substrate conveying device, it is possible to determine the reference data that is considered to have a more similar trajectory until the abnormality occurs. As a result, the predicted time of the abnormality occurrence of the substrate conveying device for which the object data is acquired can be predicted with higher accuracy.

[0081] In the output step, an image for user visual confirmation may also be outputted, which makes the trajectory of the object data correspond to the predicted time of abnormal occurrence on the curve graph. According to such a structure, the predicted time of abnormal occurrence can be displayed to the user together with its basis (change in the abnormal scale). Thus, the user can be presented with a high reliability of the estimated time of abnormal occurrence, and the user can implement planned PM (Productive Maintenance) without waste at the stage before the abnormality occurs.

[0082] In the output step, an image for user visual confirmation including the predicted trajectory connecting the end point of the trajectory of the object data and the predicted time of abnormal occurrence on a curve graph may also be output. Thus, the trajectory of the object data and the predicted time of abnormal occurrence can be presented to the user as a continuous trajectory, and the user can be presented with a situation where the reliability of the estimated time of abnormal occurrence is high. Thus, the user can be presented with a situation where the reliability of the estimated time of abnormal occurrence is high, and the user can implement a planned PM without waste at the stage before the abnormality occurs.

[0083] The characteristic quantity may also include information about the deviation of the wafer W transported by the transport arm A3 from the target transport position. In addition to the torque data of the motor of the transport arm A3, the information about the deviation of the wafer W is used as the characteristic quantity, thereby making it possible to more appropriately calculate the above-mentioned abnormal scale related to transport.

[0084] The present embodiment has been described above, but the present embodiment is not limited to the above method. For example, in the above embodiment, it is described that in the determination step, the abnormal scale of the object data is compared with the abnormal scales of a plurality of reference data to determine at least one reference data similar to the object data. More specifically, in the determination step, the trajectory RT (movement trajectory, reference data) of the abnormal scale of the object data may also be compared with the abnormal scale of a plurality of reference data. Fig.12 ) and the trajectory RT of the abnormal scale of multiple reference data (movement trajectory, reference Fig.12 ) to determine at least one reference data similar to the object data. The comparison of moving trajectories can be performed, for example, based on the correlation coefficient or the mean absolute error. The so-called trajectory RT of the abnormal scale of the object data is the moving trajectory of the abnormal scale of the object data in the first period from the starting point to the specified time point. The trajectory RT of the abnormal scale of the reference data is the moving trajectory of the abnormal scale of the reference data in the comparison object period from the starting point to the time point after the same length as the above-mentioned first period. In this way, by comparing the moving trajectories of the abnormal scale, it is possible to appropriately determine the reference data whose abnormal scale transition is similar to the object data, and the predicted time of the abnormal occurrence of the object data can be predicted with high accuracy based on the abnormal occurrence time of the reference data.

[0085] Furthermore, in the above-mentioned determination step, the "predetermined time point" during which the movement trajectory of the abnormal scale is determined to be obtained may also be the time point when the value of the abnormal scale reaches a predetermined threshold value. Fig.12 In the example shown, the value of the abnormal scale of 0.9 is set as the threshold value X1. In this case, the moving trajectory of the first period from the starting point of the object data to the time point when the value of the abnormal scale reaches the threshold value X1 (0.9) can be extracted and used in the determination step. Such a threshold value X1 can also be, for example, information based on the failure history of the device, the time point when the failure begins to increase significantly statistically, etc. Specifically, it can also be the time point when the slope of the moving trajectory becomes above a certain slope, etc. In this way, by extracting the moving trajectory until the time point when the value of the abnormal scale reaches the specified threshold and using the moving trajectory in the determination step, the moving trajectory of the appropriate period can be extracted from the perspective of the value of the abnormal scale, and the predicted time of the occurrence of the abnormality of the object data can be predicted with high accuracy.

