JP3095292B2 - Load prediction type injection control system - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a process control method for manufacturing a product, and more particularly to an input control system for controlling input of raw materials and semi-finished products to each facility.

[0002]

2. Description of the Related Art Generally, one product is manufactured through a plurality of manufacturing steps. The equipment used for each of these steps is determined for each step. The processing time of each process for one product is unique to each process.

On the other hand, there is an upper limit to the number of products that can be processed at the same time in equipment that processes each process. Also,
The same equipment may be used in the manufacturing process of different types of products. In this case, the time required for processing per product is determined by the type of product.

Usually, a plurality of types of products are often manufactured at a site where the products are manufactured. At this time, particularly when the equipment is expensive, the operation rate of each equipment is controlled to a certain level (high level) by controlling the type (a plurality of kinds may be input at the same time) and quantity of the products to be input to each equipment. )
It is important to keep At the same time, it is necessary to reduce the quantity of products waiting for processing in the process (waiting for work in progress) by the control.

[0005] As a conventional input control method, the current process waiting information (the number of products waiting for processing) of the process concerned is transmitted to the immediately preceding process, and the quantity of the product to be fed into the immediately preceding process is determined based on this information. (See Japanese Patent Application Laid-Open No. 60-15052).

Further, there is a method in which the quantity of a future product is simulated based on the quantity of the current product in the process, and the quantity of the product to be supplied is indicated based on the simulation result (Japanese Patent Application Laid-Open No. 63-312050). Gazette).

[0007]

The former method of the above-mentioned conventional method considers only the information on the current work-in-process of the process concerned, and considers the future load (the number of products to be processed). I haven't. For this reason, when the processing time of the immediately preceding process is sufficiently longer than the processing time of the process, and the type of equipment that can be processed for each product type in the process is limited, the following problem occurs. .

[0008] 1. When the facility operation rate of a certain facility in the process decreases, it is difficult to immediately take measures to increase the facility operation rate of the process because the processing time of the immediately preceding process is long.

[0009] 2. If the equipment in the immediately preceding process is to process a product having a long processing time, the facility is occupied for a long time in the immediately preceding process, which is equivalent to a reduction in process capability. In addition, in this step, the amount of product supplied from the immediately preceding step is reduced, which leads to a reduction in equipment operation rate.

[0010] 3. In order to avoid the above 1 and 2, the process may have a large amount of stock in process. In this case, the production lead time (the time required from the procurement of the material to the completion of the product) becomes long, which causes a reduction in profit.

In the latter of the above-mentioned conventional methods, the starting order is determined for each process on the basis of the status of the own process and the delivery date of the product. We do not consider the situation. For this reason, when the processing time of the immediately preceding process is sufficiently longer than the processing time of the relevant process, and the type of the product to be processed in the current process and the immediately preceding process and the equipment that can perform the process are different, the simulation is not performed even if the simulation is repeated. It is not always possible to obtain the result that the number of products to be processed by the facility increases.

[0012] It is an object of the present invention to solve the above-mentioned problems by not only grasping the current state of the work-in-process of the process concerned and the immediately preceding process and the operating condition of the equipment, but also predicting the future situation and taking this into consideration. In this way, the start of the immediately preceding process is controlled so as to prevent a decrease in the operation rate of the equipment in the process with a small waiting inventory.

[0013]

According to the present invention, there is provided, in accordance with the present invention, a process facility into which a plurality of types of work items are loaded, and wherein A first process facility including a plurality of facilities for processing a work target product in which the value of the first attribute of the target product is a value that can be processed by the user, and a work target product that has been processed in the first process facility Of the input and input work items,
A second apparatus provided with a plurality of facilities for processing a work item in which the value of the second attribute of the work item is a value that can be processed by itself.
And a process status collecting means for collecting process statuses of the first process and the second process.
A process status storage unit for storing the collected process statuses; attribute values of the work target items in the first step and the second step; and attribute values of the work target items that can be processed for each of the facilities. Production information registering means for registering information relating to production, production information storage means for storing registered production information, and an operation target for reading stored process status and production information and inputting to the first process It is possible to include a charging order determining means for determining a product based on the attribute value, and an output means for outputting the work target product determined by the charging order determining means.

The input sequence determining means predicts a future process status in the first process and the second process from the production information and the process status, and determines an operation target product to be input into the first process by an attribute value. Can also be determined based on

Further, the time required for processing per unit input amount of the product to be worked in the first process equipment is long,
The present invention can also be applied to the case where the time required for processing per unit input amount of the work target product in the second process facility is short.

Further, the input sequence determining means predicts the load on the first process equipment and the second process equipment at a predetermined time interval in the future, and sets the first process equipment and the load at each time interval. Setting a target value of the load in the second process equipment, comparing the predicted load with the target value, and increasing the load in a time interval in which a gap between the predicted load and the target value is large. The work item to be put into one process can be determined based on the attribute value.

