JP2010282541A - Method for automatically generating reference waveform of device data and method for predicting failure information - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To automatically generate a reference waveform from process data, to update the reference waveform, to analyze the device data with higher accuracy, thus, to further reduce defective products which may be produced when the device data fails by determining, for example, where the failure occurs, and to predict failure information such as the type or location of the failure. <P>SOLUTION: The method for automatically generating the reference waveform of the device data is configured to: monitor operations and manufacturing conditions of a manufacturing processor to obtain the device data; acquire waveform data from the device data; select reference waveform for non-defective product or reference waveform for defective product to be generated depending on an inspection result from an inspection device; when the acquired waveform data matches the original reference waveform for non-defective product or the original reference waveform for defective product within a predetermined range, automatically update the reference waveform by superposing the waveform data; and when they do not match, register the waveform data as a new reference waveform. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention monitors the operation of a manufacturing processing apparatus such as a semiconductor or a liquid crystal panel, collects operation data and manufacturing condition data of the manufacturing processing apparatus, analyzes the operation and manufacturing status of the manufacturing apparatus from the collected data, and manufactures The present invention relates to a method for improving the reliability and productivity of a processing apparatus.

  As a background art, a case of a color filter used for a liquid crystal panel will be described as an example. FIG. 1 is a cross-sectional view showing an example of a color filter used in a color liquid crystal display device. The color filter 1 includes a black matrix (hereinafter referred to as BM) 3, a red R colored pixel (hereinafter referred to as R pixel) 4a, a green G colored pixel (hereinafter referred to as G pixel) 4b, and a blue B colored pixel on a glass substrate 2. (Hereinafter referred to as a B pixel) 4c, a transparent electrode 5, a photo spacer (hereinafter referred to as PS) 6, and a vertical alignment (hereinafter referred to as VA) 7 are sequentially formed.

  As a method for manufacturing a color filter having the above structure, a photolithography method, a printing method, and an ink jet method are known. FIG. 2 is a flowchart showing steps of a commonly used photolithography method. In the color filter, first, a process of forming a black matrix (BM) on the glass substrate (C-1), a process of cleaning the glass substrate (C-2), and a process of applying and pre-drying a colored photoresist (C-3), prebaking step (C-4) for drying and curing the colored photoresist, exposure step (C-5), developing step (C-6), curing the colored photoresist The step (C-7), the step (C-8) for forming a film on the transparent electrode, and the step (C-9) for forming PS and VA are performed in this order.

  For example, when pixels are formed in the order of R pixel, G pixel, and B pixel, from the step (C-2) of cleaning the color filter glass substrate to the step of curing the colored photoresist (C- In 7), the color resist is changed in the order of red R, green G, and blue B, and the process is repeated three times to form R, G, and B pixels.

  Examples of defects that occur in the color filter manufacturing process include foreign matter adhesion in the cleaning process, colored resist application unevenness and pinholes in the colored resist application and preliminary drying processes, and excessive and insufficient exposure and mask alignment in the exposure process. There are defects such as over and under development and uneven development in the development process.

  In order to suppress the occurrence of the above-described defects and to detect the generated defective products, the operation status and manufacturing conditions of the manufacturing processing apparatus are monitored, and the defective products are detected by the inspection apparatus to guarantee the quality.

  In manufacturing processing apparatuses (hereinafter referred to as process apparatuses) in the manufacture of semiconductors and liquid crystal panels, process apparatus operation data (hereinafter referred to as operation data) indicating the operation status of the process apparatus, manufacturing conditions such as flow rate, temperature, and pressure of the cleaning liquid In other words, there is an EES (Equipment Engineering System) that collects and analyzes process condition data (hereinafter referred to as process data) at a high rate of several ms. EES collects equipment reference data (operation data and process data) from process equipment, checks whether the manufacturing process is performed under normal operation and normal process conditions, and checks the reliability of the process equipment. It is a system that improves productivity.

As a process abnormality analysis method using EES, for example, there are methods described in the following documents.

