TWI883808B - Production management system - Google Patents

Abstract

A production management system includes a control module. The control module is suitable for signally connecting to a plurality of product machines. The control module performs a loop calculation based on production information to schedule products or check load status of the plurality of product machines.

Description

Production Management System

The present invention relates to a production management system, and in particular to a production management system applicable to semiconductor manufacturing processes.

For semiconductor process control, existing process control methods often focus on the control inside the machine and the monitoring of the machine itself. In addition, when there is a need to execute product production or corresponding processes, the corresponding units or personnel will carry out scheduling planning or check the load status and perform corresponding processing.

However, with the development of semiconductor manufacturing processes, the increase in the number of production machines and/or the expansion of factory areas, existing process control methods are bound to face considerable bottlenecks or inefficiencies. Therefore, how to improve production management and/or improve the efficiency of the corresponding product machines has become a research topic.

The present invention provides a production management system, which is suitable for scheduling shipments or checking load status of multiple product machines.

The production management system of the present invention includes a control module. The control module is suitable for signal connection to multiple product machines. The control module performs loop calculations based on production information to arrange goods for multiple product machines or check the load status.

In one embodiment of the present invention, the production information includes production machine information, work-in-progress information, production time information, product information and/or available time information.

In one embodiment of the present invention, the production information is dynamically adjusted, which includes: the control module adjusts the production information required for the N+1th loop in the loop calculation according to the result of the Nth loop in the loop calculation, where N is a positive integer.

In one embodiment of the present invention, the loop calculation includes at least a function composed of the number of run lots in the production information, the standard working time (TC) of the product machine, and the uptime (UpTime) of the product machine. In one embodiment, the aforementioned function includes [Formula 1] described in the implementation method.

In one embodiment of the present invention, the loop calculation includes at least a function composed of the number of work-in-progress (WIP number) in the work-in-progress information, the machine capacity (Cap.) of multiple product machines, and the available operation time (UpTime) of each product machine. In one embodiment, the aforementioned function includes [Formula 2] described in the implementation method. In one embodiment, the aforementioned function further includes [Formula 3] described in the implementation method.

In one embodiment of the present invention, the loop calculation includes at least a function composed of the operating time ratio of each product machine, the standard working time (TC) of multiple product machines, and the uptime (UpTime) of each product machine. In one embodiment, the aforementioned function includes [Formula 4] described in the implementation method.

In one embodiment of the present invention, some of the plurality of product tools form a corresponding homogeneous tool group; and the loop calculation is to at least reduce the difference in quantity/ratio of executing product production and/or corresponding process of different product tools in the homogeneous tool group.

In one embodiment of the present invention, some of the multiple product machines constitute a corresponding homogeneous machine group; and the control module further simulates and adjusts the quantity/ratio of executing product production and/or corresponding processes of different product machines in the homogeneous machine group according to the calculation results of the loop calculation.

In one embodiment of the present invention, some of the multiple product machines constitute a first homogeneous machine group; other of the multiple product machines constitute a second homogeneous machine group, wherein the first homogeneous machine group and the second homogeneous machine group belong to different semiconductor processes; and the control module performs loop calculations based on production information to schedule or check the load status of different product machines in the first homogeneous machine group, and to schedule or check the load status of different product machines in the second homogeneous machine group.

In one embodiment of the present invention, the control module further determines whether there is a bottle neck machine based on the result of scheduling or checking the load status of multiple product machines.

Based on the above, the production management system can perform loop calculations based on production information. Moreover, through the corresponding loop calculations, appropriate evaluations can be performed on multiple product machines to perform corresponding dispatch or check load conditions to achieve optimal production management.

FIG1 is a system schematic diagram of a production management system according to an embodiment of the present invention.

1 , the production management system 100 may include a control module 110. The control module 110 may include corresponding hardware or software.

