Information processing device, machine difference analysis method, and substrate processing device

Abstract

To provide a technique for further improving the accuracy of machine difference analysis of substrate processing devices.SOLUTION: An information processing device for performing machine difference analysis for a plurality of substrate processing devices, includes: a reconstruction error matrix creation unit that uses a dimension reduction model trained on a data set as a reference to verify reconstruction errors of data sets of the plurality of substrate processing devices, and creates a reconstruction error matrix for the plurality of substrate processing devices; a machine difference calculation unit that calculates, as a machine difference, the magnitude of at least some reconstruction errors selected from the plurality of reconstruction errors included in the reconstruction error matrix; and a display control unit that displays the calculated machine difference on a display device.SELECTED DRAWING: Figure 3

Description

[Technical Field]

[0001] The present disclosure relates to an information processing apparatus, an apparatus difference analysis method, and a substrate processing apparatus. [Background technology]

[0002] For example, in theory, semiconductor manufacturing equipment executing processes according to the same recipe will have the same sensor behavior (sensor waveform data), and therefore a technology is known that realizes an analysis function for machine differences using log data from sensors of semiconductor manufacturing equipment executing processes according to the same recipe (see, for example, Patent Document 1). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2022-3664 Summary of the Invention [Problem to be solved by the invention]

[0004] The present disclosure provides a technique for further improving the accuracy of machine difference analysis of substrate processing apparatuses. [Means for solving the problem]

[0005] One aspect of the present disclosure is an information processing device that performs machine-to-machine error analysis for a plurality of substrate processing devices, and includes: a reconstruction error matrix creation unit that uses a dimension reduction model trained on a reference data set to verify reconstruction errors of data sets of the plurality of substrate processing devices and create a reconstruction error matrix for the plurality of substrate processing devices; a machine-to-machine error calculation unit that calculates, as a machine error, the magnitude of at least some of the reconstruction errors selected from the plurality of reconstruction errors included in the reconstruction error matrix; and a display control unit that displays the calculated machine error on a display device. [Effects of the Invention]

[0006] According to the present disclosure, it is possible to provide a technique for further improving the accuracy of machine difference analysis of substrate processing apparatuses. [Brief explanation of the drawings]

[0007] [Figure 1] 1 is a configuration diagram of an example of a substrate processing system according to an embodiment of the present invention. [Figure 2] FIG. 2 is a diagram illustrating a hardware configuration of an example of a computer. [Figure 3] FIG. 2 is a functional block diagram of an example of an equipment controller according to the present embodiment. [Figure 4] FIG. 10 is an explanatory diagram of an example of processing by a training unit and a reconstruction error matrix creation unit. [Figure 5] FIG. 10 is an explanatory diagram of an example of processing in a verification phase. [Figure 6] FIG. 10 is a diagram illustrating an example of a reconstruction error matrix. [Figure 7] FIG. 10 is a diagram illustrating an example of a reconstruction error matrix in which a range of reconstruction errors is selected. [Figure 8] FIG. 10 is a diagram illustrating an example of calculation of machine differences for each hierarchy. [Figure 9] 10 is a flowchart showing an example of a process of an apparatus difference analysis method performed by the substrate processing system according to the present embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0008] Hereinafter, the present embodiment will be described with reference to the drawings.

[0009] <System configuration> 1 is a configuration diagram of an example of a substrate processing system 1 according to the present embodiment. The substrate processing system 1 shown in FIG. 1 includes a substrate processing apparatus 10, a sensor 11, an equipment controller 12, a server apparatus 14, and an operator terminal 16. The substrate processing apparatus 10, the sensor 11, and the equipment controller 12 are installed in a manufacturing factory 2. The server apparatus 14 and the operator terminal 16 may be installed in the manufacturing factory 2 or may be installed outside the manufacturing factory 2. The operator terminal 16 is an information processing terminal such as a PC (Personal Computer) or a smartphone operated by an operator such as an operator in charge of the substrate processing apparatus 10 installed in the manufacturing factory 2.

[0010] The substrate processing apparatus 10, apparatus controller 12, server apparatus 14, and operator terminal 16 in FIG. 1 are communicably connected via networks 18 and 20 such as the Internet or a LAN (Local Area Network).

