Information processing apparatus, substrate processing apparatus and information processing method

Abstract

To provide a technology that improves prediction accuracy of a process result.SOLUTION: An information processing apparatus that comprises: a model acquisition unit that acquires a learned model in which a relationship between a first log information measured during processing in a first substrate processing apparatus and a first process result indicating a state of a post-processing processed body is learned; a data acquisition unit that acquires second log information measured during processing in a second substrate processing apparatus and a second process result indicating the state of the post-processing processed body; and a model correction unit that corrects the learned model on the basis of the second log information and the second process result, is provided.SELECTED DRAWING: Figure 7

Description

[Technical field]

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

[0002] There is known a technique for predicting the state of a processed object after processing from log information of a processing device that processes the object. For example, Patent Document 1 discloses an inference device capable of highly accurate inference regardless of the target of application. The inference device disclosed in Patent Document 1 adjusts each output data obtained by processing measurement data measured in association with the processing of the object using multiple network units that have been machine-learned in advance in a predetermined processing unit of a manufacturing process, and outputs an inference result by synthesizing each output data after adjustment. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] International Publication No. 2021 / 106646 Summary of the Invention [Problem to be solved by the invention]

[0004] The present disclosure provides techniques to improve the accuracy of predicting process outcomes. [Means for solving the problem]

[0005] According to one aspect of the present disclosure, there is provided an information processing apparatus comprising: a model acquisition unit that acquires a trained model that has learned a relationship between first log information measured during processing in a first substrate processing apparatus and a first process result indicating a state of the processed object after processing; a result acquisition unit that acquires second log information measured during processing in a second substrate processing apparatus and a second process result indicating a state of the processed object after processing; and a model correction unit that corrects the trained model based on the second log information and the second process result. Effect of the Invention

[0006] According to one aspect, the accuracy of predicting process results is improved. [Brief description of the drawings]

[0007] [Figure 1] FIG. 13 is a diagram illustrating an example of a flow of porting a reference model to another device. [Diagram 2] FIG. 13 is a diagram illustrating an example of a constant shift. [Diagram 3] FIG. 13 is a diagram illustrating an example of a sensitivity difference. [Figure 4] 1 is a block diagram showing an example of an overall configuration of a substrate processing system; [Diagram 5] 1 is a schematic cross-sectional view showing an example of a substrate processing apparatus. [Figure 6] FIG. 2 is a block diagram showing an example of a hardware configuration of a virtual measurement device. [Figure 7] 2 is a block diagram showing an example of a functional configuration of the substrate processing system; [Figure 8] 13 is a flowchart illustrating an example of a learning process. [Figure 9] 11 is a flowchart illustrating an example of a correction process. [Figure 10] 13 is a flowchart illustrating an example of a measurement process. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0008] Hereinafter, an embodiment of the present disclosure will be described with reference to the drawings. In the drawings, the same components are denoted by the same reference numerals, and duplicated descriptions may be omitted.

[0009] [Embodiment] An embodiment of the present disclosure is a substrate processing system including a substrate processing apparatus for processing a substrate, which is an example of a processing target object. In this embodiment, the substrate processing system includes a virtual measurement device that predicts a process result based on log information of the substrate processing apparatus.

[0010] Conventionally, sampling inspection using monitor wafers has been performed for quality control of semiconductor wafers, which are an example of substrates. If the process results are predicted using a virtual measurement device, virtual 100% inspection and immediate inspection after processing become possible. If the virtual measurement device predicts the process results at each point on the semiconductor wafer, it becomes possible to inspect the uniformity within the wafer. Furthermore, based on the predicted process results, it is possible to quickly detect abnormalities in substrate processing equipment.

[0011] The virtual measurement device predicts the process result from the log information using a prediction model that has learned the relationship between the log information and the process result through machine learning. In order to create a prediction model, it is necessary to perform multiple experiments while changing the control knob and collect training data. However, the relationship between the log information and the process result may vary from device to device. Therefore, if a trained prediction model based on training data collected from one substrate processing device is applied to another substrate processing device as is, the prediction accuracy may decrease.

[0012] On the other hand, it is not realistic to perform a sufficient number of experiments to create a prediction model for each substrate processing apparatus to be measured by the virtual measuring apparatus, and it is particularly difficult to perform experiments by suspending the operation of a substrate processing apparatus that is already in operation for a long period of time.

[0013] In this embodiment, a trained prediction model (hereinafter also referred to as a "reference model") based on training data acquired in a certain substrate processing apparatus (hereinafter also referred to as a "reference apparatus") is transferred to another substrate processing apparatus (hereinafter also referred to as a "recipient apparatus"), and the reference model is corrected to be suitable for the recipient apparatus using a small amount of training data acquired in the recipient apparatus.

[0014] Fig. 1 is a diagram showing an example of a flow for porting a reference model to another device. As shown in Fig. 1, model porting is performed in two phases. The first phase is the creation of a reference model. The second phase is the porting of the model to the destination device.