In one embodiment, the control module 110 includes, for example, an input unit 115, an output unit 116, a computing unit 117, and/or a storage unit 118. The input unit 115 includes, for example, a keyboard, a mouse, a touch screen, a signal receiving end (such as a corresponding data port or an antenna), and/or other similar units suitable for data input. The output unit 116 includes, for example, a screen, a printer, a signal output end (such as a corresponding data port or an antenna), and/or other similar units suitable for data output. The computing unit 117 includes, for example, a central processing unit (CPU), a graphics processing unit (GPU), a physical processing unit (PPU), or other similar units suitable for performing calculations, logical judgments, and/or data processing. The storage unit 118 includes, for example, a memory, a hard disk, a disk array, a database, and/or other similar units suitable for permanent or temporary data storage.

In terms of hardware, the control module 110 can be a single entity or a connection or integration of multiple entities. For example, the control module 110 can be an integration of multiple computers or servers.

In one embodiment, the control module 110 can be connected to the cloud system by signal. The cloud system can perform input, output, calculation, storage, monitoring, data collection, statistics and/or other appropriate operations through the control module 110 by remote control. The aforementioned cloud system includes, for example, a Fault Detection and Classification System (FDC system), a Facility Monitoring Control System (FMCS), a Computer-Integrated Manufacturing Syetem (CIM system) or other appropriate industrial control systems (ICS), but the present invention is not limited thereto.

In one embodiment, the control module 110 includes, for example, software suitable for performing logic judgment or a platform suitable for performing advanced process control (APC) and/or a programmable logic controller (PLC).

The control module 110 is suitable for signal connection to multiple product machines (which can be directly referred to as: machines). The aforementioned signal connection can generally refer to a wired signal transmission connection method, a wireless signal transmission connection method, or a combination of the above. Each product machine can transmit corresponding information (such as: corresponding parameters in the module, corresponding raw material (such as: process gas, target material, photoresist, etc.) type, etc.; but not limited) to the control module 110. The control module 110 can transmit the corresponding information to each product machine, or perform direct or indirect control. In addition, the present invention does not limit all signal connection methods to be the same or different. In addition, the aforementioned transmission may include direct transmission; or, transmission via a corresponding repeater in hardware, via corresponding encoding, decoding or other appropriate methods in software.

In one embodiment, among different product machines, a group of product machines that can perform the same process can be called homogeneous machines. That is to say, during production, homogeneous machines belonging to the same group can be used alternately according to corresponding production management. For example, machine 01, machine 02, machine 03, machine 04, and machine 05 are homogeneous machines of one kind (such as CVD machines that can be used for the same thin film process); and machine 11, machine 12, machine 13, and machine 14 are homogeneous machines of another kind (such as exposure and development machines that can be used for the same lithography process). Therefore, machine 01, machine 02, machine 03, machine 04, and machine 05 can constitute a first homogeneous machine group; and machine 11, machine 12, machine 13, and machine 14 can constitute a second homogeneous machine group.

Fig. 2 is a schematic diagram of a partial process of production management according to an embodiment of the present invention. Specifically, Fig. 2 may be a schematic diagram of a partial process of production management performed according to the production management system 100 in Fig. 1 .

Please refer to FIG2 and FIG1 , step S11 is a procedure for obtaining corresponding production information. In one embodiment, the initial production information can be created by directly or indirectly creating it manually, importing, copying or uploading information already stored in other databases, or other appropriate methods.

Production information may include, but is not limited to: information on production path, work in process (WIP), production time (TC), product information, and/or up time. Production path may be, for example, a certified (including but not limited to, certification by the customer, certification by an impartial third party) machine at a site. The aforementioned production information is described in detail below.