Further, a plurality of facilities are arranged in the first process facility, and means for classifying and registering the plurality of facilities into a plurality of facility groups based on a predetermined standard may be provided. it can. Then, in the first process facility, for each of the plurality of facility groups, the load representing the scale of the facility that is processing the work item at the time interval can be used. In the second process facility, for each of the plurality of facilities, the load representing the time during which the work item is being processed at the time interval may be used as the load.

Furthermore, in the first process facility, the number of facilities which are processing the work item at time intervals may be set as the capacity indicating the scale for each of the plurality of facility groups. In the second process facility, for each of the plurality of facilities, the integrated value of the time during which the work item is being processed at the time interval may be a capacity representing time.

Further, a plurality of facilities are arranged in the second process facility, and means for classifying and registering the plurality of facilities into a plurality of facility groups based on a predetermined standard may be provided. . In the first process facility, a load representing a time during which the work item is being processed at a time interval may be used for each of the plurality of facilities. In the second process facility, for each of the plurality of facility groups, the load that represents the scale of the facility that is processing the work item at time intervals may be used as the load.

Further, in the first process facility, for each of the plurality of facilities, the integrated value of the time during which the work item is being processed at the time interval may be a capacity representing the time. Second
In the process equipment of (1), the number of the equipment which is processing the work item at the time interval for each of the plurality of equipment groups may be set as the capacity indicating the scale.

[0021]

According to the method of the present invention, a step which is a bottleneck in a series of steps (a long processing time per unit input amount of a product, or a small amount of products which can be processed simultaneously with a small number of facilities) is included. If there is, in order to fully exploit the capability of the process, first, the product currently in process immediately before the process is grasped from the process information collected by the collection unit. Further, the immediately preceding process by these products and the load given to the process are stored in the process status storage means. In this way, the load state of the process is grasped now and in the future. As a result, when and where equipment is available in the future (there is no product to be processed)
Is predicted (evaluated) in advance to determine whether or not will occur. The input sequence determining means starts the process processing by inputting the product to the immediately preceding process so that the product is input to the facility at a time when it is expected that a vacancy will occur. As a result, it is possible to reduce the stock waiting for the process in the process, which was conventionally required to utilize the equipment in the process at a high operation rate, without lowering the operation rate of the equipment.

[0022]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A load prediction type injection control system according to the present invention will be described below with reference to FIGS. In the present embodiment, an LSI (large scale in
A case in which the present invention is applied to a product sorting process will be described.

First, before specifically describing the present invention, L
The SI product sorting process will be described.

FIG. 3 shows an outline of a flow (manufacturing flow) of a manufacturing process of an LSI product. The manufacturing flow in FIG. 3 is roughly divided into a pre-process 100 and a post-process 200.

The preprocess 100 includes a wafer process 1
10 and a wafer test step 120 are included. The wafer process step 110 is a step of drawing a wiring pattern on a wafer (a cylindrical silicon single crystal sliced). The wafer test step 120 is a step of testing the wafer after the wafer process step 110.

The post-process 200 includes an assembling process 210, a molding / forming process 220, and a sorting process 230. In the assembling step 210, a plurality of chips formed on the wafer are cut out one by one and connected to a lead frame (readout line). In the molding / forming step 220, the chip connected to the lead frame as described above is packaged with resin. The shape of a product packaged in this way is taken as an attribute and is called a package. In the selection step 230, a test of the chip packaged with the resin is performed.

FIG. 4 is a detailed manufacturing flow of the sorting step 230 in FIG.

As shown in FIG. 4, in the sorting step 230,
An aging step 231 and a test step 232 are included. In the production process, products are processed in units called lots.

The aging step 231 is a step of applying a heat load to the product. In the aging step 231, the LSI
Is mounted on a jig called an aging board by a detachable machine, and then put into an aging furnace. There are a plurality of types of aging specifications depending on the voltage applied during aging, the temperature, and the like. This is called an aging work division. This work category is determined by the product type (lot attribute). The equipment also determines which aging product to process. Facilities for processing lots of the same aging work category are grouped into the same aging furnace group. L after aging processing
The SI is removed from the aging board again by the detachable machine.

The test step 232 includes the aging step 23
This is a step of testing the function of the product that has gone through 1. In the test step 232, a tester for generating and measuring an electric signal and an LS
Processing is performed in combination with a handler that grasps (handles) I. The type of tester that can be used is limited for each type of LSI. The equipment that can be used for each handler is limited.

The processing time (aging time) in the aging step 231 and the processing time (test time) in the test step 232 both vary depending on the product type. Generally, the aging time is one of the test times.
It is about 0 times. Further, the range of variation in the aging time is very large. By the way, since the equipment (tester) in the test process is expensive, it is important to maintain a high operation rate of the tester in managing the production process.

The problems in the process control in the sorting process 230 will be described below.

Problems 1. The aging process 231 and the test process 232 differ in lot attributes and equipment combination conditions. Therefore, in the test process 232, the model (tester)
Imbalance is likely to occur in each load.

This will be described with reference to FIG.

