JPH06176031A - System and method for production control - Google Patents

Abstract

PURPOSE:To shorten a process time, to reduce a memory in use and to accurately control a line and predict the progress of a priority and an important lot by putting a short-period process schedule together with the process schedule of a lot whose progress needs to be controlled and generating the process schedule of the whole line. CONSTITUTION:All the processes of respective lot of progress controlled lots (lot group A) consisting of lots which are high in priority and importance are scheduled to generate the process schedule of the lots (1), and processes are carried out to control the progress and control the completion date. Then the lot process schedule is classified by manufacturing devices as line resources to generate the process schedule table of the devices (2). A processing and control method for progress-uncontrolled lots (lot group B) consisting of lots which are low in priority and importance is different from the lot group A. For the purpose, a short-period schedule is generated as the process schedule of the lots (3). Thus, the process schedules which are individually generated are classified by the devices and process days and put together (4) to generate the process schedule of the whole line.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a production control system and method for controlling the production of semiconductor parts such as LSI, clothing, and various industrial products such as automobiles. The management of manufacturing here means that the products of multiple types are
This includes scheduling that determines the manufacturing sequence according to each processing step and urgency, monitoring of the progress of manufacturing, and acquisition of processing data during manufacturing.

[0002]

2. Description of the Related Art A conventional production management system will be described below by taking an LSI manufacturing process as an example. FIG. 6 is a diagram for explaining the outline of the manufacturing procedure of the LSI. Depositing an insulating film such as an oxide film / nitride film on a single crystal wafer, diffusion by heat treatment, pattern formation by photolithography, film deposition by vapor deposition or sputtering, etching, introduction of impurities by ion implantation, scribing (splitting) ), Mounting, bonding and other techniques are used at least once, and many are used dozens of times, while the function is built on the semiconductor substrate,
The LSI is manufactured. In addition, each technique is further subdivided into several processes. For example, in the photolithography technique, as shown on the left side of FIG. 6, (a) application of a resist which is a photosensitizer, (b) mask alignment / exposure, (c) pattern formation by development, and (d) processing by etching. , Etc., and such a processing unit is usually called a “process”. An LSI is manufactured by repeating these steps several hundred times. A list of the steps of this manufacturing process
It is a “process chart” as shown in FIG. 7, in which used devices, processing conditions (recipe), etc. are described in the order of processing.

This process chart is designated for each lot (processing unit: usually cassette unit, single-sheet processing unit),
The LSI can be manufactured by sequentially performing the processing according to the process chart. What is characteristic here is that the same apparatus is used many times by changing the processing conditions in this step. Also, when manufacturing different types of products,
Even in the same process, manufacturing is performed under different processing conditions. The processing conditions (recipe) mentioned here include not only the manufacturing temperature of the heat treatment, the gas flow rate, the manufacturing conditions such as the mask used during exposure, but also the inspection conditions such as the position of dimension measurement.

In such manufacturing, LS has been recently used.
Not only I, but the demand for high-mix low-volume production is increasing, so many production lines can be produced on a single production line. Further, these products are not only different in the amount of production, but also different in importance and processing priority for each kind. Here, the degree of importance depends on the added value such as the design cost of a mask or the like, the manufacturing cost of requiring a special process, and the like, and is usually determined when a lot is input. On the other hand, the priority indicates which is processed first when the same device waits for processing, and is determined by a margin until the delivery date. For this reason, it is necessary to raise the priority of a lot that was low in priority because it had a sufficient margin until the delivery date at the beginning and was delayed from the schedule in the middle. On the contrary, there is a case where there is a margin on the way and the priority is lowered.