JP-A 62-42204 JP 2005-197323 A JP 2004-186445 A

  The process abnormality analysis method described in Patent Document 1 detects an abnormality from a difference from process data during normal operation, determines whether there is an abnormality, and determines the cause of the abnormality.

  The process abnormality analysis method described in Patent Document 2 extracts feature quantities (average, maximum, minimum, etc.) from process data cut out in a certain section, creates a process-quality model, determines whether there is an abnormality, This is to predict the occurrence of a failure.

  The process abnormality analysis method described in Patent Document 3 uses a feature amount extracted from process data to create a process-quality model from analysis by data mining or multivariate analysis, and determines whether there is an abnormality. .

  In the process abnormality analysis method using EES, as a method of selecting process data during normal operation described in Patent Document 1, an operator selects data during normal operation from past data and sets it as a reference waveform. do. However, since there is a variation for each operation of the apparatus even during normal operation data, the difference from the collected process data cannot be obtained accurately due to the influence of the variation.

  The method described in Patent Document 2 can determine whether or not there is an abnormality, but cannot predict what abnormality is occurring.

  Further, the method described in Patent Document 3 can determine whether or not there is an abnormality, and since inspection data is used, it is possible to predict what abnormality has occurred. However, in general, process data varies even when a non-defective product is made, and thus the obtained process data is not necessarily accurate.

  Therefore, the present invention automatically creates and updates the reference waveform from the process data, analyzes the device data with higher accuracy, and may possibly occur when the device data is abnormal. For example, an automatic reference waveform generation method for device data that suppresses the occurrence of defective products by determining which part of the device is abnormal, or predicts defect information such as the type of defect or the position of occurrence of a defect, and It is an object of the present invention to provide a defect information prediction method.

  The invention according to claim 1 of the present invention monitors the operation and manufacturing conditions of the manufacturing processing apparatus, obtains apparatus data, acquires waveform data from the apparatus data, and further, at the time of a non-defective product based on the inspection result from the inspection apparatus. Select whether to create a reference waveform or a reference waveform for a defective product, and the acquired waveform data matches the reference waveform for the original good product or the reference waveform for the original defective product within a specified range. In this case, the reference waveform is automatically updated by superimposing the waveform data, and when it does not match, it is registered as a new reference waveform.

  The invention according to claim 2 of the present invention monitors the operation and manufacturing conditions of the manufacturing processing apparatus, obtains apparatus data, acquires waveform data from the apparatus data, and the waveform data is the original non-defective product according to claim 1 A defect characterized in that defect information is predicted when the reference waveform does not match within a predetermined range and further matches with the reference waveform of the original defective product according to claim 1 within a predetermined range. Information prediction method.

  According to a third aspect of the present invention, there is provided the automatic reference waveform generation method for apparatus data according to the first aspect, wherein the defect content is linked to the reference waveform at the time of occurrence of the defect.

  According to the automatic reference waveform creation method and the defect information prediction method of the apparatus data according to the present invention, it is possible to cope with a change in the state of the process apparatus over time by automatically updating the reference waveform, and the amount of data The reference waveform with higher accuracy can be created as the value of the number is increased. As a result, it is possible to prevent a non-defective product from being erroneously inspected as a defective product and a defective product from being erroneously inspected as a non-defective product.

  In addition, when the waveform data does not match the reference waveform for the non-defective product within the predetermined range, it is further compared with the reference waveform for the defective product, and when the reference waveform for the defective product matches within the predetermined range, It is possible to predict defect information that may occur. As a result, it is possible to suppress the occurrence of defective products and to minimize the occurrence of so-called common defects due to the same cause.

  In addition, by associating defect contents such as defect type and position with the defect reference waveform, it is possible to perform maintenance of the process apparatus and manual operation, and it is expected to suppress defects occurring later.

The figure which showed an example of the color filter board | substrate in the cross section. The flowchart of the process of the photolithographic method generally used. The figure which shows an example of the information processing in the apparatus data collection device to which the automatic reference waveform preparation method of apparatus data and the prediction method of defect information which concern on this invention are applied. The figure of the waveform data which showed the case where the glass substrate is washed as an example. The figure which shows the reference | standard waveform creation flow concerning embodiment of this invention. The figure which shows the preparation method of the reference waveform at the time of the quality product concerning embodiment of this invention. The figure which shows the preparation method of the reference waveform at the time of the defect generation concerning embodiment of this invention. The figure which shows the comparison flow of the waveform data concerning embodiment of this invention.