As shown in FIG. 5A, it is assumed that each equipment (aging furnace) 510 to 530 in the aging step 231 has a limited form of a lot that can be processed. Irrespective of whether or not). FIG.
In the case of (a), the aging furnace 530 can process only a square lot. Furnace 520 can process only round lots and furnace 510 can process triangular lots only.

As shown in FIG. 5B, each equipment (tester) 540 to 560 in the test process 232 is limited in the color of the lot that can be processed (the shape of the lot can be processed). Or not). FIG. 5 (b)
In this case, the tester 540 can process only the black lot. The tester 550 can process only the gray lot, and the tester 560 can process only the white lot.

Referring to FIG. 5, in the aging step 231, when selecting a lot to be started, the shape of the lot is considered, but the color does not need to be considered. Therefore, only the white lot may be started. In that case, as shown in FIG.
Only the in-process waiting amount (the number of lots waiting for processing) for the tester 560 that processes the white lot increases, and the in-process waiting amount for the other testers 540 and 550 decreases. In this case, the tester 540 that processes the black lot
It is expected to be available.

Problem 2. Usually, the processing time of the aging step 231 is longer than the processing time of the test step 232. Further, the processing time varies greatly depending on the lot attribute. For this reason, the payout amount per unit time from the aging process 231 to the test process 232 (the number of lots to be supplied to the test process 232 after the aging process 231 is completed) is likely to fluctuate.

This will be described with reference to FIG.

FIG. 6A shows the state of the load (lot processing time) supplied to the aging furnaces 1 to 5 in the aging step 231. In FIG. 6A, near the time t1, the lots whose processing in the aging step 231 has been completed arrive at the test step 232 as indicated by the arrow group 601. Thereafter, the processing time of the lot started in the aging process 231 (610 to 614)
Aging process 231 near time t2
Is not paid out to the test step 232.

For this reason, as shown in FIG. 6B, in the tester 2, a free space 620 (a time during which the load is not processed by the tester, which is indicated by a shaded portion) occurs. Further, such a problem is more likely to occur as the tendency of the combination of the type and the equipment in the preceding and following steps to be different from each other increases as described in the problem 1. In addition, as shown in FIG. 6B, when the load of each tester is unbalanced in the test process and the tester becomes available, the tester that has been made available at the break of the aging process (node). A lot that can be processed is started in the aging process. However, since the time required for the aging process of one lot is generally long, it takes time until the lot started in the aging process is paid out to the test process and there is no free time in the test process (time lag). There is).

Problem 3. In response to these problems (problems 1 and 2), the test process 232 maintains a high operation rate of the tester by having a large amount of buffer stock (stock in process). For this reason, the lead time of the test process (the time from when the lot enters a state of waiting for the in-process test process to when the process of the test process is completed) is long.

Therefore, the load on each equipment in the test process 232 is leveled with a small waiting inventory, and the input of products in the sorting process 230 is controlled so that each equipment always maintains a high operation rate. Is required.

Now, the load prediction type injection control system according to the present invention will be specifically described.

FIGS. 1 and 12 are block diagrams showing an embodiment of the configuration of a sorting process management system in which the load prediction type input control system of the present invention is applied to a sorting process in an LSI manufacturing process.

FIG. 12 is a diagram showing a hardware configuration of the sorting process management system.

The system shown in FIG. 12 includes a CPU 51, a keyboard 21 connected thereto, and an interface device 2.
2, a CRT 61, a printer 62, and a memory 71. Further, at least two or more facility-network interfaces 31 and a CRT 33 are connected to the CPU 51 via the network 32. An aging process facility 91 or a test process facility 92 is connected to the interface 31.

FIG. 1 is a diagram showing the configuration focusing on the function realized by the hardware of FIG.

The system shown in FIG.
Production information registration means 20 and output means 60; production information registration means 10 and process status storage means 40;
0. The collection / instruction means 30 includes a process facility 90.
Is connected.

The process equipment 90 shown in FIG. 1 includes an aging process equipment 91 and a test process equipment 92 shown in FIG.

The collection / instruction means 30 shown in FIG. 1 comprises the equipment-network interface 31, CRT 33, network 32 and a part of the CPU 51 shown in FIG. The collection / instruction means 30 is a device for collecting information from the aging process equipment 91 and the test process equipment 92,
An interface between the network 92 and the network 32, and
It is responsible for transmitting the collected information over the network and storing it in memory. In addition, the operator of the equipments 91 and 92 is instructed about the lot to be put in and the timing of the lot.

The production information storage means 10 shown in FIG. 1 is the memory 71 shown in FIG. The production information storage means 10 stores information related to production, such as a combination condition of a product type and equipment and a processing time.

The process status storage means 40 of FIG. 1 is also the memory 71 of FIG. The process status storage means 40 stores information (operation status, in-process inventory, etc.) collected from the aging process equipment 91 and the test process equipment 92.

The production information registration means 20 shown in FIG. 1 includes the keyboard 21 shown in FIG. 12 and an interface device 22 with another system. The production information registration means 20
Information to be stored in the production information storage means 10 is input by a work supervisor of a production process or the like.