[0015] In the first phase, an experiment is first performed on a reference device to obtain a sufficient number of log information and process results. The number of experiments should be at least 10 runs. Next, machine learning is performed on the training data including the log information and process results obtained on the reference device. This creates a reference model.

[0016] In the second phase, an experiment is first performed on the destination device to obtain a small amount of log information and process results. The number of experiments may be at most about 3 runs, or may be as little as 1 run. Next, the reference model is corrected using training data including the log information and process results obtained on the destination device. This creates a destination model. The destination model reflects the machine difference between the reference device and the destination device, and can accurately predict process results from the log information obtained on the destination device.

[0017] In this embodiment, the machine differences of the substrate processing apparatus are classified into "constant shift" and "sensitivity difference." The constant shift is a phenomenon in which the process results for each substrate ID have a similar degree of difference. The substrate ID is identification information for identifying the substrate within the substrate processing apparatus. For example, in the case of a batch apparatus that accommodates multiple substrates in a processing vessel and processes them simultaneously, the substrate ID is information indicating the position of the substrate within the processing vessel.

[0018] Fig. 2 is a diagram showing an example of a constant shift. The example shown in Fig. 2 is a graph in which the horizontal axis indicates wafer ID (an example of a substrate ID) and the vertical axis indicates film thickness (an example of a process result), and the process results of two substrate processing apparatuses (apparatus A and apparatus B) are plotted. As shown in Fig. 2, it can be seen that the difference in the process result for each wafer ID corresponds to the machine difference, and the machine difference for each wafer ID is approximately the same value.

[0019] The sensitivity difference is a phenomenon in which a difference appears in the amount of change in the process result when sensor data changes. The sensor data is data that can be measured by a sensor installed in the substrate processing apparatus, such as temperature, pressure, gas flow rate, etc.

[0020] Fig. 3 is a diagram showing an example of the sensitivity difference. The example shown in Fig. 3 is a graph in which the horizontal axis indicates a value measured by sensor A, which is an example of sensor data, and the vertical axis indicates film thickness, which is an example of process results, and the process results of two substrate processing apparatuses (apparatus A and apparatus B) are plotted. As shown in Fig. 3, the difference in the slope of the line showing the correlation between the sensor data and the process results for each substrate processing apparatus corresponds to the machine difference, and it can be seen that the change in the process result accompanying the change in sensor data differs for each substrate processing apparatus.

[0021] In this embodiment, the prediction model is constructed using a linear regression model, and the constant shift is corrected by correcting the intercept with the difference in process results for each substrate ID. Also, in this embodiment, the sensitivity difference is corrected by performing weighted regression on the teacher data acquired in each substrate processing apparatus. This allows the machine difference between the reference apparatus and the destination apparatus to be corrected in terms of both the constant shift and the sensitivity difference, and a prediction model suitable for the destination apparatus can be created.

[0022] <System configuration> The overall configuration of the substrate processing system in this embodiment will be described with reference to Fig. 4. Fig. 4 is a block diagram showing an example of the overall configuration of the substrate processing system in this embodiment.

[0023] 4, the substrate processing system 100 includes substrate processing apparatuses 120a1-120a3 and controllers 121a1-121a3 in a factory a. The substrate processing apparatuses 120a1-120a3 and the controllers 121a1-121a3 are connected to each other by wire or wirelessly.

[0024] Furthermore, the substrate processing system 100 includes substrate processing apparatuses 120b1 and 120b2 and controllers 121b1 and 121b2 in a factory b. The substrate processing apparatuses 120b1 and 120b2 and the controllers 121b1 and 121b2 are connected to each other by wire or wirelessly.

[0025] Furthermore, the substrate processing system 100 includes substrate processing apparatuses 120c1 and 120c2 and controllers 121c1 and 121c2 in a factory c. The substrate processing apparatuses 120c1 and 120c2 and the controllers 121c1 and 121c2 are connected to each other by wire or wirelessly.

[0026] The substrate processing apparatuses 120a1-120a3, 120b1, 120b2, 120c1, 120c2 are connected to the host apparatuses 110a, 110b, 110c via the networks N1-N3. The substrate processing apparatuses perform substrate processing under the control of the respective control devices based on instructions from the host apparatuses 110a, 110b, 110c. The host apparatuses 110a, 110b, 110c are connected to the server apparatus 150 via the network N4 such as the Internet.

[0027] In the following description, the substrate processing apparatuses 120a1 to 120a3, 120b1, 120b2, 120c1, and 120c2 are also collectively referred to as substrate processing apparatus 120. The control apparatuses 121a1 to 121a3, 121b1, 121b2, 121c1, and 121c2 are also collectively referred to as control apparatus 121. The host apparatuses 110a, 110b, and 110c are also collectively referred to as host apparatus 110.

[0028] It is assumed that the substrate processing apparatuses 120a1 to 120a3, the substrate processing apparatuses 120b1 and 120b2, and the substrate processing apparatuses 120c1 and 120c2 store various data that they manage within their own apparatuses.

