JP2006294831A - Data collecting system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a data collecting system with which data processing of a semiconductor manufacturing device can appropriately be performed by shortening retrieval time of data stored in a data base. <P>SOLUTION: When the semiconductor manufacturing device 1 performs processing of a semiconductor substrate; a data collecting device 2 is operated, raw data showing operation information is acquired from the semiconductor manufacturing device 1, and raw data are stored in a primary data base 3. An event is added to raw data stored in the primary data base 3 by an operation of the data collecting device 2, and work data are stored in a secondary data base 4. Since the number of data records in work data stored in the secondary data base 4 becomes small, time can be shortened when the data collecting device 2 retrieves desired data. A method for generating work data is realized by cutting raw data by a time interval and the event, which are previously designated. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a data collection system that performs processing such as data collection / accumulation, and more particularly to a data collection system for collecting / accumulating data indicating the operating state of a semiconductor manufacturing apparatus and processing and utilizing the data. It is about.

  Conventionally, information indicating the operating state of a semiconductor manufacturing apparatus is acquired and stored in a database, and the stored information (for example, data) is displayed in a graph, superimposed graph, average, maximum A data collection system is known that performs mathematical processing such as minimum, compares the acquired data with past data or ideal data, and analyzes whether the semiconductor manufacturing apparatus is operating correctly. An example of such a data collection system is called an EES (Equipment Engineering System).

  The information indicating the operating state of the semiconductor manufacturing equipment includes information on sensors, actuators, etc. such as temperature, gas flow rate, valve open / closed state, temperature, gas flow rate, valve open / close status indicated by the semiconductor manufacturing equipment to the actuators, etc. Information such as instructions, ID information such as processed wafers and carriers, recipes, and step transition information when executing recipes. These information are combinations of events that occurred and their times. It is represented by In the case of continuous data such as temperature, the data is periodically read out at intervals of several seconds to 1 second or earlier, and the data is stored in the database as it is, or from the data acquired last time. Store new data in the database when there is a change.

  However, the data stored in the database generally stores one data in one data record, but the number of data stored in the database is very large depending on the type of data and the frequency of occurrence. For example, in the case of a semiconductor manufacturing apparatus, the total number of data when operated for several months is on the order of several hundred thousand to several million. In this way, the basic key for searching data records from a database that stores a large number of data is time and data items. However, if the time period specified in the search is long, the matching data record is inevitably required. The number increases, and as a result, the data retrieval takes a considerable time.

  FIG. 6 is a schematic configuration diagram of a conventional data collection system. That is, as shown in FIG. 6, data such as the furnace temperature and the gas flow rate indicating the operating state of the semiconductor manufacturing apparatus 11 is combined with the measured time and stored in the database 13 as it is. Then, the data collection device 12 retrieves necessary data from the database 13 and controls the semiconductor manufacturing device 11. At this time, if the data collection device 12 tries to search using the data stored in the database 13 as it is, the data search time may be long due to the large amount of data, resulting in a data collection system. 12 is not easy to use.

  The present invention has been made in view of the above problems, and provides a data collection system that can appropriately perform data processing of a semiconductor manufacturing apparatus by shortening the search time of data stored in a database. The purpose is to do.

  In order to solve the above-described problems, a data collection system according to the present invention is a data collection system in which a semiconductor manufacturing apparatus and a data collection apparatus that collects raw data indicating operation information from the semiconductor manufacturing apparatus are connected by a communication line. The data collection device stores a primary database that stores raw data collected from a semiconductor manufacturing apparatus for each record, and processing data in which records are assigned to the raw data stored in the primary database for each predetermined condition. And a secondary database to be stored. The processing data is retrieved from the secondary database and predetermined control is performed on the semiconductor manufacturing apparatus.

  According to such a configuration, when the primary database stores the raw data indicating the operation information of the semiconductor manufacturing apparatus, the secondary database for search is performed on the raw data for each predetermined condition (for example, time and event). Stores machining data to which records are assigned. As a result, the processed data stored in the secondary database for search has a reduced number of records even if the data amount of one record increases (that is, the record size increases). In addition, the data search time can be shortened. Note that the total data amount of the primary database and the total data amount of the secondary database can be equal.