Further, acquiring and monitoring the process execution data is useful not only for preventing product defects, but also for analyzing the cause at the time of failure. Therefore, it is essential to acquire the processing data including the inspection data. Here, it is desirable that this processed data be automatically collected from the device,
It does not necessarily have to be automatic. When collecting this equipment data and inspection data, the same equipment is used many times in the process chart, so that data is the data of which process (of which lot in the process chart) of which lot. The key (word) information indicating whether there is any is added and stored in the database. As this key information, it is convenient to use a processing schedule created by scheduling. That is,
This is a method of recognizing the lot by a bar code reader during lot processing, receiving the processing schedule through network communication, and storing it in a database together with the processing data. Alternatively, there is a method in which a magnetic card or the like is transported together with the lot, the card is read, the key information is read, and the processed data is stored in the database. All of these have the feature that they can also serve as lot progress management. Such a line management procedure is shown in FIG. The schedule result is distributed to the computers arranged in the line in the form of a schedule, and the worker executes the process based on the work instruction sheet according to the schedule. Execution records such as processing data are sent directly from the manufacturing / inspection equipment to collect online data, or the operator notifies the progress of offline from the terminal to feed back to the scheduler to manage the progress, and to make progress at the next schedule. Reflect the situation.

In a mass production line where a large number of products of the same type are produced, a pile-up system in which old ones are sequentially processed (first-in / first-out) from accumulated lots that are stored in each device in a waiting state. In addition, a signboard system or the like that visually displays the excess or deficiency of the arrival lot is often used as a production method. The processed results are
The (keyword) information of which process of which lot is added is added on the spot by key input or reading with a barcode, and the data is stored in the database. However, in this case, it is difficult to control the delivery date management and TAT (the number of days until the lot is completed) because the processing schedule of each process of each lot cannot be grasped.

On the other hand, in a production line called ASICLSI, as shown in FIG. 9, there are lots of lots of different types having different processing priorities and importance. For example, a custom-made product or a product with a special specification is less important in mass production such as a memory or a general-purpose product because the production amount is small and the designing / manufacturing cost is high. Among ASIC LSIs, LSIs using full custom or standard cells require the use of a dedicated mask pattern from the substrate process.
Its importance is high. However, since a common substrate is used for the gate array LSI, the importance is low, and the importance is high after the wiring process and the process of using the dedicated mask pattern. On the other hand, the priority may be changed from the middle by monitoring the margin until the delivery date, and changes as the lot processing progresses. Therefore, the priority changes as the lot processing progresses. In order to control the completion date of lots of priority and importance (hereinafter referred to as progress control lots) to keep up with the delivery date, and to create key information for data acquisition, It is necessary to monitor the progress and perform regular scheduling about once a day and irregular scheduling due to equipment troubles, lot progress delays, etc. by the scheduler to successively update the processing schedule. In other words, the production management system monitors the progress of lots, changes the priority of which lot to process first based on the processing result, and schedules after the processing has already been performed. It will be executed in consideration of information such as failure), maintenance status, and operator operating status. As a method of creating the processing schedule, for example, the following processing may be executed in order from the lot with the highest processing priority. The device and the processing time of the next process are recognized by referring to the process chart, the time zone in which the device can be used is checked from the operation schedule, and the process is assigned to the idle time period. Here, since the number of processing devices is not limited to one, it is assigned to an empty device of the same type. This process is sequentially performed for other lots and the next process. As a result, the operation schedule of the apparatus includes the processing schedule, and the operable time is limited. The processing schedule of a lot with a lower priority is assigned to the free time zone. Similarly, it is possible to create an operator's schedule while looking at the vacancy at the same time. In this case, there are a plurality of operators who can use the device (may be duplicated with operators of other devices). Thus, there are several combinations of devices and operators that can be processed.

Therefore, at the time of creating a processing schedule, it is checked which of the combinations that can be actually processed (scheduled) from such combinations that can be processed, and from the lots of high priority, The processing schedules for the entire line are created by sequentially allocating the processing schedules for the steps. In this way, the processing schedule of the device and the processing schedule of the lot (and the processing schedule of the operator) are created.