[0029] The virtual measurement device 140a is connected to the substrate processing apparatus 120a1 by wire or wirelessly. The virtual measurement device 140b is connected to the substrate processing apparatus 120b1 by wire or wirelessly. The virtual measurement device 140a is an example of a reference device. The virtual measurement device 140b is an example of a destination device. The example of FIG. 4 shows a state in which the virtual measurement device 140a is connected to the substrate processing apparatus 120a1 and the virtual measurement device 140b is connected to the substrate processing apparatus 120b1, but is not limited to this. The virtual measurement device 140a and the virtual measurement device 140b may be connected to different substrate processing apparatuses 120.

[0030] The substrate processing system 100 shown in Fig. 4 is an example, and it goes without saying that there are various system configuration examples depending on the application and purpose. The classification of devices such as the host device 110, the substrate processing device 120, the control device 121, the virtual measurement device 140, and the server device 150 shown in Fig. 4 is an example. For example, the number of factories, the number of host devices 110, the number of substrate processing device 120, the number of control devices 121, the number of virtual measurement devices 140, etc. are examples and are not limited to these.

[0031] For example, the substrate processing system 100 may have various configurations, such as a configuration in which at least two of the host device 110, the substrate processing apparatus 120, the control device 121, the virtual measurement device 140, and the server device 150 are integrated, or a configuration in which they are further divided. For example, the control device 121 may collectively control a plurality of substrate processing apparatuses 120, may be provided in a one-to-one correspondence with the substrate processing apparatus 120, or may be integrated with the substrate processing apparatus 120.

[0032] The virtual measurement device 140 may be realized by the host device 110 or may be realized by the server device 150. In this case, the virtual measurement device 140 is not necessary. The virtual measurement device 140 may also be realized by the control device 121. The virtual measurement device 140 may also be realized by a control device (not shown) that collectively controls a plurality of control devices 121.

[0033] <Substrate processing equipment> An example of a substrate processing apparatus according to the present embodiment will be described with reference to Fig. 5. Fig. 5 is a schematic cross-sectional view showing a vertical heat treatment apparatus which is an example of the substrate processing apparatus according to the present embodiment.

[0034] The vertical heat treatment apparatus 120 in this embodiment is a substrate treatment apparatus that accommodates a large number of semiconductor wafers W, which are an example of objects to be treated, at one time and performs heat treatment such as oxidation, diffusion, low-pressure CVD, etc. As shown in Fig. 5, the vertical heat treatment apparatus 120 includes a treatment vessel 10, a gas supply unit 20, an exhaust unit 30, a heating unit 40, a cooling unit 50, a control device 121, etc.

[0035] The processing vessel 10 has a substantially cylindrical shape. The processing vessel 10 includes an inner tube 11, an outer tube 12, a manifold 13, an injector 14, a gas outlet 15, a lid 16, etc. The inner tube 11 has a substantially cylindrical shape. The outer tube 12 has a substantially cylindrical shape with a ceiling, and the inner tube 11 and the outer tube 12 form a double-tube structure. The inner tube 11 and the outer tube 12 are formed of a heat-resistant material such as quartz.

[0036] The manifold 13 has a substantially cylindrical shape. The manifold 13 supports the lower ends of the inner tube 11 and the outer tube 12. The manifold 13 is formed of, for example, stainless steel. The injector 14 passes through the manifold 13 and extends horizontally into the inner tube 11, and also bends in an L-shape inside the inner tube 11 and extends upward. The base end of the injector 14 is connected to the gas introduction pipe 24, and the tip is open. The injector 14 discharges a process gas (hereinafter also simply referred to as "gas") introduced via the gas introduction pipe 24 into the inner tube 11 from the opening at the tip. There may be multiple injectors 14.

[0037] The gas outlet 15 is formed in the manifold 13. The process gas is exhausted by the exhaust unit 30 through the gas outlet 15. The lid 16 airtightly closes the opening at the lower end of the manifold 13. The lid 16 is formed of, for example, stainless steel. A wafer boat (substrate holder) 18 is placed on the lid 16 via a heat-insulating tube 17. The heat-insulating tube 17 and the wafer boat 18 are formed of, for example, a heat-resistant material such as quartz.

[0038] The wafer boat 18 holds a plurality of semiconductor wafers W substantially horizontally at a predetermined interval in the vertical direction. The wafer boat 18 is loaded into the processing vessel 10 by the lifting mechanism 19 lifting the lid 16, and is accommodated in the processing vessel 10. The wafer boat 18 is unloaded from the processing vessel 10 by the lifting mechanism 19 lowering the lid 16.

[0039] The gas supply unit 20 includes a gas source 21, an IGS (Integrated Gas System) 22, an external pipe 23, and a gas introduction pipe 24. The gas source 21 is a supply source of processing gas, and includes, for example, a film formation gas source, a cleaning gas source, and a purge gas source. The IGS 22 is an integrated circuit of gas pipes, and integrates a group of pipes connected to the film formation gas source, cleaning gas source, purge gas source, etc. of the gas source 21. A flow rate control unit is installed in the IGS 22, and controls the flow rate of gas flowing through each pipe. The flow rate control unit includes, for example, a mass flow controller, and an opening and closing valve.