[0011] The substrate processing apparatus 10 is an apparatus that performs processes such as film formation, etching, or ashing, and processes, for example, semiconductor wafers (hereinafter simply referred to as wafers). The substrate processing apparatus 10 is, for example, a semiconductor manufacturing apparatus, a heat treatment apparatus, or a film formation apparatus. The substrate processing apparatus 10 receives, for example, a recipe from an apparatus controller 12 and executes the recipe to perform a process. The recipe is a control command that combines setting values ​​of various categories, such as temperature, gas, pressure, plasma, and mechanics. The recipe of the substrate processing apparatus 10 has a plurality of control units called steps. The process of the substrate processing apparatus 10 includes a plurality of steps.

[0012] The substrate processing apparatus 10 is equipped with a plurality of sensors 11, such as a temperature sensor for measuring temperature and a pressure sensor for measuring pressure. The substrate processing apparatus 10 also is equipped with an actuator that performs mechanical operations by combining a power source and components.

[0013] The equipment controller 12 has a function of a man-machine interface that receives instructions for the substrate processing apparatus 10 from an operator and provides the operator with information related to the substrate processing apparatus 10. The equipment controller 12 receives sensor data output from a plurality of sensors 11 installed in the substrate processing apparatus 10. The equipment controller 12 may perform abnormality detection or abnormality prediction in the substrate processing apparatus 10.

[0014] 1 is provided for each substrate processing apparatus 10, it may be provided for each of a plurality of substrate processing apparatuses 10. The equipment controller 12 may be provided inside or outside the housing of the substrate processing apparatus 10. The equipment controller 12 may have a function to communicate with the equipment controller 12 for another substrate processing apparatus 10. The equipment controller 12 may have a function to communicate with the equipment controller 12 for another substrate processing apparatus 10 via a server apparatus 14. In this way, the equipment controller 12 may use information about a plurality of substrate processing apparatuses 10 (such as sensor waveform data for each step when a process is performed according to the same recipe).

[0015] The server device 14 may receive sensor data output from a plurality of sensors 11 installed in the substrate processing apparatus 10 and store the data as a process log each time a process is performed (for each Run).

[0016] The server device 14 may store, for example, information about multiple substrate processing apparatuses 10 in one or more manufacturing factories 2 (such as the recipe for the process performed in the substrate processing apparatus 10, sensor data and result data when the process is performed by executing that recipe) as a process log for each run.

[0017] The equipment controller 12 and the server device 14 may display information about the substrate processing apparatus 10 on the operator terminal 16, or may notify the operator of the operator terminal 16 by email or the like. At least one of the equipment controller 12, the server device 14, and the operator terminal 16 has a function of editing a recipe to be executed by the substrate processing apparatus 10. The equipment controller 12, the server device 14, and the operator terminal 16 in FIG. 1 are an example of an information processing apparatus according to this embodiment.

[0018] 1 is one example, and it goes without saying that there are various system configurations depending on the application and purpose. The division of the equipment into the equipment controller 12 and the server device 14 in Fig. 1 is one example. For example, various configurations are possible, such as a configuration in which the equipment controller 12 and the server device 14 are integrated, or a configuration in which they are further separated.

[0019] <Hardware configuration> The equipment controller 12, server device 14, and operator terminal 16 shown in Fig. 1 may be realized by a computer having the hardware configuration shown in Fig. 2. Fig. 2 is a hardware configuration diagram of an example of a computer 500.

[0020] 2 includes an input device 501, an output device 502, an external I / F (interface) 503, a RAM (random access memory) 504, a ROM (read only memory) 505, a CPU (central processing unit) 506, a communication I / F 507, and an HDD (hard disk drive) 508, all of which are interconnected by a bus B. The input device 501 and the output device 502 may be connected and used when necessary.

[0021] The input device 501 is a keyboard, mouse, touch panel, etc., and is used by an operator to input operation signals. The output device 502 is a display, etc., and displays the results of processing by the computer 500. The communication I / F 507 is an interface that connects the computer 500 to the networks 18 and 20 shown in Figure 1. The HDD 508 is an example of a non-volatile storage device that stores programs and data.