The input order determining means 50 in FIG. 1 is the CPU 51 in FIG. The input order determining means 50 determines the input amount and timing of lots to the process equipment 90.

The output means 60 shown in FIG.
1 and a printer 62. Output means 60
Informs the work supervisor of the equipment 90 about the lot to be put in and the timing.

Next, the operation of FIG. 1 will be described.

In FIG. 1, information related to a process such as an operation rate and a stock waiting for processing is collected from a process facility 90. These pieces of information are stored in the process status storage unit 40 via the collection / instruction unit 30. On the other hand, information relating to production, such as a combination condition of a product type and equipment and a processing time, is input via the production information registration means 20. These pieces of information are stored in the production information storage unit 10. The input order determining means 50 determines the lot to be input and its timing based on the information related to these processes and production. The determined lot and time are transmitted to the operation supervisor of the facility 90 via the display means 60. In addition,
The information is transmitted to the operator of the facility 90 via the instruction means 30.

Next, a description will be given of a procedure for determining the order of inputting the products by the input order determining means 50.

First, in this procedure, equipment is grouped for each equipment having the same product type that can be processed. For each group of the facilities, the future (time axis) is divided by a certain unit time, and each of the divided unit times is called a bucket. For each bucket, the load of the test process (the load of the bucket in the test process refers to the type of equipment that arrives at the process by the beginning of the time frame of the bucket and is represented by the bucket within the time of the bucket). Estimate the future value of the processing time of the lot to be processed in the facility). In addition, a calculation method of a criterion (load evaluation criterion) for determining whether or not there is an appropriate load for each bucket of each tester to such an extent that the equipment is not available is set. Then, the product is put into the aging step and the amount of the product to be started in the aging step is controlled so that a load is applied to the bucket determined to have a small load according to the load evaluation criteria.

FIG. 7 shows an example of the above bucket. In this example, the tester has two groups (tester 1 and tester 2).
are categorized. The time axis is divided in units of 12 hours. Each of the frames divided into 12-hour units for each equipment type is called a bucket. The size of this bucket indicates the amount of load that can be processed within that time. In FIG. 7, the hatched portions indicate the amount of load expected to be processed in the bucket. However, if the amount of load expected for a certain bucket exceeds the size of the bucket, the load is carried over to the next bucket (the load is added to the next bucket).

For a bucket that is temporally close to the start of the product introduction plan for the process, the load reference is set to 100%. This means that the load at hand is secured in order to prevent equipment from being left unoccupied.

On the other hand, the load reference of the bucket temporally far from the start of the plan is set low.

The reason will be described. If the load reference value of a distant bucket is high, the start of construction is instructed in the immediately preceding process so that a load is generated on this bucket. To generate a load on a bucket in a test step that is far from the time point is to start processing a lot with a long aging time in the aging step in the present embodiment. As described above, when the lots with a long aging time are concentrated and started at once, as described in the problem 2 of the problem to be solved by the invention, the process capability of the immediately preceding process is reduced. Eventually, the equipment for the process is left unoccupied.

On the other hand, if the load reference of the bucket in the test step far from the time is excessively low, only the lot having the short processing time will be started in the immediately preceding step. For this reason, only lots with a long processing time are in process waiting, and the load cannot be leveled. The load reference of the bucket which is far from the time should be set so that the load on the bucket does not depend on the distribution of time and quantity required for processing the product.

The concept of a bucket is referred to as an MRP system (material requirement pl.).
The system is also used in the planning system for material requirements. However, in the present invention, the lead time of one process (in-process waiting time + processing time) is
MR allows for different types of products,
Different from the P system. Also, the MRP system differs from the MRP system in that the bucket size may be different depending on the period (for example, the size of a bucket that is far in time may be large and the future in the far future may be roughly grasped).

Next, the prerequisites for this embodiment will be described, and then the details will be described.

1. The process is a two-stage flow shop type (there are two processes and the order is determined), and the flow is from the aging process to the test process.

2. Products (a) A plurality of types of products exist in the process at the same time. (b) The product flows through the process in lot units.

3. Equipment (a) In each step, there are a plurality of types of equipment, and the equipment that can be processed according to the type of product is limited. (b) In each facility, products are started one lot at a time. (c) No equipment failure is considered.

4. Processing time (a) Determined by the combination of equipment and product type. (b) In the aging step, the length is an integral multiple of the unit processing time, and in the test step, the length is an arbitrary length.

Next, the symbols used in the procedure will be described.

1. Bucket Load / Capacity The bucket in the aging process (waiting for work and during work) is a unit in which the future time axis is divided by unit time for each aging furnace group. The load of the bucket in the aging process is the number of furnaces in the aging furnace group that is processing a lot in the bucket. The capacity of the bucket in the aging process is the number of furnaces in the aging furnace group that can process a lot in the bucket.

The bucket in the test process (waiting for and in process) is a test process facility (tester)
In each case, the future time axis is divided by unit time.
The load on the bucket in the test process is an integrated value of the processing time of the lot processed in the bucket. The capacity of the bucket in the test process is the processing time of a lot that can be processed in the bucket.