[0040] The IGS 22 is connected to an external pipe 23. The external pipe 23 is connected to a gas introduction pipe 24. A heater (not shown) is wound around the outer periphery of the external pipe 23 to heat the external pipe 23. The gas introduction pipe 24 is connected to the processing vessel 10 and introduces gas into the processing vessel 10. That is, the flow rate of the processing gas from the gas source 21 is controlled by a flow rate control unit in the IGS 22, and the processing gas is heated when passing through the external pipe 23 and flows into the gas introduction pipe 24, and is supplied from the gas introduction pipe 24 into the processing vessel 10 via the injector 14. The injector 14 functions as a gas inlet of the processing vessel 10.

[0041] A gas piping joint 82 connected to the gas introduction pipe 24 is provided near the gas inlet of the processing vessel 10. The temperature sensor 80 is configured to pass through the joint 82. The temperature sensor 80 is configured to measure the temperature of the gas in the gas introduction pipe 24. The temperature sensor 80 transmits the measured temperature to the control device 121. In addition, a second heater 81 is disposed in the gas introduction pipe 24, and the second heater 81 is configured to heat the gas in the gas introduction pipe 24.

[0042] The exhaust section 30 includes an exhaust device 31, an exhaust pipe 32, and a pressure controller 33. The exhaust device 31 is, for example, a vacuum pump such as a dry pump or a turbo molecular pump. The pressure controller 33 is disposed in the exhaust pipe 32, and controls the pressure inside the processing vessel 10 by adjusting the conductance of the exhaust pipe 32. The pressure controller 33 is, for example, an automatic pressure control valve.

[0043] The heating unit 40 includes a heat insulating material 41, a first heater 42, and an outer skin 43. The heat insulating material 41 has a substantially cylindrical shape and is provided around the outer tube 12. The heat insulating material 41 is formed mainly of silica and alumina. The first heater 42 has a linear shape and is provided in a spiral or serpentine shape on the inner circumference of the heat insulating material 41. The first heater 42 is configured so that the temperature can be controlled by dividing the heat insulating material 41 into a plurality of zones in the height direction of the processing vessel 10. The outer skin 43 is provided so as to cover the outer circumference of the heat insulating material 41. The outer skin 43 retains the shape of the heat insulating material 41 and reinforces the heat insulating material 41. The outer skin 43 is formed of a metal such as stainless steel. In addition, a water-cooled jacket (not shown) may be provided around the outer skin 43 to suppress the thermal influence of the heating unit 40 on the outside. The heating unit 40 heats the inside of the processing vessel 10 by the heat generated by the first heater 42.

[0044] The cooling unit 50 supplies a cooling fluid toward the processing vessel 10 to cool the semiconductor wafer W in the processing vessel 10. The cooling fluid may be, for example, air. The cooling unit 50 supplies the cooling fluid toward the processing vessel 10 when, for example, the semiconductor wafer W is rapidly cooled after heat treatment. The cooling unit 50 has a fluid flow path 51, an outlet hole 52, a distribution flow path 53, a flow rate adjustment unit 54, and a heat exhaust port 55.

[0045] A plurality of fluid flow paths 51 are formed in the height direction between the insulation 41 and the outer skin 43. The fluid flow paths 51 are, for example, flow paths formed along the circumferential direction on the outside of the insulation 41. The blowing holes 52 are formed penetrating the insulation 41 from each fluid flow path 51, and blow out the cooling fluid into the space between the outer pipe 12 and the insulation 41. The distribution flow paths 53 are provided outside the outer skin 43, and distribute and supply the cooling fluid to each fluid flow path 51. The flow rate adjustment unit 54 is interposed in the distribution flow path 53, and adjusts the flow rate of the cooling fluid supplied to the fluid flow paths 51.

[0046] The heat exhaust port 55 is provided above the multiple blowing holes 52, and exhausts the cooling fluid supplied to the space between the outer pipe 12 and the insulating material 41 to the outside of the processing vessel 10. The cooling fluid exhausted to the outside of the processing vessel 10 is cooled, for example, by a heat exchanger, and supplied again to the distribution flow path 53. However, the cooling fluid exhausted to the outside of the processing vessel 10 may be exhausted without being reused.

[0047] The temperature sensor 60 detects the temperature inside the processing vessel 10. The temperature sensor 60 is provided, for example, inside the inner tube 11. However, the temperature sensor 60 may be provided at a position where it can detect the temperature inside the processing vessel 10, for example, in the space between the inner tube 11 and the outer tube 12. The temperature sensor 60 has, for example, multiple temperature measuring parts provided at different positions in the height direction corresponding to multiple zones. The multiple temperature measuring parts may be, for example, thermocouples or resistance temperature detectors. The temperature sensor 60 transmits the temperatures detected by the multiple temperature measuring parts to the control device 121.