[0009]

At this time, the amount of calculation increases as the number of lots and the number of devices (the number of operators when the operator is taken into consideration) increases, and especially when the number of lots processed is large. It requires memory capacity and calculation time. In a normal line, lots of mass-produced products, which have neither high importance nor priority, flow in addition to lots of high importance. Although the ratio varies depending on the manufacturing line, for example, a mass production lot that is several times to several tens of times the number of important lots often flows. In order to create a production plan on such a line, it is practically impossible to schedule all the processes of all lots over a long period of time due to restrictions such as computer memory and calculation time. is there.

An object of the present invention is to perform the above scheduling (in a production line in which lots of high priority / importance whose progress must be controlled are mixed in mass-produced products of which priority / importance is not high). It is intended to shorten the processing time at the time of (creating key information) and to reduce the memory used, and thereby to be able to accurately perform line management and progress prediction of priority / important lots. In particular, it is possible to adapt to a line in which the number of manufacturing lots and the number of types of the entire line are large.

[0011]

In order to achieve such an object, the invention of claim 1 is based on a process table showing manufacturing conditions, manufacturing order, etc. In the production management system that has a scheduler that creates a processing schedule for each manufacturing apparatus, and adds the information in the processing schedule to the processing data from the manufacturing apparatus and stores it in the database, the scheduler is a production plan,
A means for creating a long-term processing schedule of a lot whose progress needs to be controlled in consideration of a work schedule and a line resource operation schedule, and a means for separately creating a short-term processing schedule of a lot other than the above-mentioned lot. It is characterized by further comprising means for synthesizing the processing schedule of the short term with the processing schedule of the lot whose progress needs to be controlled to create a processing schedule of the entire line.

According to a second aspect of the invention, a predetermined schedule for each manufacturing lot and each manufacturing apparatus is created by a scheduler from a process table showing manufacturing conditions, manufacturing order, etc., and an operation schedule of each manufacturing apparatus. In the production management method of adding processing schedule information to the processing data from the manufacturing apparatus and storing it in the database, the scheduler needs to control the progress in consideration of the production schedule, process chart, and operation schedule of line resources. The step of creating a long-term processing schedule of a lot, the step of separately creating a short-term processing schedule of lots other than the lot, and the short-term processing schedule of the lot whose progress needs to be controlled. And a step of creating a processing schedule for the entire line by combining the processing schedule with the processing schedule.

According to the invention of claim 3, in addition to the invention of claim 2, the scheduler monitors the progress of the lot, and when an abnormality occurs in the progress, the long-term processing schedule and / or the process It is characterized by executing a step of re-creating the processing schedule of the entire line by correcting the processing schedule for a short period of time and combining it with an uncorrected processing schedule.

According to a fourth aspect of the present invention, in addition to the third aspect of the invention, even if the lot has a low priority, if an abnormality occurs that delays its progress, the lot is apparently regarded as a high priority lot. The step of re-creating the processing schedule of the entire line is executed.

[0015]

In the inventions of claims 1 and 2, a long-term processing schedule is created for an important lot whose progress is controlled. For other less important lots, make a short-term treatment schedule. Create an overall processing schedule by synthesizing long-term and short-term processing schedules. Therefore, as compared with the case where long-term processing schedules are created for all lots, the capacity of the memory for storing schedule information can be reduced. In addition, the processing schedule for the completed line reflects the long-term plan for important lots. In addition, data is stored for all lots and is subject to scheduler monitoring.

According to the third aspect of the invention, since the scheduler can be partially modified, it is possible to flexibly deal with an abnormal situation.

According to the fourth aspect of the present invention, even if the lot is of low importance, and the delivery date is short, the schedule (progress) can be managed as a lot of high importance.

[0018]

Embodiments of the present invention will now be described in detail with reference to the drawings.