[0022] The external I / F 503 is an interface with an external device. The computer 500 can read from a recording medium 503a such as an SD (Secure Digital) memory card via the external I / F 503. The external I / F 503 may also be able to write to the recording medium 503a such as an SD memory card via the external I / F 503.

[0023] The ROM 505 is an example of a non-volatile semiconductor memory (storage device) that stores programs and data. The RAM 504 is an example of a volatile semiconductor memory (storage device) that temporarily stores programs and data. The CPU 506 is an arithmetic unit that reads programs and data from a storage device such as the ROM 505 or HDD 508 onto the RAM 504 and executes processing to realize overall control and functions of the computer 500.

[0024] The equipment controller 12, the server device 14, and the operator terminal 16 of the substrate processing system 1 shown in FIG. 1 execute programs on the computer 500 shown in FIG. 2 to realize various functions described below.

[0025] <Functional configuration> In the following, an example will be described in which the information processing device that performs the machine difference analysis for the plurality of substrate processing apparatuses 10 is the equipment controller 12. Note that the information processing device that performs the machine difference analysis for the plurality of substrate processing apparatuses 10 may be the server device 14 or the operator terminal 16.

[0026] The equipment controller 12 of the substrate processing system 1 according to this embodiment is realized by, for example, the functional blocks shown in Fig. 3. Fig. 3 is a functional block diagram of an example of the equipment controller 12 according to this embodiment. Note that the functional block diagram of Fig. 3 omits illustration of components that are not necessary for explaining this embodiment.

[0027] The device controller 12 executes a program for the device controller 12 to implement an instrument error calculation unit 30, a reconstruction error matrix creation unit 32, a training unit 34, a data storage unit 36, a display control unit 38, an operation reception unit 40, and a dataset acquisition unit 42. The reconstruction error matrix creation unit 32 has a dimension reduction model 50. The training unit 34 has a dimension reduction model 52. The data storage unit 36 ​​has a reference dataset storage unit 54 and an analysis dataset storage unit 56.

[0028] The data set acquisition unit 42 acquires a reference data set and stores it in the reference data set storage unit 54. The reference data set includes sensor data output from the sensor 11 while the reference substrate processing apparatus 10 is executing a process in accordance with a recipe. The reference substrate processing apparatus 10 is, for example, a golden device, which is a reference device whose perfect operation is guaranteed.

[0029] Furthermore, the dataset acquisition unit 42 acquires an analysis dataset and stores it in the analysis dataset storage unit 56. The analysis dataset includes sensor data output from the sensor 11 while the substrate processing apparatus 10 performing the machine difference analysis is executing a process in accordance with a recipe. The data storage unit 36 ​​has a reference dataset storage unit 54 that stores a reference dataset and an analysis dataset storage unit 56 that stores an analysis dataset.

[0030] The training unit 34 trains the dimension reduction model 52 with the reference data set stored in the reference data set storage unit 54. The dimension reduction model 52 projects data in a high-dimensional space onto data in a low-dimensional space.

[0031] The dimension reduction model 52 is PCA (Principal component analysis), GPLVM (Gaussian Process Latent Variable Model), MPPCA (Mixture Probabilistic Principal Component Analysis), Kernel PCA, or Probabilistic PCA.

[0032] The reconstruction error matrix creation unit 32 uses the dimensionality reduction model 50 trained by the training unit 34 to verify the reconstruction error of the analysis dataset stored in the analysis dataset storage unit 56. The reconstruction error is the difference between the original data when the dimensions of the original data are reduced from a high-dimensional space to a low-dimensional space and then reconstructed from the low-dimensional space into a high-dimensional space. The reconstruction error matrix creation unit 32 also verifies the reconstruction error and creates a reconstruction error matrix, which will be described later.

[0033] The processing of the training unit 34 and the reconstruction error matrix creation unit 32 will be described with reference to Fig. 4. Fig. 4 is an explanatory diagram of an example of the processing of the training unit 34 and the reconstruction error matrix creation unit 32.

[0034] In the training phase, the training unit 34 trains the dimension reduction model 52 on the reference dataset stored in the reference dataset storage unit 54. The training unit 34 trains the dimension reduction model 52 by, for example, unsupervised learning.