[0048] The control device 121 controls the operation of the vertical heat treatment device 120, thereby controlling the semiconductor process performed in the vertical heat treatment device 120. The control device 121 may be, for example, a computer.

[0049] <Hardware configuration> The host device 110, the control device 121, the virtual measurement device 140, and the server device 150 included in the substrate processing system 100 shown in Fig. 4 are realized by a computer having a hardware configuration as shown in Fig. 6. Fig. 6 is a block diagram showing an example of the hardware configuration of the computer in this embodiment.

[0050] 6, a computer 500 in this embodiment 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 connected to each other via a bus B. The input device 501 and the output device 502 may be connected and used when necessary.

[0051] The input device 501 is a keyboard, mouse, touch panel, etc., and is used by an operator to input various 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 a network. The HDD 508 is an example of a non-volatile storage device that stores programs and data.

[0052] The external I / F 503 is an interface with an external device. The computer 500 can read and / or write data from and to a recording medium 503a such as a Secure Digital (SD) memory card via the external I / F 503. The ROM 505 is an example of a non-volatile semiconductor memory (storage device) in which programs and data are stored. The RAM 504 is an example of a volatile semiconductor memory (storage device) that temporarily holds programs and data.

[0053] The CPU 506 is a calculation device that reads out programs and data from storage devices such as the ROM 505 and the HDD 508 onto the RAM 504 and executes processes, thereby realizing overall control and functions of the computer 500 .

[0054] <Functional configuration> The functional configuration of the virtual measurement device in this embodiment will be described with reference to Fig. 7. Fig. 7 is a diagram showing an example of the functional configuration of the virtual measurement device in this embodiment.

[0055] <Virtual measuring device 140a (reference device)> As shown in FIG. 7, the virtual measurement device 140a in this embodiment includes a data acquisition unit 201, a model generation unit 202, a model storage unit 203, a result prediction unit 204, and an anomaly detection unit 205.

[0056] The model storage unit 203 is realized, for example, by the RAM 504 or the HDD 508 shown in Fig. 6. The data acquisition unit 201, the model generation unit 202, the result prediction unit 204, and the anomaly detection unit 205 are realized, for example, by the CPU 506 shown in Fig. 6 executing a program loaded on the RAM 504.

[0057] The data acquiring unit 201 acquires log information (an example of first log information) and a process result (an example of a first process result) from the substrate processing apparatus 120a1 (an example of a first substrate processing apparatus).

[0058] The log information is sensor data measured while a process is being performed in the substrate processing apparatus 120a1 (in other words, while the workpiece is being processed). In the present embodiment, the log information may include, for example, temperature, gas flow rate, accumulated film thickness, plasma output, etc. The log information is not limited to these, and may include any data that can be measured by a sensor included in the substrate processing apparatus 120 (for example, the temperature sensor 60, the temperature sensor 80, etc. shown in FIG. 5).

[0059] The process result is measurement data obtained by measuring the state of the processed substrate after the process is completed in the substrate processing apparatus 120a1. In the present embodiment, the process result may include, for example, a film thickness, a refractive index (RI), a film stress, a film density, a film etching rate, a particle count, and the like. The process result is not limited to these, and may include any data that can be measured from the substrate after the process in the substrate processing apparatus 120a1.

[0060] The model generation unit 202 generates a reference model based on the log information and the process results acquired by the data acquisition unit 201. The model generation unit 202 performs machine learning on the relationship between the log information and the process results using training data including the log information and the process results. In this embodiment, the model generation unit 202 performs regression analysis on the training data to generate a linear regression model as the reference model.

[0061] A reference model is stored in the model storage unit 203. The reference model is a trained prediction model generated by the model generation unit 202.

[0062] Based on the reference model read out from the model storage unit 203, the result prediction unit 204 predicts a process result (an example of a third process result) from the log information (an example of a third log information) acquired from the substrate processing apparatus 120a1.

[0063] The anomaly detection unit 205 detects an anomaly in the substrate processing apparatus 120a1 based on the predicted result of the process result predicted by the result prediction unit 204. The anomaly detection unit 205 may detect an anomaly in the substrate processing apparatus 120a1, for example, by comparing items included in the process result with a predetermined threshold value. The anomaly detection unit 205 may detect an anomaly in the substrate processing apparatus 120a1 based on a predetermined rule or a machine learning model.

[0064] When the abnormality detection unit 216 detects an abnormality in the substrate processing apparatus 120a1, it notifies a predetermined notification destination of the detection of the abnormality. The notification destination may be, for example, an external monitoring device, or a display of the virtual measurement device 140a or the control device 121a1.

[0065] <Virtual measurement device 140b (transplantation device)> 7, the virtual measurement device 140b in this embodiment includes a model acquisition unit 211, a data acquisition unit 212, a model correction unit 213, a model storage unit 214, a result prediction unit 215, and an anomaly detection unit 216. The model correction unit 213 includes a weighted regression unit 221 and an intercept correction unit 222.