  Hereinafter, embodiments of an apparatus data automatic reference waveform creation method and defect information prediction method according to the present invention will be described with reference to the drawings, taking a glass substrate cleaning process in the case of manufacturing a glass substrate of a color filter as an example. .

  FIG. 3 shows an example of information processing in an apparatus data collection apparatus to which an apparatus data automatic reference waveform creation method and defect information prediction method according to the present invention are applied.

  The apparatus data collection apparatus (hereinafter referred to as EES) 11 has a function of collecting apparatus information 13 from the target apparatus 12 at a high speed in a cycle of several ms and a communication function of receiving inspection result information 15 from the inspection apparatus 14.

  Device information (process operation data such as stage vibration and information including process data such as process conditions such as power consumption, pump pressure, and cleaning liquid flow rate) 13 including process operation data and process data collected by the EES 11 is: It is displayed in real time on a data collection personal computer (hereinafter referred to as a data collection PC) 16, and waveform data is acquired from the process operation data and the process data. The data collection PC 16 determines whether the acquired waveform data is normal, and determines whether the apparatus is operating normally. In this case, for example, FIG. 4 shows an example in which a glass substrate is cleaned. Data on the operation time of the cleaning liquid supply pump and the amount of cleaning liquid is collected, and the horizontal axis indicates ms ( Waveform data indicating the elapsed time indicated in milliseconds (milliseconds) and the liquid volume indicated in ml (milliliters) on the vertical axis are generated.

  The ID information 17 of the glass substrate that has been subjected to process processing (in this case, cleaning processing by a cleaning device) while waveform data is being collected by the EES 11 is linked to the waveform data, and the linked waveform data is the waveform. It is stored in the data file 19. The waveform data file 19 is a file that stores a plurality of waveform data collected from the target device 12 and associated with ID information of the glass substrate for each item such as heat, pressure, liquid amount, and power consumption. is there.

  The glass substrate ID information 17 can be obtained from a host system such as an LCS (Line Control System) 18. On the other hand, the ID information of the glass substrate is also associated with the inspection result information 15 output from the inspection device 14 (in this case, a foreign matter inspection device performed for inspecting the degree of cleaning of the glass substrate). A reference waveform selected from the reference waveforms stored in the reference waveform data file 20 based on the non-defective / defective product information in the inspection result information 15 and an ID associated with the inspection result from the waveform data file 19 The waveform data extracted from the information is compared.

  Waveform data acquired in the past is used as the reference waveform. As the waveform data used for the reference waveform, data of a glass substrate ID determined as a non-defective product from the inspection result of the inspection apparatus is used. At present, a reference waveform is registered using an average value obtained by superimposing one or more pieces of waveform data determined to be non-defective by inspection results. Therefore, an operator's operation is required when changing the reference waveform.

  If the comparison with the waveform data is always performed using the same reference waveform without changing the reference waveform, there is the following problem, and there is a possibility that an accurate comparison cannot be performed.

  One of the problems is a variation in data when the product is non-defective. That is, there is a possibility that the waveform data varies even when a good product is made.

  Another problem is a change in waveform data over time. That is, it is an influence on waveform data due to a change with time such as deterioration of the apparatus.

  To solve the above problem, create a reference waveform with higher accuracy. Create the reference waveform by superimposing the latest waveform data when as many non-defective products as possible are produced and calculating the average value. There is a need.

  In addition, the waveform data when a defective product is produced is created as a reference waveform at the time of occurrence of a defect and linked to information such as the type and position of the defect in the same manner as the reference waveform for a non-defective product. As a result, if the waveform data is judged as an abnormal value compared with the reference waveform at the time of non-defective product, the waveform data is compared with the reference waveform at the time of occurrence of a defect. It becomes possible to predict the type of defect.

  FIG. 5 shows a reference waveform creation flow.