The production tool (product path) information may include the corresponding tool information of the semiconductor process sites and/or processes required for the production of each batch (lot) of chips and/or the corresponding substrate (such as wafer). The main semiconductor processes required for chip production include, for example, diffusion processes, thin film processes, lithography processes and/or etching processes. Diffusion processes may include but are not limited to: oxidation, annealing, and implantation. Thin film processes may include but are not limited to: physical vapor deposition (PVD) and chemical vapor deposition (CVD). Lithography processes may include but are not limited to: coating, exposing, and development. The etching process may include but is not limited to: dry etching, wet etching, ashing, polishing. Therefore, the production tool information may include the product tool usage process and/or sequence of the lot of products or substrates (such as wafers). It is worth noting that for the semiconductor process required for chip production, it can be carried out in different product tools (which can be called homogeneous tools) that can perform the same process. It is also worth noting that for the semiconductor process sites required for chip production, the same or similar processes may be repeatedly performed between the same or similar product tools. In terms of BEOL processes, corresponding thin film processes, lithography processes and/or etching processes may be repeatedly performed to form multiple layers of interconnect layers.

Work in Process (WIP) information may include information (e.g., quantity) related to raw materials (e.g., bare wafers or blanket wafers) ready for processing and/or semi-finished products (e.g., processing wafers) that have been processed. For semiconductor wafer factories, work in process can directly correspond to wafers in process (WIP). For example, work in process information may include corresponding product batches or other information that may be directly or indirectly related to product manufacturing.

Production time information can include the standard working time (TC) for product production. Standard working time can include personnel working time (Ts), production gap between products (Tc), product processing time under a single operation (Tm), product duration under a single operation (Tl), or the above-mentioned integration for estimation. The above-mentioned estimation basis can be obtained through existing log files or general production management methods (such as: standard manpower calculation, equipment capacity calculation, cost allocation calculation, etc.).

Product information (product info.) may include the number of substrates (e.g., wafers) corresponding to each batch (lot), product feature values, product prices, product delivery dates, or other information that may be directly or indirectly related to the final product.

Up Time can include Run Time, Ready Time, Engineering Time and/or the integration of the above, which is mainly considered from the perspective of the corresponding product machine. Up Time can include Regular Production time and/or rework time. Waiting Time includes time when there is no operation (No operator), no work in progress (No WIP) or other possible reasons that cause the product machine not to operate. Engineering Time can include the time required for machine testing requirements, R&D requirements or other engineering requirements (Engineering requirement).

Please refer to FIG. 2 and FIG. 1 , step S12 is a procedure for calculating and/or estimating the actual operating time per unit time of each product machine.

In one embodiment, the calculation and/or estimation of the actual operating time of each product machine per unit time includes a function composed of the number of run lots per unit time (e.g., per day), the standard working time (TC), and the up time. The above function is represented by [Formula 1] as follows: [Formula 1]

In one embodiment, considering multiple products and/or different sites or product tools involved in the overall semiconductor process, the aforementioned actual utilization time can be further calculated and/or estimated based on the corresponding manufacturable tool information, work-in-progress information and/or product information for each product tool.

Please refer to FIG. 2 and FIG. 1 , step S13 is a procedure for calculating and/or estimating the time ratio of each product machine to execute product production or corresponding process (such as Run goods).

In one embodiment, considering multiple products, different sites involved in the overall semiconductor process and/or product machines involved in the semiconductor process, the time ratio of each of the aforementioned product machines to execute product production or corresponding processes (such as: Run goods) can be calculated and/or estimated based on the corresponding producible machine information, work-in-progress information and/or product information. The aforementioned calculation and/or estimation at least includes a function composed of the number of work-in-progress (WIP number), the machine calculation capacity (Cap.) of the same type of machine, and each product machine and the up time (Up Time). The above function is represented by [Formula 2] as an example as follows: [Formula 2]

In one embodiment, under the premise of the corresponding number of products (WIP number), the support capacity (such as cross-factory support) and/or other capacity (such as R&D capacity) can be further considered to make further adjustments under the premise of the aforementioned [Formula 2]. Moreover, under the premise of the corresponding available time (Up Time), the rework rate (RWK) and/or other variation rate (UTIL) can be further considered to make further adjustments under the premise of the aforementioned [Formula 2]. Therefore, the Run time ratio of each product machine can be further calculated and/or estimated by the function represented by the example of [Formula 3]: [Formula 3]