[0066] The model storage unit 214 is realized, for example, by the RAM 504 or the HDD 508 shown in Fig. 6. The model acquisition unit 211, the data acquisition unit 212, the model correction unit 213, the result prediction unit 215 and the anomaly detection unit 216 are realized, for example, by the CPU 506 shown in Fig. 6 executing a program loaded on the RAM 504.

[0067] The model acquisition unit 211 acquires the reference model from the virtual measurement device 140a. The model acquisition unit 211 may read out the reference model stored in the virtual measurement device 140a, or may acquire the reference model by receiving the reference model transmitted from the virtual measurement device 140a.

[0068] The data acquiring unit 212 acquires log information (an example of second log information) and process results (an example of second process results) from the substrate processing apparatus 120b1 (an example of a second substrate processing apparatus). The log information is sensor data measured while a process is being performed in the substrate processing apparatus 120b1. The process results are measurement data obtained by measuring the state of a processed substrate after the process is completed in the substrate processing apparatus 120b1. Items included in the log information and process results are similar to those acquired by the data acquiring unit 201 of the virtual measurement apparatus 140a.

[0069] The model correction unit 213 corrects the reference model acquired by the model acquisition unit 211, based on the log information and process results acquired by the data acquisition unit 212. The model correction unit 213 executes the weighted regression unit 221 and the intercept correction unit 222, respectively, to generate a corrected prediction model (transplantation model).

[0070] The weighted regression unit 221 corrects the sensitivity difference between the substrate processing apparatus 120a1 and the substrate processing apparatus 120b1. Specifically, the weighted regression unit 221 first acquires teacher data (an example of first teacher data) when a reference model is generated from the virtual measurement device 140a. The teacher data includes log information and process results acquired by the data acquisition unit 201. Next, the weighted regression unit 221 generates teacher data (an example of second teacher data) including the log information and process results acquired by the data acquisition unit 212.

[0071] Next, the weighted regression unit 221 determines the weights to be given to the first teacher data and the second teacher data. At this time, the weighted regression unit 221 sets the weight to be given to the second teacher data to be greater than the weight to be given to the first teacher data. Then, the weighted regression unit 221 performs weighted regression on the first teacher data and the second teacher data. As a result, a new prediction model in which the parameters of the reference model are updated is generated.

[0072] The intercept correction unit 222 corrects the constant shift between the substrate processing apparatus 120a1 and the substrate processing apparatus 120b1. Specifically, the intercept correction unit 222 first predicts the process result from the log information acquired by the data acquisition unit 212 based on the reference model. Next, the intercept correction unit 222 calculates the difference between the predicted result of the process result and the process result acquired by the data acquisition unit 212. Then, the intercept correction unit 222 corrects the intercept of the reference model by the difference between the predicted result and the process result.

[0073] The transplantation model is stored in the model storage unit 214. The transplantation model is a corrected prediction model generated by the model correction unit 213.

[0074] Based on the destination model read from the model storage unit 214, the result prediction unit 215 predicts a process result (an example of a third process result) from log information (an example of a third log information) acquired from the substrate processing apparatus 120b1.

[0075] The abnormality detection unit 216 detects an abnormality in the substrate processing apparatus 120b1 based on the prediction result of the process result predicted by the result prediction unit 215. When the abnormality detection unit 216 detects an abnormality in the substrate processing apparatus 120b1, it notifies a predetermined notification destination that the abnormality has been detected.

[0076] <Processing Procedure> An information processing method executed by the substrate processing system 100 in this embodiment will be described with reference to Fig. 8 to Fig. 10. The information processing method in this embodiment includes a learning process (see Fig. 8), a correction process (see Fig. 9), and a measurement process (see Fig. 10).

[0077] Learning process 8 is a flowchart showing an example of the learning process in this embodiment. The learning process is a process for generating a reference model.

[0078] In step S1, the data acquisition unit 201 of the virtual measurement apparatus 140a acquires log information from the substrate processing apparatus 120a1. Next, the data acquisition unit 201 sends the acquired log information to the model generation unit 202.

[0079] In step S2, the data acquiring unit 201 of the virtual measurement apparatus 140a acquires the process result from the substrate processing apparatus 120a1. Next, the data acquiring unit 201 sends the acquired process result to the model generating unit 202.

[0080] In step S3, the model generation unit 202 of the virtual measurement device 140a receives the log information and the process result from the data acquisition unit 201. Next, the model generation unit 202 generates teacher data including the log information and the process result. Then, the model generation unit 202 learns the relationship between the log information and the process result using the teacher data. As a result, a reference model is generated.

[0081] In step S4, the model generating unit 202 of the virtual measurement device 140a stores in the model storage unit 203 the reference model generated in step S3.

[0082] <Correction processing> 9 is a flowchart showing an example of the correction process in this embodiment. The correction process is a process for correcting a reference model and generating a destination model.

[0083] In step S11, the model acquisition unit 211 of the virtual measurement device 140b acquires a reference model from the virtual measurement device 140a. Next, the model acquisition unit 211 sends the acquired reference model to the model correction unit 213.