  The present invention can also provide a data collection method in a semiconductor manufacturing apparatus. That is, in the data collection method according to the present invention, the data collection apparatus collects the raw data indicating the operation information from the semiconductor manufacturing apparatus via the communication line, and stores the collected raw data in the primary database for each record. A procedure, a procedure for storing processed data in which records are assigned to the raw data stored in the primary database for each predetermined condition in the secondary database, and a data collecting device retrieves the processed data from the secondary database and performs semiconductor processing. And a procedure for performing predetermined control on the manufacturing apparatus.

  According to the data collection system of the present invention, machining data in which records are assigned to the operation information collected from the semiconductor manufacturing apparatus for each predetermined condition (for example, time and event) is stored in the secondary database. Therefore, by searching the secondary database, the number of records matched by the search can be reduced, so that the data search can be performed efficiently and in a short time. Therefore, it is possible to increase the search speed of data (record).

<< Summary of Invention >>
The data collection system of the present invention is a database (hereinafter referred to as a primary database) that stores raw data indicating operation information acquired from a semiconductor manufacturing apparatus by a data collection apparatus in order to increase the search speed of data accumulated in the database. In addition to the above, a search database (hereinafter referred to as a secondary database) is constructed. That is, the secondary database stores processed data obtained by processing the raw data of the primary database, and reduces the time for retrieving data from the secondary database by reducing the number of data records retrieved from the secondary database. I am trying.

  That is, when a data processing system for a semiconductor manufacturing apparatus is constructed by the data collection system of the present invention, when raw data indicating the operating state of the semiconductor manufacturing apparatus itself acquired from the semiconductor manufacturing apparatus is accumulated and processed. First, the raw data accumulated in the primary database is processed and stored in the secondary database as processed data. Then, by retrieving the machining data from the secondary database, the retrieval time is shortened and the utilization in the application is facilitated.

<< Embodiment of the Invention >>
Hereinafter, embodiments of a data collection system according to the present invention will be described in detail with reference to the drawings. FIG. 1 is a schematic configuration diagram of a data collection system according to an embodiment of the present invention.

  The data collection system according to the present embodiment shown in FIG. 1 includes a semiconductor manufacturing apparatus 1 that processes a semiconductor substrate, a data collection apparatus 2 that collects raw data indicating operation information from the semiconductor manufacturing apparatus 1, and a data collection apparatus 2 A primary database 3 that stores raw data collected from the semiconductor manufacturing apparatus 1 for each record, and a secondary database 4 that stores machining data in which records are assigned to the raw data stored in the primary database 3 for each predetermined condition. It is the composition provided with. The data collection device is configured to search the processing data from the secondary database and perform predetermined control on the semiconductor manufacturing device.

  The primary database 3 is a database that originally exists in the data collection system, such as the database 13 in the conventional data collection system shown in FIG. A database conversion program for converting raw data stored in the primary database 3 into machining data is prepared, and the database that stores the machining data converted by the database conversion program is the secondary database 4. Such a secondary database 4 is a database that is actually used for searching, and is designed so that data records can be searched faster than the primary database 3.

  Further, the predetermined condition for processing the raw data stored in the primary database 3 includes a time during a semiconductor substrate processing process in the semiconductor manufacturing apparatus 1 and events in various processing steps. For example, the predetermined conditions include a step of starting a recipe, raising / lowering the temperature in the furnace, or controlling the temperature inside the furnace, or supplying / stopping the gas supplied to the furnace.

  Next, the operation of the data collection system shown in FIG. 1 will be described. In the configuration of the data collection system shown in FIG. 1, when the semiconductor manufacturing apparatus 1 is processing a semiconductor substrate, the data collection apparatus 2 is operated to obtain raw data indicating operation information from the semiconductor manufacturing apparatus 1, Raw data is stored in the primary database 3. Further, by operating the data collection device 2, processing data is generated by adding an event or the like to one or more raw data stored in the primary database 3, and the processing data is stored in the secondary database 4. As a result, since the number of data records of the processed data stored in the secondary database 4 is reduced, the time for the data collection device 2 to search for desired data can be shortened. The method of processing the raw data of the primary database 3 to generate the processed data can be realized by cutting out the raw data at a predetermined time interval or a predetermined event.