[0035] In the validation phase, the reconstruction error matrix creation unit 32 verifies the reconstruction error of the analysis dataset stored in the analysis dataset storage unit 56 using the dimension reduction model 50 trained by the training unit 34 .

[0036] The processing in the verification phase will be further explained with reference to Fig. 5. Fig. 5 is an explanatory diagram of an example of the processing in the verification phase.

[0037] In step S1, the reconstruction error matrix creation unit 32 uses the dimension reduction model 50 trained by the training unit 34 to reduce the dimension of the original data of the analysis data set (actual points in FIG. 5) from a high-dimensional space to a low-dimensional space.

[0038] In step S2, the reconstruction error matrix creation unit 32 uses the dimension reduction model 50 trained by the training unit 34 to reconstruct the data whose dimensions have been reduced in step S1 into a high-dimensional space.

[0039] The position of the data reconstructed in the high-dimensional space is reconstructed into the original high-dimensional space using the dimension reduction model 50 trained with the reference data set, and therefore becomes the position where the data of the reference substrate processing apparatus 10 should be. Therefore, the distance between the position of the original data in the high-dimensional space and the position of the data reconstructed in the high-dimensional space from the data whose dimension has been reduced in step S1 is the instrument difference between the reference substrate processing apparatus 10 and the substrate processing apparatus 10 of the analysis data set.

[0040] In step S3, the reconstruction error matrix creation unit 32 verifies the reconstruction error between the points where data is reconstructed from low-dimensional space to high-dimensional space and the original data of the analysis dataset (the actual points in FIG. 5). Note that data with low frequency has a large reconstruction error because appropriate conversion to low-dimensional space is not learned. Furthermore, the dimension reduction model 50 trained by the training unit 34 has been trained on a reference dataset with changes in setting values. For example, even when the same recipe is being executed, the setting values ​​of the substrate processing apparatus 10 may be adjusted.

[0041] Therefore, the dimension reduction model 50 trained by the training unit 34 can verify the reconstruction error other than the influence of the change in the setting value, even for an analysis data set in which the setting value has changed. In this way, the reconstruction error verified in the verification phase represents the machine difference that absorbs the data fluctuation due to the change in the setting value.

[0042] The reconstruction error matrix creating unit 32 creates, for example, the reconstruction error matrix 1000 shown in Fig. 6 by repeating the processing of the verification phase shown in Fig. 5. Fig. 6 is a diagram showing the configuration of an example of the reconstruction error matrix 1000.

[0043] 6, the reconstruction error for each step of the sensor 11 that detects the state of the substrate processing apparatus 10 is taken as the smallest unit of reconstruction error. In the reconstruction error matrix 1000, the horizontal axis represents the process run. In addition, the vertical axis represents the step of the sensor 11, the sensor 11, and the group of the sensor 11. The reconstruction error matrix 1000 is created for each substrate processing apparatus 10.

[0044] Returning to FIG. 3, the machine error calculation unit 30 calculates the magnitude of at least a portion of the reconstruction errors selected from the reconstruction error matrix 1000 created by the reconstruction error matrix creation unit 32 as the machine error.

[0045] The machine error calculation unit 30 calculates the machine error for each layer according to the range of the minimum unit reconstruction error selected from the plurality of minimum unit reconstruction errors included in the reconstruction error matrix 1000. Fig. 7 is an explanatory diagram of an example of the reconstruction error matrix 1000 in which the range of the reconstruction error has been selected.

[0046] 7 is the range of the smallest unit of reconstruction error selected when calculating the machine error for the run "6-8" in all steps of the group of sensors 11 to which "sensor A1" belongs. When the reconstruction error range 1010 in Fig. 7 is selected, the machine error calculation unit 30 calculates the L2 norm as the magnitude of the smallest unit of reconstruction error included in the range 1010, thereby being able to calculate the machine error for the run "6-8" in all steps of the group of sensors 11 to which "sensor A1" belongs.

[0047] 7 is the range of the reconstruction error of the smallest unit selected when calculating the instrument error for all runs in step "1" of "sensor A1." When the reconstruction error range 1012 in Fig. 7 is selected, the instrument error calculation unit 30 can calculate the instrument error for all runs in step "1" of "sensor A1" by calculating the L2 norm as the magnitude of the reconstruction error of the smallest unit included in the range 1012.