[0086] In addition, in the determination step, when reference data similar to the object data cannot be determined, a fault threshold value of the value of the abnormality scale may be determined. In this case, in the estimation step, the predicted time of abnormality occurrence may be estimated based on the time until the value of the object data reaches the fault threshold value. The case where reference data similar to the object data cannot be determined includes, for example, a case where there is little fault history information and little reference data.

[0087] As a premise for performing the above-mentioned processing, for example, an exponential degradation model can be used. The exponential degradation model is a model that expresses the transition of a health indicator until a failure occurs. Fig.13 This is a diagram that explains the exponential degradation model. Fig.13 In the chart, the horizontal axis represents the remaining useful life (Remaining Useful Life (RUL)) and the vertical axis represents the health index (abnormality scale). The health index is set to "1" at the start of operation, and the lower the health (the more advanced the deterioration over the years), the higher the value is set. Fig.13 In the example shown in (a), for example, in three cases, Case 1, Case 3, and Case 4, the changes in the health index calculated based on the characteristic amount of the time series torque data are shown. Fig.13 As shown in (b), by fitting the changes in health indicators in each situation with an exponential degradation model, it can be expressed as a function of time based on a combination of coefficients. The exponential degradation model is represented by the following formula (1), for example. Theta (i.e., θ) and beta (i.e., β) are coefficients (parameters) in the exponential degradation model.

[0088] h(t)=phi+theta*exp(beta*t)(1)

[0089] Fig.14 is a graph illustrating the effects of theta and beta in an exponential degradation model. Fig.14 In (a), phi and beta are set to fixed values, and only theta value is changed to represent the behavior of health indicators in multiple situations. Fig.14 As shown in (a), theta can show the overall progress of degradation. Fig.14 In (b), phi and theta are set to fixed values, and only the value of beta is changed to represent the behavior of health indicators in multiple situations. Fig.14 As shown in (b), beta can express a sharp degradation slope just before failure.

[0090] Fig.15 This is a diagram for explaining the estimation of the remaining useful life. Fig.15 In (a), the horizontal axis represents the elapsed time, and the vertical axis represents the health index (abnormality scale). Fig.15 As shown in (a), when using the exponential degradation model, a slope detection threshold Y1 and a fault threshold Z1 are set. The slope detection threshold Y1 can also be set to the maximum value of the health index observed in a healthy belt state. The fault threshold Z1 is preferably determined based on historical data such as past failure cases and domain knowledge, but can also be temporarily set when data is insufficient. Fig.15 As shown in (a), when the value of the object data exceeds the slope detection threshold value Y1, the time until the value of the object data reaches the failure threshold value Z1 can be derived to estimate the remaining service life. Fig.15 In (b), the horizontal axis represents the elapsed time, and the vertical axis represents the estimated remaining service life. Fig.15 As shown in (b), the remaining service life is estimated from the time when the value of the object data exceeds the slope detection threshold value Y1. Fig.15 The dotted line in (b) indicates the assumed error range.

[0091] The values ​​of theta and beta, which are coefficients of the exponential degradation model, are successively estimated by, for example, Bayesian updating. When the values ​​of theta and beta, which are coefficients of the exponential degradation model, change by a certain value or more, the fault threshold Z1 is automatically offset up and down in linkage. That is, when the value of the coefficient increases by a certain value or more, the fault threshold decreases (for example, from 0.40 to 0.35), and when the value of the coefficient decreases by a certain value or more, the fault threshold increases (for example, from 0.40 to 0.45). In this way, the remaining service life is estimated and output while the value of the fault threshold is offset.