  FIG. 2 is a diagram showing an example of a data format of raw data stored in the primary database 3 in the data collection system shown in FIG. That is, the primary database 3 has a data format structure of raw data as shown in FIG. 2 as an example. In FIG. 2, one line such as (occurrence time) 19:05:01, (item name) temperature 1, (value) 700, etc. represents one record. The data format of the primary database 3 shown in FIG. 2 has one record per second per item. Accordingly, the data collection device 2 retrieves 3,600 records per hour and 86,400 records per day.

  FIG. 3 is a diagram showing an example of the data format of the machining data stored in the secondary database 4 in the data collection system shown in FIG. In FIG. 3, (start time) 19:05:01, (end time) 19:15:00, (item name) temperature 1, (time) 19:05:01, (value) 700, (time) 19: One line such as 05:02, (value) 701... Represents one record.

  For example, suppose that the data format of the machining data stored in the secondary database 4 is structured as shown in FIG. Further, assume that the data conversion cycle of the secondary database 4 is set to 10 minutes. Thus, the data format of the secondary database 4 shown in FIG. 3 has one record for 600 seconds per item (for example, 10 minutes from 19:05:01 to 19:15:00). That is, 6 records are searched in one hour and 144 records are searched in one day. Instead, one record has 600 data.

  With the configuration of the data format of the secondary database 4 processed as shown in FIG. 3, the number of records that match in the search can be reduced. That is, in the data format of FIG. 2, 86,400 records must be searched in one day, but in the data format processed as in FIG. 3, 144 records can be searched in one day. The number of records can be considerably reduced. Therefore, the search time can be shortened by searching the data stored in the secondary database 4.

  In the data format processed as shown in FIG. 3, the number of data per record is 600 times larger than the data format shown in FIG. 2, so the processing time per unit record increases, but the total processing Since the processing effect due to the reduction in the number of records is considerably larger, the search speed can be increased by using the data format processed as shown in FIG.

  In the example of the data format as shown in FIG. 3, the data conversion process is performed with the conversion period of the data format fixed to 10 minutes. However, the data conversion process is performed using an event indicating the content of processing the semiconductor substrate. It is also possible to make the processing asynchronous. For example, data conversion processing can be performed on the secondary database 4 at every event of starting a recipe on a semiconductor substrate or an event of changing a target temperature during processing of a semiconductor substrate.

  FIG. 4 is a temperature characteristic diagram showing an example of a temperature control pattern in the semiconductor manufacturing apparatus 1 shown in FIG. In the example of FIG. 4, step start of each process such as recipe start and temperature rise start is used as an event. Further, the processing from the start of the step to the start of the next step is processed as being in step processing. Further, the completion of the last step is an event of the recipe end.

  FIG. 5 is a diagram showing an example of the data format of the processing data stored in the secondary database 4 for each event in the temperature control pattern shown in FIG. That is, FIG. 5 shows the processing data for each step where the content of the event changes. For example, from the start time of Step 1 at which the recipe is started to the start time of Step 2 at which the temperature rise is started, the event is set as Step 1 (Start Time) 2:03:00, (End Time) 2: One record is represented in one line as 08:00, (item name) temperature 1, (time) 2:03:00, (value) 600, (time) 2:03:01, (value) 601. Yes.

  Further, from the start time of Step 2 at which the temperature rise is started to the start time of Step 3 at which the temperature rise is ended, the event is set as Step 2 (Start time) 2:08: 00, (End time) 2: 1 record is expressed in the next line as 11:00, (item name) temperature 1, (time) 2:08:00, (value) 600, (time) 2:08:01, (value) 601. Has been.