A large number of lots having different priorities and importance flow on the production line. In order to manage the progress of these lots and control the completion date, the progress of the lots is managed by the scheduler, the completion date is predicted, and the delayed ones are scheduled again by processing such as increasing the priority. It is important to rebuild and operate the line according to this schedule. At this time, the scheduling is usually in the range of several months because it is necessary to perform the period until the lot is completed in order to predict the completion date. This processing schedule is sequentially updated according to the progress status of lots on the line, troubles in the equipment, changes in the production plan, and the like. In order to create such a processing schedule, in the conventional method, it is necessary to make a processing schedule for all the processes of all lots on the line at a time as shown by the range of the dotted line in FIG. It was virtually impossible to create a long-term schedule due to the problem of calculation time. In the present invention, only lots with high priority / importance that control progress are first extended from the entire lot. A processing schedule is set for a period, and then a short-term processing schedule for a large number of lots for which it is not necessary to accurately control the completion date of other gate array substrates, etc., and both are combined by a scheduler later. Therefore, it is possible to reduce the memory used when scheduling at once and the calculation time, and it is possible to schedule for a long period of time, so look at the long-term forecast once scheduled and raise the priority of lots that are likely to be delayed By processing and scheduling again,
Accurate progress management and delivery date management are possible. This processing schedule includes, for each processing step, information such as what step in which lot, what processing conditions are (recipe name), and scheduled processing time. A lot processing schedule is created in the same manner, and a schedule table in which information on which device, when, and under what conditions is to be processed is arranged in the order of steps is created. All or part of this processing schedule information is used as key information for storing the processing data in the database. For example, lot name, process table name, device name,
It is a recipe name or the like. It goes without saying that these may be numbers instead of names. These production control methods may be the same as the conventional method shown in FIG.

FIG. 1 is a flow chart showing an outline of the schedule processing according to the present invention. The computer used as the scheduler executes the scheduling related to the present invention by executing the software program in which the control procedure of FIG. 1 is described. Since the hardware configuration of the scheduler can be the same as the conventional one,
Detailed description will be omitted, and only the schedule processing according to the present invention will be described. Further, the processing steps shown in the figure correspond to the respective steps of the present invention.

For a progress control lot (lot group A) consisting of lots of high priority / importance, according to the procedure on the left side of the figure, all the processes are scheduled for each lot, and a lot processing schedule is created. By executing the processing based on this, the progress is accurately managed and the completion date is controlled. This scheduling is the period until the lot is completed, typically in the range of months. This predetermined schedule is
It is monitored by the scheduler and is sequentially updated in accordance with the progress status of the lot on the line, the production plan change caused by the abnormal occurrence of the manufacturing apparatus, and the like. After creating the processing schedule of, the processing schedule of each lot is categorized by the manufacturing equipment as line resources (also for operators if necessary),
Create a processing schedule table for the device (operator). On the other hand, the processing / management method for the non-progress control lot (lot group B) consisting of lots of low priority / importance is different from the lot group A. Like the gate array substrate process, these lots are lots that are usually produced in large numbers on the line, and it is not necessary to control the completion date of each lot accurately. It suffices to know the lot to be processed in about 1 week. Therefore, the processing schedule for these lots should be short-term. This processing schedule only needs to know the processing date and the priority, and it is not necessary to schedule exactly until the time. Therefore, the method of preparing the processing schedule is different from that of the progress control lot and can be simply obtained. The method will be described later. In this way, the processing schedules that have been created separately are classified according to device and processing date, and are combined to create a processing schedule for the entire line. According to the present invention, the progress control lot is scheduled based on a long-term plan even if it is planned to be processed for a short period, so that the completion date of each target lot can meet the delivery date. As described above, the processing order is controlled. At the same time, other mass production lots are scheduled to be processed at the same time. The creation of such a scheduling can be performed while reducing the memory and the calculation time.