  After the start (K-1), the inspection result is received from the inspection device (K-2). When the received inspection result is non-defective (YES in (K-3)), a new reference waveform is created by superimposing the current reference waveform on the substrate ID waveform data of the glass substrate (K-4), and the reference The waveform is updated to a new reference waveform (K-5), and the process ends (K-13).

  If the inspection result is not good (NO in (K-3)), the waveform data of the current board ID is compared with the reference waveform for the good product to determine whether the waveform data is normal (K-6). If normal (YES in (K-6)), the reference waveform is not updated and the process is terminated (K-13).

  If it is not normal (NO in (K-6)), then it is determined whether there is a reference waveform at the time of occurrence of a defect having the same defect information as the defect information of the inspection result (K-7). If yes (YES in (K-7)), then superimpose the reference waveforms at the time of failure (K-8), and further determine whether there is failure information such as maintenance method (K-9) If it exists (YES in (K-9)), defect information such as maintenance information is added to the reference waveform at the time of defect occurrence (K-10), and the process ends (K-14). If there is no defect information (NO in (K-9)), the process ends as it is (K-14).

  In (K-7), when there is no reference waveform at the time of failure occurrence (NO of (K-7)), a new reference waveform at the time of failure occurrence is created (K-11), and further failure Defect information is added to the reference waveform at the time of occurrence (K-12), and the process ends (K-14).

  Next, the method of creating a reference waveform will be described in detail separately for non-defective products and defective products. FIG. 6 is a diagram showing a method of creating a reference waveform for a non-defective product.

  Inspection result information obtained by inspecting (G-1) the glass substrate by an inspection device (in this case, a foreign matter inspection device as an example) is received by a data collection PC (not shown) (G-2). In this case, the inspection result information includes non-defective product information and substrate ID information of the glass substrate. From the substrate ID of the glass substrate, for example, waveform data indicating the amount of cleaning liquid in the cleaning process when the glass substrate is cleaned is extracted (G-3). Reference numeral 31 denotes the extracted waveform data. The extracted waveform data 31 is a waveform data file that stores waveform data indicating the power consumption of the cleaning device, the discharge pressure of the cleaning solution, the flow rate of the cleaning solution, the temperature of the cleaning solution, etc. in addition to the amount of the cleaning solution (see FIG. (Not shown).

  On the other hand, a non-defective reference waveform is selected from reference waveforms (G-4) stored in a reference waveform data file (not shown) (G-5). The reference waveform selected in this case indicates the amount of the cleaning liquid at a non-defective product in the cleaning process. Reference numeral 32 denotes the selected reference waveform.

  The extracted waveform data 31 and the selected reference waveform 32 are superposed, and the average is taken to create a new reference waveform 33. In this way, a reference waveform for a non-defective product is created. The created reference waveform 33 is updated to a new reference waveform and stored in a reference waveform data file.

  For example, in the case of a cleaning process, the extracted waveform data includes all the waveform data indicating the power consumption of the cleaning device, the discharge pressure of the cleaning solution, the flow rate of the cleaning solution, the temperature of the cleaning solution, etc. Data is extracted, a reference waveform corresponding to the non-defective product is selected, and the extracted waveform data and the selected reference waveform are overlapped and averaged to create a new reference waveform. As waveform data for extraction, it is desirable to acquire as much waveform data as possible for each process. Thereby, the accuracy of the new reference waveform is improved, and the defect information can be accurately predicted.

  Since EES collects process operation data and process data separately for each process equipment such as cleaning, coating, and drying, when creating a reference waveform, every process operation and all process data It is desirable to create a reference waveform every time and store it in the reference waveform database. Table 1 shows a part of the reference database. The waveform data acquired from the process device may be stored in the waveform database.

  When creating a reference waveform for a non-defective product, the method is not limited to a method of creating a new reference waveform by superimposing the waveform data and the reference waveform as described above, and taking the average, for example, waveform data from a glass substrate ID. Under the calculation conditions that have been acquired and specified, the waveform data collected within 10 days, for example, is automatically superimposed and averaged, and the data is updated as a reference waveform. The reference waveform may be created.