[0092] Furthermore, by making the initial values ​​of the coefficients (parameters) of the exponential degradation model close to the optimal values, the estimation accuracy of the remaining useful life can be improved from the start of the estimation. Fig.16 This is a diagram for explaining the estimation result of the remaining service life when the initial value of the parameter is close to the optimal value. Fig.17 This is a diagram for explaining the estimation result of the remaining service life when the initial value of the parameter deviates from the optimal value. Fig.16 (a) and Fig.17 (a) shows the initial value and optimal value (final) of each parameter. Fig.16 (b) and Fig.17 In (b), the horizontal axis represents the elapsed time, and the vertical axis represents the health index. Fig.16 (c) and Fig.17 In (c), the horizontal axis represents the actual remaining service life, and the vertical axis represents the estimated remaining service life. Fig.16 As shown in (c), when the initial value of the parameter is close to the optimal value, the remaining service life is estimated with high accuracy (close to the actual remaining service life) from the beginning of the estimation. Fig.17 As shown in (c), when the initial value of the parameter deviates (deviates) from the optimal value, the estimation accuracy of the remaining service life is low (deviates from the actual remaining service life) immediately after the estimation. That is, if the sampling (number of data) is not increased to a certain extent, it will not converge to the optimal exponential degradation model.

[0093] The initial values ​​of the coefficients (parameters) of the exponential degradation model can be determined, for example, based on past fault history information. In this case, the coefficients of the exponential degradation model are associated with the past fault history information. When diagnosing a new transport arm, the databased fault history information can be referenced to compare information such as machine type, arm type, and usage conditions (number of processed items per day, transport speed), and the closest information can be extracted, and the coefficients of the associated exponential degradation model can be set as the initial values.

[0094] Finally, refer to Fig.18 An example of the procedure of the abnormality management method according to the modification will be described. Fig.18 FIG. 1 is a flowchart showing the steps of an abnormality management method according to a modified example. Fig.18 As shown, first, the movement trajectory (graph trajectory) of the abnormal scale in the target data for a certain period of time is extracted based on the feature value (step S11). Then, the movement trajectory of the abnormal scale of the target data is compared with the movement trajectory of the abnormal scale of a plurality of reference data (past health index graphs) (step S12).

[0095] Then, it is determined whether there is a movement trajectory of reference data having a certain range of similarity with respect to the movement trajectory of the object data (step S13). In step S13, if it is determined that there is a movement trajectory of reference data having similarity, the movement trajectory (health index curve graph) of the reference data is determined (step S14), and the predicted time of abnormal occurrence is estimated based on the movement trajectory of the reference data. On the other hand, if it is determined in step S13 that there is no movement trajectory of reference data having a certain range of similarity, the initial value of the coefficient of the exponential degradation model is determined based on the past fault history information, and the predicted time of abnormal occurrence is estimated (step S15).

[0096] Description of Reference Numerals

[0097] 2…Coating and developing device (management device)

[0098] 100…Controller (control unit)

[0099] A3...Transport arm (substrate transport device)

[0100] W...wafer (substrate).

Claims

1. An abnormality management method, which is an abnormality management method for a substrate transporting device for transporting substrates, wherein the abnormality management method is characterized by comprising: A step of acquiring object data, wherein the object data is change data of an abnormality scale calculated based on a feature quantity related to a transport operation of a substrate transport device; a determination step of comparing the object data acquired in the acquisition step with a plurality of reference data acquired in the past and stored in advance as the transition data in which an abnormality has occurred, and determining at least one reference data similar to the object data based on the comparison result; an estimating step of estimating a predicted time of occurrence of an abnormality in the substrate transport device for which the object data is acquired, based on the time of occurrence of an abnormality in at least one of the reference data determined in the determining step; and An output step of outputting information indicating the predicted abnormality occurrence time estimated in the estimating step.

2. The abnormality management method according to claim 1, characterized in that: In the determination step, the abnormality scale of a first period from a starting point to a specified time point of the object data is compared with the abnormality scale of a comparison object period from a starting point to a time point having a length equal to that of the first period of a plurality of reference data, and at least one of the reference data similar to the object data is determined.

3. The abnormality management method according to claim 1, characterized in that: The object data and the plurality of reference data are associated with additional information, the additional information including at least one of a type of substrate transport device, a type of substrate processing device including the substrate transport device, and a use condition of the substrate transport device. In the determination step, the additional information associated with the target data is compared with the additional information associated with each of the plurality of reference data, and the reference data is determined in consideration of the similarity of the additional information.

4. The abnormality management method according to claim 1, characterized in that: In the output step, an image for user visual confirmation is outputted in which the trajectory of the object data and the predicted abnormality occurrence time are associated with each other on a graph.