Please refer to Figures 2 and 1. Step S14 is a procedure for calculating and/or estimating the ratio of the number of run lots (Run Lots) that can be operated per unit time for each product machine. It is worth noting that the aforementioned ratio is based on the operating time ratio (such as operating hour ratio), the standard working time (TC) of homogeneous machines and the up time (Up Time) of each product machine, and the ratio of the number of run lots per unit time (such as daily) between each is calculated and/or estimated. The standard working time (TC) can also be the time it takes to produce a batch of goods. The standard working time (TC) may also be referred to as working hours. Taking the ratio of the number of run lots that can be operated per day for each product machine as an example, it can be calculated and/or estimated by the function represented by [Formula 4]: [Formula 4]

Through the aforementioned procedures of step S11, step S12, step S13 and step S14, the first corresponding calculation and/or estimation can be substantially completed. Furthermore, the result of the Nth time (e.g., the first time) can be fed back to the production information in step S11 to perform the N+1th time (e.g., the second time) corresponding calculation and/or estimation, so as to achieve dynamic loop calculation and achieve optimal production management.

For example, the Nth result can be properly estimated to obtain the N+1th adjustment direction. For example, the deviation or outlier of a certain product machine among the homogeneous machines described can be calculated or understood by appropriate statistical methods (such as: estimation of the mean, standard deviation or coefficient of variation) as the basis for the adjustment direction. For another example, the corresponding adjustment direction can be evaluated by other mathematical methods (such as: Random walk (RW), Monte Carlo method or Newton’s method). In one embodiment, the aforementioned adjustment direction can be to adjust the Run quantity or proportion of the corresponding product machine. In addition, multiple loop calculations can reduce the differences in the Run quantity/ratio of different product machines in the same machine (i.e., the difference or difference ratio converges) to achieve optimal production management.

In one embodiment, the number of loops is not limited in the present invention, and can be tens, hundreds, or thousands of times. In addition, the limiting condition for the end of the loop can be a number limit and/or a threshold limit, which is not limited in the present invention.

Fig. 3 is a schematic diagram of a result of production management according to an embodiment of the present invention. Specifically, Fig. 3 may be a schematic diagram of a part of the result of production management performed according to the process shown in Fig. 2 by the production management system 100 in Fig. 1 .

In Figure 3, the horizontal axis represents the corresponding product machines. For example, machine 01, machine 02, machine 03, machine 04, and machine 05 are homogeneous machines (e.g., CVD machines that can be used for the same thin film process), and machine 11, machine 12, machine 13, and machine 14 are homogeneous machines (e.g., exposure and development machines that can be used for the same lithography process).

In Figure 3, the horizontal axis represents the ratio of the number of batches that can be operated per day for each product machine. It is worth noting that Figure 3 only illustrates two different homogeneous machines and/or nine different product machines, and in the practice of semiconductor manufacturing, there may be hundreds, thousands or more corresponding product machines. In addition, in terms of the way of presentation, appropriate selection and/or presentation can be performed through an appropriate graphical user interface (GUI).

In FIG3 , “Series 1” represented by a dotted line is the ratio of the number of batches that can be operated per day for each product machine when the process shown in FIG2 is performed for the Nth time (e.g., the first time), and “Series 2” represented by a solid line is the ratio of the number of batches that can be operated per day for each product machine when the process shown in FIG2 is performed for the N+1th time (e.g., the second time).

As shown in Figure 3, through multiple loop processes, the proportion of available batches and/or workload of each product machine in homogeneous machines can be made consistent, thereby improving the corresponding production capacity utilization.

FIG. 4 is a partial flow diagram of production management according to an embodiment of the present invention.

Please refer to FIG4 , step S10 is a procedure for performing a production management calculation. In one embodiment, the production management calculation can be as described in FIG2 and/or its corresponding implementation method, and will not be elaborated here.