(A) Aging process in process waiting WALOAD (ag, at, th, p): Aging work category (ag) (ie, aging furnace group), aging time (at)
Separately, test process equipment (th) Load of bucket WA (ag, at, th, p) waiting for in-process aging in p period (b) Aging process in progress ・ ALOAD (ag, p): Aging work category (ag) Load of bucket A (ag, p) for each p period (by aging furnace group) ・ AABL (ag, p): Capacity of bucket A (c) Waiting for work in process for the test process ・ WTLOAD (th, p) : Load on test process work-in-process waiting bucket WT (th, p) for each p period in test process equipment type (th) ・ WTABL (th, p): Capacity of bucket WT WTABL (th, p) It means the upper limit of the amount of work in progress. (d) Test process in progress ・ TLOAD (th, p): Bucket T of p period by test process equipment
Load on (th, p) • TABL (th, p): Capacity of bucket T (th, p) • TC (th, p + ap): Coefficient of load pile upper limit on bucket at the end of ap for each test process facility.

2. Step (a) The current time is set to 0, and the unit of time (when all bucket sizes are equal, the unit of time for one bucket)
Tu. (b) The current term is the first term bucket. (c) The maximum value of the period p is defined as Mp.

3. Product (a) p_atr-th attribute of product type j: PRO (j, p_atr) PRO (j, 1): Aging work category (ag) PRO (j, 2): Aging time (at) PRO (j, 3): Usable test process equipment type (th) ・ PRO (j, 4): Test time (b) l_atr-th attribute of lot i: LOT (i, l_atr) ・ LOT (i, 1) ): Type (j) LOT (i, 2): In-process place (sn) LOT (i, 2) includes the following 1 to 6.

1; not input; 2; waiting for the aging process; 3; during the aging process; 4; waiting for the testing process; 5; during the testing process; 6; end. LOT (i, 3) : Time when the current state (sn) is reached ・ LOT (i, 4): Time when the aging process can be input.

4. Processing time (a) Aging time PRO (j, 2) (aging time of product type j) is n times Tu (n is a natural number and the maximum value of n is Mn) (b) Test time PRO (j, 4) is r times Tu (r is a real number). The procedure (procedures 1 to 5) according to the present invention will be described in detail below.

Steps 1, 2, and 3 are preparation stages, and step 4
Perform evaluation and instructions. Procedure 5 is a process for evaluating all the periods p.

Procedure 1. The load of the test process is set to bucket T (t
h, p). However, th = PRO (LOT (i, 1), 3) That is, as for the available test process equipment type, th procedure 1 is realized by the following (1) and (2).

(1) LOT (i, 2) = 5 (Test process is in progress)
Calculated by Formula 1 for lots.

[0083]

TLOAD (th, p) = PRO (LOT (i, 1), 4) − (current time−LOT (i, 3)) That is, this equation is: (p period bucket for each test process equipment) T (th, p) load = (test time)-((current time)-(current time)). According to this formula, the remaining processing time of the lot currently undergoing the test processing is loaded in the bucket.

(2) LOT (i, 2) = 4 (Waiting for test process in process) is calculated by the formula (7).

[0085]

TLOAD (th, p) = TLOAD (th, p) + PRO (LOT (i, 1),
4) That is, this equation is: (Load on bucket T (th, p) in p period for each test process facility) = (bucket T (th, p) in p period for each test process facility)
(Load on test) + (test time) By this formula, the processing time in the facility of the lot waiting to be processed for the facility is added. However, the values of TLOAD (th, p) and TABL (th, p) are compared, and if the magnitude relationship is as shown in Expression 8, the remaining load is p = p + 1
And load the next bucket of the same type of equipment.

[0086]

TLOAD (th, p)> TABL (th, p) That is, this formula is: (load on bucket T (th, p) in period p for each test process facility)> (bucket T (th, p) ), Which means that the added value of the load exceeds the capacity of the bucket. Therefore, when the equation 8 is satisfied, the remaining load is loaded on the next bucket of the same type of equipment as p = p + 1. This means that if the added value exceeds the capacity of the bucket, the next load is calculated for the excess. It means to carry over to the bucket.

The procedure 1 will be described with reference to FIG. To load the bucket means to add up the processing time of the lot processed by the bucket. In (1), the remaining processing time of the lot currently being processed is loaded in the bucket.
In (2), the processing time in the equipment of the lot waiting for the in-process of the equipment is added. here,
Equation 8 means that the added value of the processing time exceeds the capacity of the bucket. In such a case, the excess is carried over to the next bucket.

Procedure 2. The load during the aging process (LOT (i, 2) = 3 lots) is converted into a bucket A (ag, p) (p-phase bucket of the aging furnace group ag) and a bucket T
(th, p ') (p' period bucket for each test process equipment). Where ag = PRO (LOT (i, 1), 1), p '= LOT (i, 3) + PRO (i, 2) That is, ag is in the aging furnace group p' is in the current state Addition value of timing and aging time Step 2 is realized by the following (1) and (2).