[0084] In step S12, the data acquiring unit 212 of the virtual measurement apparatus 140b acquires log information from the substrate processing apparatus 120b1. Next, the data acquiring unit 212 sends the acquired log information to the model correcting unit 213.

[0085] In step S13, the data acquiring unit 212 of the virtual measurement apparatus 140b acquires the process result from the substrate processing apparatus 120b1. Next, the data acquiring unit 212 sends the acquired process result to the model correcting unit 213.

[0086] In step S14, the model corrector 213 of the virtual measurement device 140b receives the reference model from the model acquirer 211. In addition, the model corrector 213 receives the log information and the process result from the data acquirer 212.

[0087] The model correction unit 213 executes the weighted regression unit 221. The weighted regression unit 221 acquires first teacher data when a reference model is generated from the virtual measurement device 140a. Next, the weighted regression unit 221 generates second teacher data including the log information and the process result received from the data acquisition unit 212. Next, the weighted regression unit 221 determines weights to be given to the first teacher data and the second teacher data. Then, the weighted regression unit 221 performs weighted regression on the first teacher data and the second teacher data.

[0088] In step S15, the model correction unit 213 of the virtual measurement device 140b executes the intercept correction unit 222. The intercept correction unit 222 predicts a process result from the log information received from the data acquisition unit 212 based on the reference model acquired in step S11. Next, the intercept correction unit 222 calculates the difference between the predicted result of the process result and the process result received from the data acquisition unit 212. Then, the intercept correction unit 222 corrects the intercept of the prediction model generated in step S14 by the difference between the predicted result and the process result.

[0089] In step S16, the model corrector 213 of the virtual measurement device 140b acquires the prediction model corrected in step S15 as a transplantation model. Then, the model corrector 213 stores the acquired transplantation model in the model storage unit 214.

[0090] <Measurement processing> 10 is a flowchart showing an example of a measurement process in this embodiment. The measurement process is a process of virtually measuring a process result from log information based on a prediction model. In other words, the measurement process is a process of predicting a process result based on a prediction model.

[0091] In the following description, an example of performing the measurement process on a destination device will be described, but the measurement process may be performed on a reference device. When performing the measurement process on a reference device, the virtual measurement device 140b in the following description may be replaced with the virtual measurement device 140a.

[0092] In step S21, the data acquisition unit 212 of the virtual measurement apparatus 140b acquires log information from the substrate processing apparatus 120b1. Next, the data acquisition unit 212 sends the acquired log information to the result prediction unit 215.

[0093] In step S22, the result prediction unit 215 of the virtual measurement device 140b receives the log information from the data acquisition unit 212. Next, the result prediction unit 215 reads out the destination model from the model storage unit 214. Then, the result prediction unit 215 inputs the log information to the destination model. The destination model predicts the process result based on the input log information, and outputs the prediction result.

[0094] The result prediction unit 215 acquires the prediction result of the process result output from the transplantation destination model. The result prediction unit 215 sends the acquired prediction result to the anomaly detection unit 216. The result prediction unit 215 may display the prediction result of the process result on the output device 502 or the like. The result prediction unit 215 may accumulate the prediction result of the process result in the HDD 508 or the like.

[0095] In step S23, the abnormality detection unit 216 of the virtual measurement equipment 140b receives the predicted result of the process result from the result prediction unit 215. Next, the abnormality detection unit 216 performs abnormality detection of the substrate processing apparatus 120b1 based on the predicted result of the process result.

[0096] If an abnormality is detected in the substrate processing apparatus 120b1 (YES), the abnormality detection unit 216 proceeds to step S24. On the other hand, if an abnormality is not detected in the substrate processing apparatus 120b1 (NO), the abnormality detection unit 216 skips step S24 and ends the measurement process.

[0097] In step S24, the abnormality detection unit 216 of the virtual measurement apparatus 140b notifies a predetermined notification destination that an abnormality has been detected in the substrate processing apparatus 120b1. For example, the abnormality detection unit 216 may notify an external monitoring device of the abnormality detection via the communication I / F 507. Also, for example, the abnormality detection unit 216 may display the abnormality detection on the output device 502.

[0098] <Effects of the embodiment> The virtual measurement apparatus 140b in this embodiment corrects a trained model that has learned the relationship between the log information acquired by the substrate processing apparatus 120a and the process result, based on the log information and the process result acquired by the substrate processing apparatus 120b. The trained model after the correction reflects the machine difference between the substrate processing apparatus 120a and the substrate processing apparatus 120b, and the process result can be accurately predicted from the log information acquired by the substrate processing apparatus 120b. Therefore, according to this embodiment, the prediction accuracy of the process result is improved.

[0099] In the virtual measurement apparatus 140b in this embodiment, the trained model is a linear regression model, and the intercept of the trained model may be corrected by the difference between the predicted result of the process result predicted based on the trained model from the log information acquired in the substrate processing apparatus 120b and the process result acquired in the substrate processing apparatus 120b. Therefore, according to this embodiment, a trained model in which the constant shift between the substrate processing apparatus 120a and the substrate processing apparatus 120b is corrected can be obtained, and the prediction accuracy of the process result is improved.