  Furthermore, the event is set as Step 3 from the start time of Step 3 where the temperature rise is finished and the control of the constant temperature is started to the start time of Step 4 where the control of the constant temperature is finished and the temperature drop is started. (Start time) 2:11:00, (End time) 2:41:00, (Item name) Temperature 1, (Time) 2:11:00, (Value) 803, (Time) 2:11:01, Further, one record is represented in the next line as (value) 802.

  By using such a data format for each event, for example, when it is desired to retrieve the data of step 3 in which the processing of the semiconductor substrate is performed while controlling the inside of the processing furnace at a constant temperature, another record (that is, Only by acquiring one record of step 3 without combining with the record of step 1 or the record of step 2, data necessary for the control at that time can be extracted and optimal control can be performed. In this way, it is possible to construct the secondary database 4 in units of data that the application wants to recognize.

  Regarding the secondary database, not a single data format such as temperature, but a plurality of data formats such as temperature, gas flow rate, valve open / close state, etc. are assumed depending on the application. It is also possible to have a plurality of secondary databases corresponding to such a plurality of data requests. Furthermore, by having a data format with multiple data such as temperature, gas flow rate, valve open / close state, etc. for multiple secondary databases, it is possible to handle various processes such as temperature control and gas flow rate control. it can.

  Also, the secondary database does not have to be persistent. In other words, it takes time to convert the data format from the primary database to the secondary database, but the secondary database can be easily reconstructed from the primary database, so the secondary database can be deleted accordingly. It is.

  That is, in the above embodiment, only the temperature of the processing furnace in the semiconductor manufacturing apparatus is taken up as an item handled by the data collection system. However, the present invention is not limited to this, and the flow rate of the gas supplied to the processing furnace or the exhaust gas processing valve is not limited. You can also handle the open / closed state as an item. Also, as shown in FIG. 4 and FIG. 5, by recording data by dividing each event, only necessary data for each event can be searched in a short time and appropriate processing can be executed.

In the present embodiment, as a predetermined condition, a database can be constructed for each item (temperature, gas flow rate, valve opening / closing, etc.) or for each item name (temperature 1, temperature 2, etc.). In this case, a plurality of secondary databases are constructed, and the total data amount of these secondary databases and the total data amount of the primary database are the same.
In the above embodiment, the semiconductor manufacturing apparatus for processing the semiconductor substrate is applied as the apparatus connected to the data collecting apparatus. However, the present invention is not limited to this, and the glass substrate used in the LCD (Liquid Crystal Display) apparatus. It is applicable also to the manufacturing apparatus which processes. The semiconductor manufacturing apparatus can be applied not only to a vertical apparatus but also to a horizontal apparatus. Further, any processing of the semiconductor substrate in the processing furnace may be performed. For example, the data collection of the present invention can also be applied when performing CVD (Chemical Vapor Deposition) processing, oxidation processing, diffusion processing, annealing processing, or the like. Needless to say, a plurality of devices may be connected to the data collection device.

It is a schematic block diagram of the data collection system in one embodiment of this invention. It is a figure which shows an example of the data format of the raw data stored in the primary database in the data collection system shown in FIG. It is a figure which shows an example of the data format of the process data stored in the secondary database in the data collection system shown in FIG. FIG. 2 is a temperature characteristic diagram showing an example of a temperature control pattern in the semiconductor manufacturing apparatus shown in FIG. It is a figure which shows an example of the data format of the process data stored in the secondary database 4 for every event in the temperature control pattern shown in FIG. It is a schematic block diagram of the conventional data collection system.

Explanation of symbols

1 Semiconductor manufacturing equipment 2 Data collection equipment 3 Primary database 4 Secondary database

Claims (1)

A data collection system in which a semiconductor manufacturing apparatus and a data collection apparatus that collects raw data indicating operation information from the semiconductor manufacturing apparatus are connected by a communication line,
The data collection device includes:
A primary database for storing raw data collected from the semiconductor manufacturing apparatus for each record;
A secondary database that stores processing data in which the records are assigned for each predetermined condition with respect to the raw data stored in the primary database;
A data collection system, wherein processing data is retrieved from the secondary database and predetermined control is performed on the semiconductor manufacturing apparatus.

Images