(1) Load is applied to the bucket A according to the equation (2).

[0090]

ALOAD (ag, 1 + n) = ALOAD (ag, 1 + n) +1 where n = 0,1, ..., LOT (i, 3) + PRO (i, 2) -2 , (Aging furnace group ag bucket load in 1 + n period)
= (Load of bucket in 1 + n period of aging furnace group ag) +1 where n = 0, 1,..., (Time of current state) + (aging time) -2. According to this equation, loads are loaded one by one from the bucket in the first period of the aging process to the bucket in the period of ((current state) + (aging time) −1).

(2) Load is applied to the bucket T according to the equation (3).

[0092]

TLOAD (th, p ') = TLOAD (th, p') + PRO (LOT (i, 1), 4) If TABL (th, P ') <TLOAD (th, p') , Step 1
As in (1), the remaining loads are sequentially loaded on the next bucket. This formula is as follows: (load of p 'period bucket for each test process equipment) = (load of p' period bucket for each test process equipment) + test time, where p 'is the time when the current state is reached. This shows the added value of the aging time. The load of the bucket in the p ′ period for each test process facility calculated by this equation is compared with the capacity of the bucket in the p ′ period for each test process facility. As a result of this comparison, if the load exceeds the capacity, the load exceeding the capacity is sequentially loaded on the next bucket.

The procedure 2 (1) will be described with reference to FIG. When the aging furnace starts a certain lot, the equipment is in use for the length of the processing time of the lot. For this reason,
The load and the capacity are respectively counted by the number of facilities being processed and the number of facilities belonging to the facility group. In FIG.
At the beginning of the first stage of the bucket, a new lot with a processing time of three periods is started in the aging furnace group 1 (shaded area).
The number of facilities in group 1 is four, and two of them are already being processed (shaded area), so bucket A (1,1)
Will increase from 2 to 3. Also, n in Equation 2
+1 indicates the period when the processing of the lot ends. In the lot we are currently focusing on, LOT (i, 3) is 1 and PRO (i, 2) is 3, so the load on Group 1 is 1 for the first to third periods.
Increase by one.

Procedure 3. The in-process waiting load of the aging process is loaded on the bucket WA (ag, at, th, p).

(1) LOT (i, 2) = 2 (waiting for the aging process in process) lot is calculated by Equation 4.

[0096]

[Expression 4] WALOAD (ag, at, th, p) = WALOAD (ag, at, th, p) +
1 where j = LOT (i, 1), ag = PRO (j, 1), at = PRO (j, 2), th
= PRO (j, 3), p = LOT (i, 4)
It indicates that the load is loaded on the bucket at at, the available test process equipment type is th, and the aging process inputtable time is the p period.

Step 3 will be described with reference to FIG. In step 3, the number of lots waiting for the aging process is counted.
When counting, it is considered to classify by aging work category (that is, by aging furnace group), by aging time, by test process equipment group, and by aging process start period.

Procedure 4. The aging process start order is determined so that a load is generated in the bucket T having a high value of the load evaluation criterion CTLOAD (th, p + ap). See FIG. ap is 1, Mn, Mn-1,... Here, Mn is the maximum value of n in the aging time = Tu × n. Procedure 4 is realized by the following (1) to (8).

(1) Set ap = 1.

The criteria for selecting ap for determining from which stage of the test process the load is to be applied to the bucket is based on the method currently determined by the site manager. This is a rule that first secures the load on the current bucket and then starts aging for a long time. This prevents aging products from being left behind for a long time.

(2) For all th, the load evaluation criterion of the bucket T (th, p + ap) is calculated by Equation 5.

[0102]

Equation 5 CTLOAD (th, p + ap) = TABL (th, p + ap) × TC (th, p +
ap) −TLOAD (th, p + ap) This formula is: (Evaluation standard for load of bucket in p + ap period of test process equipment type th) = (Capacity of bucket in p + ap period of test process equipment type th) × (Test process equipment type th th p + ap period bucket load pile upper limit determination coefficient)-(Test process equipment type th
p + ap period bucket load). The load stacking upper limit determination count TC for each bucket used in the evaluation criterion calculation formula (Equation 5) was determined according to the following formula in accordance with the production quantity ratio for each product for each processing time of the aging process.

[0103]

TC (th, p + ap) = (PRO (j, 2) ≧ (ap−1) × total production of product j of Tu) / (total production) This indicates that the load pile upper limit determination count in the ap period) = (total production amount of products whose aging time is (ap−1) times or more of Tu (predetermined unit time)) / (total production amount). .

This is based on the premise that the processing times of the aging step and the test step are uncorrelated, and that products having various processing times arrive at the aging step randomly in proportion to the production amount. By providing such an upper limit of the pile of loads, it is possible to prevent a product of a type having a long aging processing time from occupying the furnace and temporarily stopping the delivery to the test process.