The scheduler first creates a processing schedule (important processing schedule) for the progress control lot, but this portion is often a fraction of the total operating time of the line because the lot is selected. There is little waiting on the device. Therefore, it is easy to create a schedule centered on the processing schedule for each lot. The outline of the creation process is shown in FIG. For simplicity, the number of processes that can be processed in one day is reduced. In the case of 24-hour operation, the operating hours may be ignored. First, the progress control lots i, j, and
The current process starting steps (a, b, c, d,
...) is recognized. Next, referring to the process chart of each lot, the following processes are arranged in order in the operation time zone of the device (a).
Make a tentative schedule like. Since this tentative schedule does not take into consideration the overlap with other lots, it cannot be processed because it actually overlaps with the processing of other lots like the shaded parts (devices M2 and M3) in the figure. There is. If there is an overlap, the processing schedule of the low priority lot is moved backward as shown in (b). In the case of the same priority, it is sufficient to determine rules such as giving priority to the person who put in first and giving priority to the person who has advanced the process. However, if the steps are shifted as shown in (b), duplication (apparatus M2) occurs in another time zone. This also shifts one processing schedule in the same manner. If the processing time of the process is shifted, the influence should be minimized. If the overlap is small, the process that comes in later is shifted backward. In addition, if the processing priority of the lots that have been shifted once is increased by proceeding, the specific lots will not be displaced. It should be noted that there is a gap between steps in the figure, so it is the transportation time between steps. In this way, the processing schedule (c) may be created by arranging the processing schedules of all the processes of each lot (or arranging the processing schedules during the designated period). In the case of a line which is not in full operation for 24 hours, if the processing time is shifted and deviates from the operating time zone of the day, the processing is continued on the next day. When the scheduler makes a schedule for processing lots, the lots are classified for each device and a process schedule (d) for the device is created. This schedule is only for the progress control lot, and there is a vacancy.

Although illustrated in the form of a line table here, these processing schedules may be created as a list in the form of a processing schedule table as shown in FIG. 4 for each device.

On the other hand, a non-progress control lot with low priority and low importance can be scheduled for processing in the same manner as the progress control lot. The following method can be used as the processing schedule creation method.

(1) A process schedule for a progress-controlled lot is set, its operating time is registered by considering it as a non-operating time of the device, and the remaining time is scheduled as an available time to schedule the process.

(2) The facilities and processing time zones of the production line are divided into a lot for progress control and a lot for non-progress control lot in advance. Schedule only for this non-progress control lot. For example, if there are a plurality of devices of the same type, one or two of them are dedicated to the progress control lot, and the other devices are assigned to the processing schedule for the non-progress control lot. The processing of the device can be treated in the same way by ignoring this division. In that case, the device to be processed is replaced with the device actually processed at the time of data storage and stored.

As a method of more easily creating a processing schedule of a non-progress control lot, there is a method of analogizing from the average processing time and waiting time. The method will be described with reference to FIG. Figure 3
In the same manner as shown in (a) above, the steps after the current processing step are simply arranged in time series, and the scheduled processing order is set for each device. On the other hand, the number of processes that can be processed in one day by each device (the number of processes that can be processed) is separately calculated, and the number of processes is assigned to process from the youngest in the above order, and the process schedule table (FIG. ) To register.
Here, one box in each table indicates that the process of a certain process of a certain lot is assigned. Although only lots and processes are shown in the figure, the processing time, priority, startable time (scheduled time when the previous process ends), and pointer to the next process should be stored together in this table. This is convenient when composing the next processing schedule. With this table, the lots to be processed on a certain day can be estimated. As shown in the figure, if progress control lots are registered in this table in advance, and if they are registered in this table, the rest will be non-progress control lots. There are few corrections when combining the processing schedules.