  Next, FIG. 7 shows a method of creating a reference waveform when a defect occurs. First, the glass substrate is inspected by the inspection device 50 (F-1), and inspection result information 51 is acquired (F-2). The inspection result information at the time of defect detection includes information such as the substrate ID, the inspection result (in this case, defect information), the type of defect, and the like. Next, the waveform data 53 of the same substrate ID as the substrate ID of the inspection result information is extracted from the waveform data file (not shown) from the plurality of acquired waveform data 52. The waveform data 53 extracts all the waveform data that is considered to be the basis of the inspection result from the waveform data file. For example, if a foreign object is defective as a result of inspection by a foreign substance inspection device, the waveform indicates the power consumption of the cleaning device, the discharge pressure of the cleaning liquid, the flow rate of the cleaning liquid, the liquid temperature of the cleaning liquid, etc. All waveform data of the data is extracted. If the inspection result is defective, there is a possibility that a defect has occurred due to an apparatus other than the process apparatus collecting the waveform data. Therefore, the extracted waveform data 53 is first compared with the reference waveform 54 for a non-defective product. (F-3).

  As described above, the comparison is to determine whether the waveform data is normal. The waveform data (the waveform data 53 in FIG. 7) and the reference waveform (the reference waveform 54 in the non-defective product in FIG. Data).

  As a result of the comparison, if the waveform data 53 matches the reference waveform for a non-defective product (when the waveform data is normal), it is considered that a defect is caused by another process apparatus. In this case, since the waveform data is considered to be caused by another process apparatus, it is not added to the waveform data when the apparatus is defective. Also, it is not added as a normal reference waveform.

  On the other hand, when they do not match (when the waveform data is abnormal), it is considered that the occurrence of the failure is caused by the process device collecting the waveform data.

  If the waveform data 53 does not match the reference waveform 54 for a non-defective product, that is, if the waveform data 53 is abnormal, how abnormal it is (compared to the reference waveform 54) (abnormal upper limit value, abnormal lower limit value, etc.) Get the cause of the abnormality. Further, whether or not a reference waveform corresponding to a defective product is registered (stored) from information such as the type and position of the defect from the inspection result is determined based on the cause of the abnormality and the type of the defect. Search for a reference waveform. The reference waveform at the time of occurrence of a defect here is a reference based on the type of defect, the position of the defect, how abnormal it was when compared to the reference waveform (abnormal upper limit value, abnormal lower limit value, etc.) It is a waveform. If there is a reference waveform that matches these in the searched reference waveforms, the reference waveforms that match the waveform data 53 are added to obtain a reference waveform when a new defect occurs.

  A reference waveform 55 at the time of occurrence of a defect stored in a reference waveform data file (not shown) is an example of a searched reference waveform at the time of occurrence of a defect. The reference waveform is stored for each condition or defect of the inspection result that has occurred, such as the reference waveform group 56 when 1 occurs, and the reference waveform group 57 when defect 2 occurs. In addition, for each reference waveform when a failure occurs, a reference waveform for the upper limit value abnormality and a reference waveform for the lower limit value abnormality are provided.

  Of the reference waveform group 56 when the defect 1 occurs, for example, the reference waveform 58 of the upper limit value abnormality or the reference waveform 59 of the lower limit value abnormality is a foreign object defect that has occurred in the past (in this case, a defect size defect). ) Is updated from the reference waveform, and in the case of a non-defective product indicating the power consumption of the cleaning device, the discharge pressure of the cleaning liquid, the flow rate of the cleaning liquid, the temperature of the cleaning liquid, etc. For example, the reference waveform is created from waveform data that exceeds an upper limit value and a lower limit value of ± 5%. In addition, among the reference waveform groups when the defect 2 occurs, the defect 1 is the case of the foreign matter size defect. However, the reference waveform group when the defect 2 occurs is, for example, the same foreign substance defect, Reference waveform at the time of occurrence of a defect relating to a height defect, which is a reference waveform updated from waveform data at the time of occurrence of a past defect, in which a reference waveform of an upper limit value abnormality is 60 and a reference waveform of a lower limit value abnormality is 61 .