[0100] The virtual measurement apparatus 140b in this embodiment may perform weighted regression on teacher data including log information and process results acquired by the substrate processing apparatus 120a and teacher data including log information and process results acquired by the substrate processing apparatus 120b. In this case, the weight given to the teacher data acquired by the substrate processing apparatus 120b may be larger than the weight given to the teacher data acquired by the substrate processing apparatus 120a. Therefore, according to this embodiment, a trained model in which the sensitivity difference between the substrate processing apparatus 120a and the substrate processing apparatus 120b is corrected can be obtained, and the prediction accuracy of the process result is improved.

[0101] The virtual measurement device 140b in this embodiment may predict a process result from log information acquired by the substrate processing apparatus 120b based on the trained model. Therefore, according to this embodiment, it is possible to accurately perform a virtual 100% inspection of the objects processed by the substrate processing apparatus 120b or a real-time inspection of a predetermined process.

[0102] The virtual measurement device 140b in this embodiment may detect an abnormality in the substrate processing apparatus 120b based on the predicted result of the process result. Therefore, according to this embodiment, an abnormality in the substrate processing apparatus 120b can be quickly detected.

[0103] [supplement] The information processing apparatus and substrate processing apparatus according to the embodiments disclosed herein are illustrative in all respects and are not restrictive. The embodiments can be modified and improved in various forms without departing from the scope and spirit of the appended claims. The matters described in the above embodiments can be configured in other ways as long as they are not inconsistent, and can be combined as long as they are not inconsistent.

[0104] The substrate processing apparatus for performing the process including the substrate processing method of the present disclosure is not limited to a thermal processing apparatus, and may be 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).

[0105] The substrate processing apparatus of the present disclosure can be applied to both plasma-using and non-plasma-using apparatuses that perform a predetermined process (e.g., film formation process, etching process, etc.) on a substrate. The substrate processing apparatus of the present disclosure can be applied to any of single-wafer apparatuses that process substrates one by one, batch apparatuses that process multiple substrates at once, and semi-batch apparatuses that process a smaller number of substrates at once than the number processed at once by a batch apparatus. [Explanation of symbols]

[0106] 100 Substrate processing system 110 Host device 120 Substrate Processing Equipment 121 Control device 140 Virtual Measuring Device 150 Server device 201 Data Acquisition Department 202 Model Generation Unit 203 Model Memory Unit 204 Result Prediction Department 205 Anomaly detection unit 211 Model Acquisition Department 212 Data Acquisition Section 213 Model Correction Section 214 Model Memory Section 215 Result Prediction Department 216 Anomaly detection unit 221 Weighted Regression Section 222 Intercept correction section

Claims

1. a model acquisition unit configured to acquire a trained model that has learned a relationship between first log information measured during processing in the first substrate processing apparatus and a first process result indicating a state of the processed object after processing; a data acquisition unit configured to acquire second log information measured during processing in the second substrate processing apparatus and a second process result indicating a state of the processed object after processing; A model correction unit configured to correct the trained model based on the second log information and the second process result; An information processing device comprising:

2. 2. The information processing device according to claim 1, The model correction unit is configured to correct a prediction result of the trained model based on a difference between a prediction result obtained by predicting a process result from the second log information based on the trained model and the second process result. Information processing device.

3. 3. The information processing device according to claim 2, The trained model is a linear regression model, The model correction unit is configured to correct an intercept of the trained model by the difference. Information processing device.

4. 2. The information processing device according to claim 1, The trained model is a linear regression model, the model correction unit is configured to perform weighted regression on first teacher data including the first log information and the first process result, and second teacher data including the second log information and the second process result. Information processing device.

5. 5. The information processing device according to claim 4, The weight given to the second teacher data is greater than the weight given to the first teacher data. Information processing device.

6. 6. An information processing device according to claim 1, and a result prediction unit configured to predict a third process result indicating a state of the target object after the processing from third log information measured during the processing in the second substrate processing apparatus based on the learned model corrected by the model correction unit. Information processing device.

7. 7. The information processing device according to claim 6, an anomaly detection unit configured to detect an anomaly in the second substrate processing apparatus based on a result of the third process; Information processing device.

8. A substrate processing apparatus, comprising: a model acquisition unit configured to acquire a trained model that has learned a relationship between first log information measured during processing in another substrate processing apparatus different from the substrate processing apparatus and a first process result indicating a state of the processed object after processing; a data acquisition unit configured to acquire second log information measured during processing in the substrate processing apparatus and a second process result indicating a state of the target object after processing; A model correction unit configured to correct the trained model based on the second log information and the second process result; A substrate processing apparatus comprising:

9. An information processing device, A step of acquiring a trained model that has learned a relationship between first log information measured during processing in the first substrate processing apparatus and a first process result indicating a state of the processed object after processing; acquiring second log information measured during processing in the second substrate processing apparatus and second process results indicating a state of the processed object after processing; correcting the trained model based on the second log information and the second process result; An information processing method for performing the above.

Images