A method of synthesizing the schedulers scheduled to process lot group A and lot group B will be described with reference to FIG. The composition of the processing schedule is performed for each device. Prepare a processing schedule table for the device. In this table, the lot, process, processing time, and process startable time of each processing process are described. The one with the higher priority is described above, and the one with the same priority is described earlier with the earlier scheduled processing time, and the processing schedule is assigned first. In these tables, results obtained by separately scheduling progress control lots and non-progression control lots are simply joined together into a table (may be prepared as a separate table). The processing startable time is the quantity estimated time of the previous process plus the phase time.
As the synthesizing method, first, the one with a high priority (for the progress control lot) is assigned at the processing scheduled time created first. For the non-progress control lot, the processing startable time is checked, and after that time, it is determined whether there is a free time within the processing time range, and the lot is allocated to a free space. This is also sequentially performed in other steps. The part that does not include the process is turned on the next day.
It should be noted that in this synthesizing process, the first process startable time and the actual scheduled process time deviate. For this reason, a pointer to the next process is prepared for each process in the device processing schedule table so that the next process of which device is recognized can be recognized. If it is reflected in the processing startable time (the same in the following steps), an accurate processing schedule can be created. Compared to the method shown in FIG. 3 in which all lots and all processes are developed at once, this method creates a processing schedule by closing it only in a specific device, which reduces the memory when creating a processing schedule. It is possible.

In the above method, the processing schedule of the progress control lot is fixed, and the processing schedule of the non-progress control lot is inserted into the rest. Therefore, if the idle time is halfway,
The process may be difficult to enter. As an improvement method, moveable time for each processing process for priority lots is put together in a table, and the processing schedule of progress control lots is moved backward within that time range to make free time, and no progress is made. There is a method to add processing for the control lot. For example, the processing schedule t of the progress control lot is from 11:00 to 13 hours, but it may be moved backward if the following processing is up to 2 hours. On the other hand,
When it is attempted to enter for 2 hours from 10:00, the processing schedule u does not exist because there is the processing schedule t. However, the processing schedule t is 1
By shifting the time backward, the processing schedule u can be entered from 10:00 to 12:00. Thus, the use of the movable time improves the efficiency of scheduling.

Here, the order of the apparatus for preparing the processing schedule does not matter in particular, but it is efficient to start processing from the one having a lot of staying lots. Strictly speaking, since it is carried out for each device, when a certain process is moved backward, it is necessary to shift the subsequent process backward and sequentially create a processing schedule for each device. In this case, in the case of an apparatus whose process schedule has already been created, it will be necessary to reallocate it. However, in the case of a mass production line in which the number of lots in the entire line is large and there is a wait in each device, the next process is kept waiting in the next device, so there is almost no effect of shifting the subsequent processes. Absent.
Therefore, for each device, for each short-term processing unit time (for example, about 1 day in the case of an LSI process having a processing time of several tens of minutes to several hours, if the processing of one process is within 1 minute, for example, Processing schedules (for example, about time) may be created independently and in parallel, and the processing start possible time of the next process may be shifted from the result to create a processing schedule for the next processing unit time. In this way, the amount of processing schedule created at one time is only for a certain device for a short period of time, and the processing is sequentially performed, so that the memory used at one time can be extremely small. In addition, since the calculation can be performed in parallel, the processing time can be shortened.

In the above-described method of FIG. 1, since the non-progress control lot has only a short-term processing schedule, the production plan based on the long-term schedule, the progress management of the lot, and the delivery date control cannot be performed. In order to prevent this, there is a method of using a marker lot as a method of making a long-term processing schedule and reducing the time and memory required for the scheduling to roughly manage the progress and control the delivery date. When lots of the same product type are present on a large number of lines like a gate array substrate, lots of N lots (for example, 10 lots) are treated as marker lots in the same manner as the progress control lots, and a long-term processing schedule is created. At this time, the processing schedule is created while the priority is low. This process schedule was created according to the actual process on the line by designating the lot start time or the process start process of the day (the end process may be notified at a certain time). It is possible to make a processing schedule and manage the progress of lots on the line. On the other hand, for lots other than the marker lot, a processing schedule for only a short period is created as a non-progress control lot. The lots are processed in the order of the input lots, and if a plurality of lots are put in a waiting state in the same process, it is determined that the lots that have been input first will be processed first. The other N-1 lots between the marker lots are between the two marker lots, and the progress of the rough lot can be controlled. In addition, data collection can be performed in the same manner as a lot that has been subjected to a normal schedule.