  As a result of comparison, if there is no matching reference waveform at the time of occurrence of a defect, information on the content of the defect such as the type of defect is extracted from the inspection result, and the information is linked to the waveform data when a new defect occurs. Is registered as a reference waveform (F-4). Reference numeral 62 is the same as the waveform data 53, and shows a newly generated reference waveform when a defect occurs.

  Further, by registering information related to the maintenance work performed at that time in association with the reference waveform at the time of the occurrence of the defect, it is possible to collect a coping method when the defect occurs.

  On the other hand, as a result of the comparison, if there is a matching reference waveform at the time of occurrence of a failure, the reference waveform 63 at the time of failure occurrence (for example, waveform data 63 if the waveform matches the reference waveform 53 of the upper limit abnormality) and the waveform data 53 Are automatically created and updated when a defect occurs (F-5). Reference numeral 64 denotes an updated reference waveform when a defect occurs.

  The comparison of the waveform data can be performed by comparing the waveform data with the reference waveform as soon as the processing for each device is completed before performing the inspection. A comparison flow of waveform data is shown in FIG. After the start (S-1), the waveform data of the process device is acquired (S-2). The acquired waveform data is compared with a reference waveform for a non-defective product (S-3), and it is determined whether the waveform data is within the upper and lower limits of the reference waveform for a non-defective product (S-4). . If it is within the upper and lower limits of the reference waveform for a non-defective product (YES in (S-4)), it is next determined whether or not the waveform data exceeds a threshold value (S-5). .

  Monitor the operation and manufacturing conditions of the manufacturing processing device to obtain device data, acquire waveform data from the device data, compare the waveform data with a predetermined threshold upon completion of the manufacturing processing device, and the waveform data When the threshold value is exceeded, it is possible to predict a defect in the manufacturing apparatus.

  The reason for determining whether the waveform data does not exceed the threshold value (S-5) is that the process apparatus generally changes with time, and accordingly, the waveform data also changes slightly, resulting in the inspection result. There is a possibility that a defect will occur after a certain period of time without affecting the process. In addition, the reference waveform for non-defective products is overlapped with the waveform data that changes over time, so even with waveform data that changes over time, the inspection results from the inspection device may be non-defective. The waveform also changes with time. As a result, even when a certain amount of time has elapsed and a defect has occurred due to a change in the process equipment over time, there is a possibility that even if a reference waveform at the time of a non-defective product is compared with waveform data, it is not determined to be abnormal .

  Therefore, threshold values are set for the upper and lower limits of the initial waveform data, and when the threshold values are exceeded, it is determined that the change due to changes in waveform data over time is acceptable, that is, the threshold value is deviated. An abnormality in waveform data due to a change over time is notified on the PC, the waveform data is displayed, and an alarm is generated to prompt replacement or maintenance of the part. As a result, for example, when the power consumption of the motor is measured and the power consumption increases significantly, it can be determined that the motor has failed or is about to fail, and motor replacement and maintenance can be performed.

  If it is determined that the waveform data does not exceed the threshold (YES in (S-5)), it is determined that normal device operation has been performed and the process ends (S-12). If the waveform data exceeds the threshold (NO in (S-5)), it is determined that an abnormality has occurred in the operation of the apparatus, the waveform data in which the abnormality has occurred is displayed on the data collection PC, and an alarm is issued. Occurs (S-6), the motor is replaced or maintained, and then the process ends (S-12).

  On the other hand, if the waveform data exceeds the upper and lower limits of the reference waveform (NO in (S-4)), the waveform data is compared with the reference waveform at the time of occurrence of a failure, and matches the reference waveform at the time of occurrence of the failure. If it is (YES in (S-8)), the waveform data and defect information (defect type, position, maintenance information, etc.) are displayed on the data collection PC and an alarm is generated (S-9) Then, the process ends (S-12). As a result, it is possible to predict the content of the failure such as the type of failure that has occurred, and it is possible to take measures such as maintenance immediately before the inspection result is issued by the inspection device immediately after the process device. It is possible to minimize the defects caused by.