Please refer to FIG. 4 , step S20 is a procedure for adjusting the actual corresponding scheduling method or checking the load status according to the result of the production management calculation.

Please refer to FIG4 , step S30 is a decision, which is to determine whether there is a bottle neck machine. For example, it can be determined whether a product machine is a bottle neck machine (Bottleneck Machine / Key Machine) based on the corresponding wait time (Wait time), run time or other appropriate methods.

Please refer to FIG. 4 . If the result of step S30 is that there is no bottle neck machine, the current production management process can be terminated accordingly.

Please refer to FIG. 4 . If the result of step S30 is that there is a bottle neck machine, then step S40 can be performed accordingly. Step S40 is a decision to improve the strategy. The strategy improvement method may include but is not limited to: confirmation of the bottle neck machine, improvement of the process, evaluation of the purchased product machine, etc.

In summary, the production management system of the present invention can perform loop calculations based on production information. Furthermore, through corresponding loop calculations, appropriate evaluations can be performed on multiple product machines to perform corresponding scheduling or check load conditions, thereby achieving optimal production management.

100: Production management system 110: Control module 115: Input unit 116: Output unit 117: Calculation unit 118: Storage unit S11, S12, S13, S14: Steps S10, S20, S30, S40: Steps

FIG. 1 is a system schematic diagram of a production management system according to an embodiment of the present invention. FIG. 2 is a partial process schematic diagram of a production management according to an embodiment of the present invention. FIG. 3 is a result schematic diagram of a production management according to an embodiment of the present invention. FIG. 4 is a partial process schematic diagram of a production management according to an embodiment of the present invention.

S11, S12, S13, S14: Steps

Claims (8)

A production management system, comprising: A control module, suitable for signal connection to multiple product machines, wherein: The control module performs loop calculations based on production information to schedule or check the load status of multiple product machines, wherein: The production information includes information on producible machines, work-in-progress information, production time information, product information, and available time information; and The production information is dynamically adjusted, and each loop in the loop calculation is performed based on the corresponding production information, wherein: The production information required for the first loop is the initial production information; and The control module adjusts the production information required for the N+1th loop in the loop calculation based on the result of the Nth loop in the loop calculation, wherein N is a positive integer. A production management system as described in claim 1, wherein the loop calculation includes at least a function composed of the number of operating batches in the production information, the standard working time of the product machine, and the available operating time of the product machine. A production management system as described in claim 1, wherein the loop calculation includes at least a function composed of the number of work-in-process in the work-in-process information, the machine calculated capacity of multiple product machines, and the available operating time of each of the product machines. A production management system as described in claim 1, wherein the loop calculation at least includes a function composed of the operating time ratio of each of the product machines, the standard working time of multiple product machines, and the available operating time of each of the product machines. A production management system as described in claim 1, wherein: Some of the plurality of said product machines constitute a corresponding homogeneous machine group; and The loop calculation is to at least reduce the difference in the quantity or proportion of the executed product production or corresponding process of different said product machines in the homogeneous machine group. A production management system as described in claim 1, wherein: Some of the plurality of said product machines constitute a corresponding homogeneous machine group; and The control module further simulates and adjusts the quantity or proportion of the executed product production or corresponding process of different said product machines in the homogeneous machine group according to the calculation result of the loop calculation. A production management system as described in claim 1, wherein: Some of the plurality of product machines constitute a first homogeneous machine group; Other of the plurality of product machines constitute a second homogeneous machine group, wherein the first homogeneous machine group and the second homogeneous machine group belong to different semiconductor processes; and The control module performs the loop calculation according to the production information to schedule or check the load status of different product machines in the first homogeneous machine group, and schedule or check the load status of different product machines in the second homogeneous machine group. In the production management system as described in claim 1, the control module further determines whether there is a bottle neck machine based on the results of scheduling or checking the load status of multiple product machines.