In the production control method described above,
If you want the scheduler to correct the processing schedule for the entire line, perform the following processing (execute steps). Modify the long-term (progress control lot) and / or short-term processing schedule created by the processing procedure in Figure 1 (including new creation)
Then, the processing schedule of the entire line is recreated by combining with the uncorrected processing schedule stored in the scheduler.
It should be noted that since the progress of the lot can be monitored by the scheduler in the past, this function is used to monitor the progress of the lot, and when the scheduler detects that a predetermined abnormal situation has been reached, the above-mentioned processing schedule of the entire line is corrected. Can be applied.

In particular, when the delivery date of the non-progressive control lot is short, the non-progressive control lot is regarded as the progress control target lot and the entire line is processed as described in this embodiment. By recreating the schedule, the progress of the lot can be expedited.

[0034]

According to the production management system and method of the present invention, a long-term processing schedule is first set for only a progress control lot selected from all lots without scheduling all lots. For a lot with low dielectric constant / importance, the calculation time when creating a schedule and the memory used can be reduced, and a long-term schedule is possible. For this reason, it is possible to perform accurate progress control and delivery date management by looking at the long-term forecast once scheduled and performing processing such as raising the priority of the lot that is likely to be delayed, and rescheduling. In addition, by shortening the period of all other lots and creating a processing result, and synthesizing these results, it is possible to control the TAT of important lots and
It is possible to create a processing schedule.

[Brief description of drawings]

FIG. 1 is a flowchart showing a process of creating a process schedule by a production management system of the present invention.

FIG. 2 is a diagram showing a schedule target range at the time of creating a processing schedule.

FIG. 3 is a diagram showing processing schedule creation contents for a progress control lot.

FIG. 4 is a diagram showing the content of creation of a processing schedule for a non-progress control lot.

FIG. 5 is a diagram showing a method of combining a processing schedule for a progress control lot and a processing schedule for a non-progress control lot.

FIG. 6 is a diagram illustrating an outline of a manufacturing procedure of an LSI.

FIG. 7 is a diagram showing an example of a process table.

FIG. 8 is a diagram showing an outline of a conventional LSI production procedure.

FIG. 9 is a diagram showing how to divide a progress control lot and a non-progress control lot.

Claims (4)

[Claims] 1. A process schedule showing manufacturing conditions, a manufacturing order, etc., and a schedule for creating a processing schedule for each manufacturing lot and each manufacturing device from an operation schedule of each manufacturing device. In a production management system in which information on processing schedules is added to the processing data of and stored in a database, the scheduler needs to control the progress in consideration of the production schedule, process schedule, and operation schedule of line resources. A means for creating a long-term processing schedule, a means for separately creating a short-term processing schedule for lots other than the lot, a short-term processing schedule, and a lot processing schedule for which it is necessary to control the progress. And a means for synthesizing to create a processing schedule for the entire line. 2. A predetermined schedule for each manufacturing lot and each manufacturing apparatus is created by a scheduler from a process chart showing manufacturing conditions, manufacturing order, etc., and an operation schedule of each manufacturing apparatus, and processing from the manufacturing apparatus is performed. In a production management method of adding information on processing schedule to data and storing it in a database, the scheduler needs to control the progress in consideration of a production plan, a process chart, and an operation schedule of line resources for a long term of a lot. And a step of separately creating a short-term processing schedule of lots other than the lot, and combining the short-term processing schedule with the processing schedule of the lot whose progress needs to be controlled. And a step of creating a processing schedule for the entire line. 3. The scheduler monitors the progress of the lot, and when an abnormality occurs in the progress, corrects the long-term processing schedule and / or the short-term processing schedule,
3. The production management method according to claim 2, wherein the step of recreating the processing schedule of the entire line by combining with the uncorrected processing schedule is executed. 4. Even if the lot has a low priority, if an abnormality occurs in which the progress is delayed, the step of recreating the processing schedule of the entire line is executed as a lot having a high priority apparently. The production management method according to claim 3, wherein