  If it does not coincide with the reference waveform at the time of occurrence (NO in (S-8)), it is determined whether or not the comparison with the reference waveform at the time of occurrence of all defects has been completed (S-10). When the comparison with the reference waveform is completed (YES in (S-10)), the waveform data is displayed on the data collection PC, an alarm is generated (S-11), and then the process is terminated (S-). 12).

  When the comparison with the reference waveform at the time of occurrence of all the defects has not been completed (NO in (S-10)), the waveform data is compared with another reference waveform at the time of occurrence of the defect (S-7), The waveform data is compared with the reference waveform at the time of occurrence of a defect until it matches the reference waveform at the time of occurrence (YES in (S-8)) or until the comparison with the reference waveform at the time of occurrence of failure is completed.

  As described above, according to the automatic reference waveform creation method and defect information prediction method of the apparatus data of the present invention, it is possible to cope with a change in the state of the process apparatus over time by updating the reference waveform. As the amount of data increases, it becomes possible to prevent erroneous inspection of a non-defective product as a defective product or erroneous inspection of a defective product as a non-defective product.

  In addition, when the waveform data does not match the reference waveform for the non-defective product within the predetermined range, it is further compared with the reference waveform for the defective product, and when the reference waveform for the defective product matches within the predetermined range, It is possible to predict defect information that may occur. As a result, the occurrence of defects can be suppressed, and the occurrence of so-called common defects due to the same cause can be minimized.

  In addition, by associating defect contents such as defect type and position with the defect reference waveform, it is possible to perform maintenance of the process apparatus and manual operation, and it is expected to suppress defects occurring later.

DESCRIPTION OF SYMBOLS 1 ... Color filter 2 ... Glass substrate 3 ... Black matrix 4a ... Red R colored pixel 4b ... Green G colored pixel 4c ... Blue B colored pixel 5 ... Transparent Electrode 6 ... Photo spacer 7 ... Vertical alignment 11 ... Equipment data collection device (EES)
12 ... Target device 13 ... Device information 14 ... Inspection device 15 ... Inspection result information 16 ... Data collection PC
17 ... ID information 18 ... LCS
19 ... Waveform data file 20 ... Reference waveform data file 31 ... Extracted waveform data 32 ... Selected reference waveform 33 ... New reference waveform 33
50 ... Inspection device 51 ... Inspection result information 52 ... Acquired waveform data 53 ... Waveform data 54 ... Reference waveform 55 for non-defective product ... Reference waveform 56 for failure occurrence ... Reference waveform group 57 when failure 1 occurs ... Reference waveform group 58 when failure 2 occurs ... Reference waveform 59 when the upper limit value is abnormal ... Reference waveform 60 when the lower limit value is abnormal Reference waveform 61 of upper limit value abnormality ... Reference waveform 62 of lower limit value abnormality ... Newly created reference waveform 63 when failure occurs ... Reference waveform 64 when failure occurs ... Updated failure Reference waveform at the time of occurrence

Claims (3)

  Monitors the operation and manufacturing conditions of the manufacturing processing equipment, obtains equipment data, obtains waveform data from the equipment data, and further determines whether the reference waveform for a non-defective product or the reference waveform for a defective product depends on the inspection result from the inspection equipment. If the acquired waveform data matches the reference waveform of the original non-defective product or the reference waveform of the original defective product within a specified range, the reference waveform is superimposed by overlapping the waveform data. An automatic reference waveform generation method for apparatus data, wherein the apparatus data is automatically updated and registered as a new reference waveform if they do not match.   The operation data and the manufacturing conditions of the manufacturing processing apparatus are monitored to obtain apparatus data, waveform data is acquired from the apparatus data, and the waveform data coincides with the reference waveform of the original good product according to claim 1 within a predetermined range. In addition, the failure information prediction method is characterized in that the failure information is predicted when compared with the reference waveform of the original defective product according to claim 1 and matched within a predetermined range.   2. A method for automatically generating a reference waveform of apparatus data according to claim 1, wherein a defect content is associated with a reference waveform when a defect occurs.

Images