[0048] Note that if the number of sensors 11 belonging to a group and the number of runs in a data set differ, the magnitudes cannot be directly compared due to the influence of the difference in the total number of sensors 11 belonging to the group and the difference in the number of runs in the data set. Therefore, in this embodiment, the instrument difference may be adjusted using the following formula (1) taking into account the difference in the total number of sensors 11 belonging to the group.

[0049] Adjusted instrument error = L2 norm / number of sensors 11 belonging to a group × average number of sensors 11 in all groups... (1) In this embodiment, the machine difference may be adjusted by the following formula (2) taking into account the number of runs in the data set.

[0050] Adjusted machine error = L2 norm / number of runs in the dataset × average number of runs across all datasets…(2) Furthermore, according to the reconstruction error matrix 1000 shown in FIGS. 6 and 7, the number of steps for each sensor 11 and the number of summary types such as “max,” “mean,” “min,” and “std” for each step are the same, so comparisons can be made between different steps or sensors 11.

[0051] Furthermore, according to the reconstruction error matrix 1000 shown in FIGS. 6 and 7, as shown in FIG. 8, the reconstruction error for each step of the sensor 11 that detects the state of the substrate processing apparatus 10 is taken as the smallest unit of reconstruction error, and the magnitude of the reconstruction error for each step of the sensor 11, all steps of the sensor 11, all steps of the group of sensors 11, and all steps of the substrate processing apparatus 10 can be calculated as an instrumental error for each hierarchical level.

[0052] 8A and 8B are diagrams illustrating an example of calculation of the machine error for each hierarchical level. Fig. 8A shows the range of the reconstruction error of the smallest unit selected when calculating the magnitude of the reconstruction error for all steps of the substrate processing apparatus 10. Fig. 8B shows the range of the reconstruction error of the smallest unit selected when calculating the magnitude of the reconstruction error for all steps of the group of sensors 11.

[0053] Fig. 8(C) shows the range of reconstruction error for the smallest unit selected when calculating the magnitude of reconstruction error for all steps of the sensor 11. Fig. 8(D) shows the range of reconstruction error for the smallest unit selected when calculating the magnitude of reconstruction error for each step of the sensor 11.

[0054] 3, the display control unit 38 displays the machine error calculated by the machine error calculation unit 30 on an output device 502 such as a display device. The operation reception unit 40 receives various operations from the operator. For example, the operation reception unit 40 receives a selection from the operator of the reconstruction error range 1010 or 1012 shown in FIG. 7.

[0055] <Processing> FIG. 9 is a flowchart showing an example of the process of the machine difference analysis method performed by the substrate processing system 1 according to this embodiment.

[0056] In step S10, the training unit 34 of the equipment controller 12 trains the dimension reduction model 52 on the reference data set stored in the reference data set storage unit 54. The dimension reduction model 52 can use PCA, GPLVM, MPPCA, Kernel PCA, or Probabilistic PCA, but it is preferable to use PCA and GPLVM from the viewpoint of accuracy.

[0057] In step S12, the dataset acquisition unit 42 acquires an analysis dataset and stores it in the analysis dataset storage unit 56.

[0058] In step S14, the reconstruction error matrix creation unit 32 uses the dimension reduction model 50 trained by the training unit 34 to reduce the dimension of the original data of the analysis dataset from a high-dimensional space to a low-dimensional space, and reconstructs the data from the low-dimensional space to a high-dimensional space.

[0059] In step S16, the reconstruction error matrix creation unit 32 verifies the reconstruction error, which is the difference between the original data and the data reconstructed from the low-dimensional space to the high-dimensional space, and creates the reconstruction error matrix 1000 shown in FIG.

[0060] In step S18, the machine error calculation unit 30 determines whether a reconstruction error has been selected from the created reconstruction error matrix 1000, for example, as shown in the reconstruction error ranges 1010 or 1012 in Fig. 7. If a reconstruction error has not been selected, the device controller 12 ends the processing of the flowchart in Fig. 9.

