Data processing: generic control systems or specific application – Specific application – apparatus or process – Product assembly or manufacturing
Reexamination Certificate
2001-05-31
2004-09-07
Paladini, Albert W. (Department: 2125)
Data processing: generic control systems or specific application
Specific application, apparatus or process
Product assembly or manufacturing
C700S095000, C700S291000, C703S022000
Reexamination Certificate
active
06788985
ABSTRACT:
CROSS-REFERENCE TO RELATED APPLICATIONS
This application is based upon and claims the benefit of priority from the prior Japanese Patent Applications No. 2000-163042, filed May 31, 2000; and No. 2000-163043, filed May 31, 2000, the entire contents of both of which are incorporated herein by reference.
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to an electronic commerce method and system for semiconductor products in case of conducting electronic commerce for semiconductor products via a network and a production system, a production method, a production equipment design system, a production equipment design method, and a production equipment manufacturing method for effective operations in a factory.
2. Description of the Related Art
Conventionally, a typical semiconductor factory monthly produces general-purpose products such as memory chips on the basis of several thousand lots. A production line includes too may lots and requires a long period of production. Because of this, it has been difficult to estimate the completion of the product after it went into production. Even in this situation, general-purpose products need not be especially considered regarding input of a lot in accordance with the delivery time, causing no serious problems. Generally, one lot can take in about 25 to 50 wafers. Of course, the lot can take in about 1 to 50 or 100 wafers.
On the other hand, a semiconductor factory in a SOC (System On Chip) age is considered to chiefly produce system LSI chips on a scale of several hundred lots as a monthly production in accordance with customer requests. Such a small-scale factor (hereafter referred to as the mini-fab) needs to input a necessary amount of lots and follow the delivery time by conducting a proper lot progress management. Further, it is necessary to determine whether it is possible to actually manufacture the product in accordance with customer requests such as specification, quantity, delivery time, price, and the like.
However, it has been difficult for conventional mini-fabs to strictly control the lot progress management and to correctly estimate whether the product can be manufactured by following the delivery time. In semiconductor products such as LSI chips, it is considered to drastically increase business opportunities by constructing an electronic commerce using networks such as the Internet. However, since it is difficult to conduct the lot progress management and estimate the product manufacturing, it has been very difficult to implement an electronic commerce for these semiconductor products.
Hence, it has been difficult for conventional semiconductor factories to estimate whether it is possible to conduct the lot progress management and manufacture the product. This has been a cause of losing business opportunities for mini-fabs in a SOC age.
Generally, conventional typical semiconductor factories use as many as dozens of apparatuses for the same purpose at various processes. The same type of apparatuses process many lots, making it difficult to control a flow of lots. As a system for controlling a flow of lots, there is provided the software called “ManSim” from TYECIN Systems, Inc. Input information includes apparatuses used for each process of a product, processing times, apparatus groups, and the like. Lots are allowed to flow on a computer virtually. The system aims at controlling a flow of lots, optimizing production lines, and conducting production scheduling.
To optimize production lines and conduct production scheduling, it is necessary to transfer various information such as lot progress information on an actual production line, information about apparatus states, product's process information, and the like to a computer system. A progress estimate is computed through the use of these types of information as input data. The resulting information needs to be transferred to the actual production line as a work instruction. However, on a large-scale production system characterized by a monthly production of several thousand lots, the progress estimate is computed by simplifying various processes due to restrictions on computer throughput. Accordingly, such a system does not necessarily conduct accurate simulation.
A similar method is proposed in Jpn. Pat. Appln. KOKAI Publication No. 10-207506. The manufacturing management system proposed therein exchanges trial production system information via shared information and uses a result of the simulation to manage a manufacturing process for the production or trial production. According to this technique, a computer system chiefly contains a device simulation function, a process simulation function, circuit, shape, logic simulations functions, and the like, but not a simulation function for flowing lots. This has been the problem of not estimating a lot flow.
FIG. 1
exemplifies a result of computing a throughput and a work period by using ManSim. In this figure, the abscissa axis shows the number of lots (work in process: WIP) within a production line. The ordinate axis shows the throughput (monthly quantity of output) and the work period. Solid lines indicates results of computing a throughput and a work period, and a dotted line indicates actual result of a throughput for reference. According to this figure, when the WIP is small, the throughput is proportional to the WIP and the work period remains constant. This state causes little wait conditions in a lot. When the WIP increases, the throughput gradient decreases gradually, and finally becomes a constant value. It is known that this throughput corresponds to the throughput of a bottlenecked apparatus. Within this region, the work period increases in proportion to the WIP.
Increasing productivity of the production line requires increasing the throughput and shortening the work period. Shortening the work period needs to decrease the number of waiting lots. In this figure, the WIP needs to be set approximately to value A. However, this is not practical because the throughput is too small. By contrast, increasing the WIP approximate to value C in the figure maximizes the throughput, but lengthens the work period. Accordingly, it is considered to be appropriate for operations to use values approximate to B in the figure.
As indicated with a broken line in
FIG. 1
, however, the throughput and productivity decreases due to maintenance or failures of apparatuses, inconsistent arrival of products to a bottlenecked apparatus, and the like. To prevent the throughput from decreasing, it is necessary to accurately predict the progress of lots and conduct optimal processing for increasing the throughput and shortening the work period. As mentioned above, however, a large-scale production system must simplify various processes for computation due to restrictions on computer throughput. It has been difficult to accurately estimate the progress of lots.
Besides, several choices may occur when a certain apparatus processes lots. For example, it is assumed that there is provided a batch apparatus which can process a plurality of lots at a time. When a given lot waits for processing, it is necessary to determine whether to process that lot immediately or to wait until another lot arrives. On a given apparatus, a lot with a low priority waits and a lot with a high priority is expected to occur after a specified time. In this case, it is necessary to determine whether to process the low-priority lot first or to process the high-priority lot first by suspending the low-priority lot. In addition, when there is provided a continuous process such as pre-treatment, oxidation (or CVD), and then post-treatment within 24 hours, it is necessary to determine at which timing the processing should start.
There may be a variety of methods for selecting an optimal one from a plurality of choices as mentioned above depending on situations. Above-mentioned ManSim uniquely determines a rule for selecting choices and computes a lot progress under the corresponding condition. When the above-mentioned choices occur, Man
Mitsutake Kunihiro
Okumura Katsuya
Kosowski Alexander
Paladini Albert W.
LandOfFree
ELECTRONIC COMMERCE METHOD FOR SEMICONDUCTOR PRODUCTS,... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with ELECTRONIC COMMERCE METHOD FOR SEMICONDUCTOR PRODUCTS,..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and ELECTRONIC COMMERCE METHOD FOR SEMICONDUCTOR PRODUCTS,... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3240639