5. The abnormality management method according to claim 4, characterized in that: In the output step, the user visual confirmation image including a predicted trajectory connecting an end point of the trajectory of the object data and the abnormality occurrence predicted time on the graph is output.

6. The abnormality management method according to any one of claims 1 to 5, characterized in that: The feature amount includes information on a deviation amount of a substrate transported by the substrate transport device relative to a target transport position.

7. The abnormality management method according to claim 2, characterized in that: In the determination step, the movement trajectory of the abnormal scale of the first period from the starting point to the specified time point of the object data is compared with the movement trajectory of the abnormal scale of the comparison object period from the starting point to the time point of the same length as the first period of a plurality of reference data, and at least one of the reference data similar to the object data is determined.

8. The abnormality management method according to claim 2, characterized in that: The specified time point is a time point when the value of the abnormality scale reaches a specified threshold value.

9. The abnormality management method according to claim 2, characterized in that: In the determination step, when the reference data similar to the target data cannot be determined, a failure threshold value of the value of the abnormality scale is determined; In the estimating step, the abnormality occurrence prediction time is estimated based on the time until the value of the target data reaches the failure threshold value.

10. A management device, which is a management device for a substrate transport device for transporting substrates, wherein: The invention comprises a control unit configured to execute the following steps: A step of acquiring object data, wherein the object data is change data of an abnormality scale calculated based on a feature quantity related to a transport operation of a substrate transport device; A determination step of comparing the acquired object data with a plurality of reference data of the transition data in which an abnormality has occurred, which have been acquired and stored in advance, and determining at least one reference data similar to the object data based on the comparison result; A calculation step of calculating a predicted time of occurrence of an abnormality of the substrate transport device that acquires the object data based on the determined time of occurrence of an abnormality of at least one of the reference data; and The output step outputs information indicating the estimated predicted time of occurrence of the abnormality.

11. The management device according to claim 10, characterized in that: The control unit compares the abnormality scale of a first period from a starting point to a specified time point of the object data with the abnormality scale of a comparison object period from a starting point to a time point having a length equal to that of the first period of a plurality of reference data, and determines at least one of the reference data that is similar to the object data.

12. The management device according to claim 10, characterized in that: The object data and the plurality of reference data are associated with additional information, the additional information including at least one of a type of substrate transport device, a type of substrate processing device including the substrate transport device, and a use condition of the substrate transport device. The control unit further compares the additional information associated with the target data with the additional information respectively associated with the plurality of reference data, and determines the reference data in consideration of the similarity of the additional information.

13. The management device according to claim 10, characterized in that: The control unit outputs an image for user visual confirmation in which the trajectory of the object data and the abnormality occurrence prediction time are associated with each other on a graph.

14. The management device according to claim 13, characterized in that: The control unit outputs the user visual confirmation image including a predicted trajectory connecting an end point of a trajectory of the object data and the abnormality occurrence predicted time on the graph.

15. The management device according to any one of claims 10 to 14, characterized in that: The feature amount includes information on a deviation amount of a substrate transported by the substrate transport device relative to a target transport position.

16. The management device according to claim 11, characterized in that: The control unit compares the movement trajectory of the abnormal scale of the first period from the starting point to the specified time point of the object data with the movement trajectory of the abnormal scale of the comparison object period from the starting point to the time point of the same length as the first period of the plurality of reference data, and determines at least one of the reference data that is similar to the object data.

17. The management device according to claim 11, characterized in that: The specified time point is a time point when the value of the abnormality scale reaches a specified threshold value.

18. The management device according to claim 11, characterized in that: When the reference data similar to the target data cannot be identified, the control unit determines a failure threshold value of the abnormality scale value and estimates the abnormality occurrence prediction time based on the time until the target data value reaches the failure threshold value.

19. A storage medium, which is a computer-readable storage medium, characterized in that: A program for causing a device to execute the abnormality management method according to any one of claims 1 to 9 is stored.

Images