[0061] If a reconstruction error has been selected, the process proceeds to step S20, where the machine-difference calculation unit 30 calculates, as the machine-difference, the magnitude of the reconstruction error in, for example, the reconstruction error range 1010 or 1012 in Fig. 7 selected from the reconstruction error matrix 1000. For example, the machine-difference calculation unit 30 calculates the L2 norm of multiple reconstruction errors included in the range 1010 or 1012 as the magnitude of the reconstruction error in the reconstruction error range 1010 or 1012 in Fig. 7. The machine-difference calculation unit 30 only needs to calculate the L2 norm of multiple reconstruction errors, which reduces calculation costs.

[0062] In step S22, the display control unit 38 displays the machine difference calculated by the machine difference calculation unit 30 on the output device 502 such as a display device. The operator can check the machine difference displayed on the output device 502 such as a display device.

[0063] <Summary> Conventional machine difference analysis assumes that if the recipe setting values ​​are the same, the behavior of sensor data during the process is also the same, and determines that there is a machine difference if the behavior of sensor data during the process differs. However, even with the same recipe, setting values ​​are sometimes adjusted. As a result, conventional machine difference analysis sometimes detects differences in setting values ​​as machine differences, making it impossible to analyze the true machine differences.

[0064] Furthermore, in conventional machine difference analysis, when attempting to hierarchically calculate machine differences for each step of the sensor 11, all steps of the sensor 11, all steps of the group of the sensor 11, and all steps of the substrate processing apparatus 10, a separate model is required for each hierarchical level. In conventional machine difference analysis, it was not possible to compare machine differences for each step of the sensor 11, all steps of the sensor 11, all steps of the group of the sensor 11, and all steps of the substrate processing apparatus 10 within or between hierarchical levels.

[0065] A comparison of the machine difference within a hierarchical level is, for example, a comparison between the machine difference of "sensor A1" and the machine difference of "sensor B1" that belong to a different group of sensors 11. A comparison of the machine difference between hierarchical levels is, for example, a comparison between the machine difference of the substrate processing apparatus 10 and the machine difference of "sensor A1".

[0066] In this embodiment, one overall dimension reduction model 52 is trained with the reference data set, and the machine error for each layer is obtained by selecting the range of the reconstruction error of the smallest unit in the reconstruction error matrix 1000 shown in FIG. 6, for example, and calculating the L2 norm. This allows for direct comparison of machine errors between or within layers.

[0067] In this way, in this embodiment, machine difference analysis can be performed even on logs such as trace logs that have variations in setting values, thereby expanding the range of logs that can be analyzed for machine differences. For example, the variation in setting values ​​is a change in setting value that does not affect the process. For example, in this embodiment, in machine difference analysis of a log that has variations in setting values, if the temperature is fine-tuned to optimize the process results, the difference in temperature is not analyzed as a machine difference.

[0068] In addition, in this embodiment, the machine differences between or within layers can be verified using a single dimension-reduced model 50, making it possible to directly compare machine differences from the overall to the detailed aspects of the substrate processing apparatus 10. Furthermore, in this embodiment, the machine differences between or within layers can be verified using a single dimension-reduced model 50, making it possible to reduce calculation costs.

[0069] As described above, in this embodiment, the dimension reduced model 52 is trained using a reference data set (reference normal data) of the reference substrate processing apparatus 10, and the reconstruction error of the analysis data set (other data) is verified using the trained dimension reduced model 50. In this case, the reconstruction error serves as an index for evaluating whether the other data is normal.

[0070] The reconstruction error is the difference between the original data and the reconstructed data. The reconstruction error of other data with similar characteristics to normal data tends to be relatively small. On the other hand, the reconstruction error of abnormal data containing machine differences or noise can be relatively large.

[0071] Therefore, in this embodiment, a dimension reduction model 52 is trained using a reference data set of a reference substrate processing apparatus 10, and the trained dimension reduction model 50 is used to verify the reconstruction error of the verification data set, and the reconstruction error is taken as the machine error.

[0072] According to the substrate processing system 1 of this embodiment, a technique for further improving the accuracy of machine difference analysis of the substrate processing apparatus 10 can be provided.