(3) For all th (test process equipment type), if CTLOAD (th, p + ap) ≦ 0, go to the next ap. If ap = 2, go to step 5. The above description indicates that if (the load evaluation criterion of the bucket in the p + ap period of the test process equipment type th) is 0 or less, the next ap is evaluated. However, when ap = 2, it indicates that the procedure proceeds to the procedure 5.

(4) Selection of test process equipment th subject to load allocation Selection rule: th of CTLOAD (th, p + ap) is large, that is, (load evaluation criteria of p + ap period of test process equipment type th) Select in order from the largest test process equipment type.

(5) Selection of load-assigned aging work category ag Selection rule: if ap = 1, select ag with large WALOAD (ag, at = ap, th, p) If ap ≠ 1, then aABL (ag , p + 1) −ALOAD (ag, p + 1)
g is selected. That is, if ap = 1, the aging furnace group ag, the aging time ap, and the ag with a large load on the p-stage aging process ready bucket of the test process equipment th are selected. If ap ≠ 1, (Aging furnace group ag, capacity of bucket in period p + 1)
-(Aging furnace group ag, bucket load in period p + 1) Select the aging furnace group ag with a large calculated value.

As described above, in the selection of the aging furnace to which the load is to be assigned, when the load of the next test process is considered (ap = 1), there is no effect after the next test period. Consider reducing in-process waiting. If there is an impact after the next period (ap ≠ 1),
The aging furnace is selected from furnaces with large spare capacity. This aims at leveling the load on the aging furnace.

(6) If ALOAD (ag, p) <AABL (ag, p) is true
(7) If false (8) The above compares (aged furnace group ag, load of bucket in p period) and (aged furnace group ag, capacity of bucket in p period), and the load is greater than the capacity. If smaller,
The process proceeds to (7), and indicates that the process proceeds to (8) if the load is equal to or more than the capacity. That is, it is checked whether the aging furnace has enough capacity.

(7) Start Lot Selection and Processing Lots of the bucket WA (ag, at = ap, th, p) are started with the LOT (i, 3) as early as possible, and buckets are processed in the same manner as in the procedure 2. The load is loaded on A and the bucket T, and the load evaluation criterion of the bucket is updated according to Equation 5, and the process returns to (3). If there is no corresponding lot, go to (8). That is, (4) and (5)
Are selected in the order of arrival time, that is, first-in first-out, from the lots of the waiting buckets WA determined by the combination of the equipment selected in the above.

(8) The next aging work section of interest is selected according to the same rules as in (5). If there is no capacity for all aging work categories ag or there is no corresponding lot, set CTLOAD (th, p + ap) = 0 and go to (3).

This equation indicates (load evaluation criterion of bucket in p + ap period of test process equipment type th) = 0.

That is, if there is no appropriate lot in the selected aging work category, the lot is excluded, and the next assigned work category is selected from the remaining work categories according to the rule (5).
If there is no appropriate lot in all the work categories, the load evaluation criterion of the test process bucket is set to 0, and is excluded from the load allocation.

Procedure 5. If p = Mp−1, end. p ≠ M
If p−1, update the data as p = p + 1,
Return to step 3.

FIG. 2 is a schematic flowchart showing the processing from procedure 1 to procedure 5 described above.

In FIG. 2, step S300 shows the above procedure 1. Step S310 shows procedure 2, and step S320 shows procedure 3. Step S330 to step S350 show procedure 4. Step S360,
S370 shows the procedure 5.

In the present embodiment, a plurality of facilities are arranged in the aging process facility, and the plurality of facilities are classified into a plurality of facility groups and registered. In the aging process equipment, the load is the number of pieces of equipment that are processing the work item in the bucket for each equipment group. In the test process equipment, the load is the integrated value of the time during which the work item is being processed in the bucket for each equipment. However, when a plurality of facilities are arranged in the test process facility, the plurality of facilities may be classified into a plurality of facility groups and registered. In this case, in the test process equipment, for each equipment group, the load may be the number of equipment that is processing the work item in the bucket. In the aging process equipment, the load may be the integrated value of the time during which the work target product is being processed in the bucket for each equipment.

[0118]

According to the present invention, it is possible to predict in advance the time when the equipment operation rate of the bottleneck process decreases in the flow shop type process. At that time (the time when the operating rate will decrease), the process of the product is started in the upstream process (the process performed before the process) so that the load is generated in the process. can do. In this way, it is possible to prevent the equipment operation rate of the process serving as a bottleneck from decreasing. This eliminates the necessity of having a large amount of stock waiting for processing in order to increase the equipment operation rate in the neck process as in the related art.

[Brief description of the drawings]

FIG. 1 is a diagram showing a configuration of an embodiment of a sorting process management system according to the present invention.

FIG. 2 is a schematic diagram of a load evaluation method for making control according to the present invention.

FIG. 3 is a diagram showing a manufacturing flow of an LSI product.

FIG. 4 is a diagram showing a detailed manufacturing flow of a sorting step.

FIG. 5 is a view showing a problem of a sorting step (part 1).

FIG. 6 is a view showing a problem of a sorting step (part 2).

FIG. 7 is a view showing a concept of a bucket.