[0073] Although the preferred examples of this embodiment have been described in detail above, this embodiment is not limited to the above-described examples, and various modifications and substitutions can be made to the above-described examples without departing from the scope of this embodiment.

[0074] The substrate processing apparatus 10 of the present disclosure can be applied to any type of apparatus, such as an atomic layer deposition (ALD) apparatus, a capacitively coupled plasma (CCP), an inductively coupled plasma (ICP), a radial line slot antenna (RLSA), an electron cyclotron resonance plasma (ECR), or a helicon wave plasma (HWP).The substrate processing apparatus 10 of the present disclosure can also be applied to a chemical vapor deposition (CVD) apparatus and an oxidation / annealing apparatus.

[0075] The substrate processing system 1 of the present disclosure is not limited to the configuration shown in Fig. 1, and there are various system configuration examples depending on the application and purpose. The substrate processing apparatus 10 of the present disclosure can be applied to any of a single-wafer apparatus that processes substrates one by one, a batch apparatus that processes multiple substrates at once, and a semi-batch apparatus. Processes performed by the substrate processing apparatus 10 of the present disclosure include, for example, a film formation process or an etching process. [Explanation of symbols]

[0076] 1. Substrate Processing System 10. Substrate processing equipment 11 Sensors 12 Equipment Controller 14 Server equipment 16 Worker terminal 30 Machine difference calculation section 32 Reconstruction error matrix creation unit 34 Training Department 50, 52 Dimensionality reduction model 54 Reference Data Set Storage Unit 56 Analysis Dataset Storage 502 Output Device

Claims

1. An information processing device that performs machine difference analysis for a plurality of substrate processing devices, a reconstruction error matrix creation unit that uses a dimension reduction model trained on a reference data set to verify reconstruction errors of the data sets of the plurality of substrate processing apparatuses and creates a reconstruction error matrix of the plurality of substrate processing apparatuses; an instrumental error calculation unit that calculates, as an instrumental error, the magnitude of at least some of the reconstruction errors selected from the plurality of reconstruction errors included in the reconstruction error matrix; a display control unit that displays the calculated machine error on a display device; An information processing device having the above.

2. the reconstruction error matrix creation unit verifies a minimum unit reconstruction error and creates the reconstruction error matrix for each of the substrate processing apparatuses; The machine-difference calculation unit calculates the machine-difference for each layer according to a range of the minimum unit reconstruction error selected from the plurality of minimum unit reconstruction errors included in the reconstruction error matrix.

2. The information processing device according to claim 1.

3. The machine-difference calculation unit defines the reconstruction error for each step of a sensor that detects the states of the plurality of substrate processing apparatuses as the minimum unit reconstruction error, and calculates the magnitude of the reconstruction error for each step of the sensor, all steps of the sensor, all steps of the group of the sensors, and all steps of the substrate processing apparatus as the machine-difference for each hierarchical level.

3. The information processing device according to claim 2.

4. the data set is stored for each substrate processing operation performed, with a set value associated with log data for each sensor of the substrate processing apparatus during the substrate processing operation in accordance with the set value; The log data may include changes in the set values ​​for each substrate processing operation that has been performed.

4. The information processing device according to claim 3,

5. The machine difference calculation unit calculates the L2 norm as the magnitude of the reconstruction error. The information processing device according to claim 1 .

6. 5. The information processing device according to claim 1, wherein the dimension reduction model is principal component analysis (PCA), Gaussian process latent variable model (GPLVM), mixture probabilistic principal component analysis (MPPCA), kernel PCA, or probabilistic PCA.

7. A machine difference analysis method performed by an information processing device that performs machine difference analysis on a plurality of substrate processing devices, comprising: verifying reconstruction errors of the data sets of the plurality of substrate processing apparatuses using a dimension reduction model trained on a reference data set, and creating a reconstruction error matrix for the plurality of substrate processing apparatuses; calculating, as an instrumental error, magnitudes of at least some of the reconstruction errors selected from the plurality of reconstruction errors included in the reconstruction error matrix; displaying the calculated machine error on a display device; A machine difference analysis method having the following.

8. A substrate processing apparatus comprising the information processing apparatus according to any one of claims 1 to 4.

Images