FIG. 8 is a diagram showing the concept of procedure 1.

FIG. 9 is a diagram showing the concept of procedure 2.

FIG. 10 is a diagram showing the concept of procedure 3.

FIG. 11 is a diagram showing the concept of procedure 4.

FIG. 12 is a diagram showing a hardware configuration of an embodiment of the sorting process management system according to the present invention.

[Explanation of symbols]

10 production information storage means, 20 production information registration means, 2
DESCRIPTION OF SYMBOLS 1 ... Keyboard, 22 ... Interface apparatus with other systems, 30 ... Collection / instruction means, 31 ... Equipment-network interface, 32 ... Network, 40 ... Process status storage means, 50 ... Input order determination means, 60 ... Display means, 61: CRT, 62: Printer, 91: Aging process equipment, 92: Test process equipment.

──────────────────────────────────────────────────続 き Continuing from the front page (72) Inventor Takemasa Iwasaki 292, Yoshida-cho, Totsuka-ku, Yokohama-shi, Kanagawa Pref. JP-A-60-137705 (JP, A) JP-A-2-226468 (JP, A) JP-A-4-346106 (JP, A) JP-A-60-155092 (JP, A) JP-A-63-312050 (JP JP, A) JP-A-4-3477769 (JP, A) Carlson WH, Shafer PH, "MATERIAL RE QUIREMENTS PLANNING FOR SEMICONDUCTOR RS AND OTHER PROCESS SS INDUSTRIES" PLAN NING, USA, AMERICA PRODUCTION AND INV ENTRY CONTROL SOC IETY, 1985, p. 28-32 (58) Fields investigated (Int. Cl. ⁷ , DB name) G06F 17/60 G05B 19/418 INSPEC (DIALOG) JICST file (JOIS) WPI (DIALOG)

Claims (8)

(57) [Claims] An object of the present invention is a process facility into which a plurality of types of work items are input, and among the input work items, a value of a first attribute of the work item is a value that can be processed by itself. A first process facility having a plurality of facilities for processing a certain work target product, and a work target product that has undergone processing in the first process facility is input, and among the input work target products, A second process facility including a plurality of facilities for processing a work item in which the value of the second attribute is a value that can be processed by the user; and collecting process statuses of the first process and the second process. Process status collecting means; process status storing means for storing the collected process status; attribute values of work target items of the first process and the second process; Information on production, including at least the attribute values of the Production information registering means for registering, production information storage means for storing the registered production information, reading the stored process status and production information, and setting the work item to be input to the first process as an attribute value A load prediction type injection control system, comprising: an input order determining unit that determines based on the input order; and an output unit that outputs the work item determined by the input order determining unit. 2. The method according to claim 1, wherein the input order determining means predicts a future process status in the first process and the second process from the production information and the process status, and A load prediction type input control system, wherein a work item to be input to one process is determined based on an attribute value. 3. The method according to claim 2, wherein the first process equipment requires a long time for processing per input unit amount of the work object, and the second work equipment requires a long time for processing. A load prediction type input control system characterized in that the time required for processing per input unit amount is short. 4. The method according to claim 3, wherein the input order determining means predicts a load on the first process facility and the second process facility at a predetermined time interval in the future, and The first process equipment and the second
Setting a target value of the load in the process equipment, comparing the predicted load with the target value, and increasing the load in a time interval where the gap between the predicted load and the target value is large. A load prediction type input control system, wherein a work item to be input to the first step is determined based on an attribute value. 5. The apparatus according to claim 4, wherein a plurality of facilities are arranged in the first process facility, and the plurality of facilities are classified and registered into a plurality of facility groups based on a predetermined standard. The load is, in the first process facility, a capacity representing a scale of a facility which is processing the work item at the time interval for each of the plurality of facility groups; In the process equipment, for each of the plurality of equipment,
A load prediction type injection control system, wherein the time interval is a capacity indicating a processing time of the work target item. 6. The capacity according to claim 5, wherein the capacity indicating the scale is:
For each of the plurality of equipment groups, the number of facilities that are processing the work item in the time interval, the capacity representing the time, in the second process facility, for each of the plurality of facilities, A load prediction type injection control system, characterized in that it is an integrated value of the processing time of the work target item at a time interval. 7. The apparatus according to claim 4, wherein a plurality of facilities are arranged in the second process facility, and the plurality of facilities are classified into a plurality of facility groups based on a predetermined standard and registered. The load is, in the first process equipment, for each of the plurality of equipment,
It is a capacity representing the processing time of the work item in the time interval, In the second process equipment, for each of the plurality of equipment groups, the scale of the equipment that is processing the work item in the time interval A load prediction type injection control system characterized in that the capacity is expressed. 8. The system according to claim 7, wherein the capacity representing the time is an integrated value of the processing time of the work item at the time interval for each of the plurality of facilities in the first process facility; The capacity indicating the scale is, in the second process facility,
A load prediction type injection control system, wherein for each of the plurality of equipment groups, the number of equipment that is processing the work target item at the time interval is the number of equipment.