JP4849773B2 - Automatic production system - Google Patents

Description

  The present invention relates generally to the field of semiconductor device manufacturing, and more particularly to an automated production environment that uses autonomous, active software agents to control recipes for wafer processing.

  In the semiconductor industry, there is a continuous motivation for improving the quality, reliability and throughput of integrated circuit devices such as microprocessors, memory devices and the like. This motivation is driven by consumer demand for high quality computers and electrical products that operate with higher reliability. Such a requirement leads to continuous improvement in the manufacture of semiconductor devices such as transistors, and further to continuous improvement in the manufacture of integrated circuit devices including such transistors. Furthermore, by reducing defects in the manufacture of typical transistor components, the overall cost per transistor can be reduced, and further the cost of integrated circuit devices including such transistors can be reduced.

  In general, a set of processing steps is performed on a wafer using various processing (process) apparatuses. The processing apparatus includes a photolithography stepper, an etching apparatus, a deposition apparatus, a polishing apparatus, a rapid thermal processing apparatus, an implantation apparatus, and the like. The technology underlying semiconductor processing (process) equipment has generated greater interest over the past few years, which has led to substantial improvements. However, despite the progress made in this area, many of the currently commercially available processing equipment suffers from several drawbacks. In particular, such devices often do not have advanced process data monitoring capabilities. Advanced process data monitoring functions, for example, in a user-friendly format, with real-time graphical display of both event logs, current process parameters and process parameters for the entire run, as well as progress ( The ability to provide historical (parametric) data. These drawbacks may not optimize the processing of critical process parameters such as throughput, accuracy, stability and reproducibility, process temperature, machine equipment parameters. This variability appears as run-to-run variability, run-to-run variability, and device-to-device variability, which can lead to product quality and performance variability. In contrast, an ideal monitoring and diagnostic system for such devices provides a means to optimally control important parameters as well as a means to monitor this variability.

  One technique for improving the operation of semiconductor processing lines is to employ a factory-wide control system that automatically controls the operation of various processing equipment. A manufacturing device communicates with a network of manufacturing frameworks or processing modules. Each manufacturing device is typically coupled to a device interface. The device interface is coupled to a machine interface that facilitates communication between the manufacturing device and the manufacturing framework. The machine interface may generally be part of an advanced process control (APC) system. The APC system invokes a control script based on a manufacturing model. The control script can be a software program that automatically reads data necessary to execute the manufacturing process. Often, semiconductor devices pass through multiple manufacturing equipment for multiple processes and generate data regarding the quality of the processed semiconductor devices.

During the manufacturing process, various events can occur that affect the performance of the device being manufactured. That is, variations during the manufacturing process steps lead to variations in device performance. Factors such as critical dimensions of the structure, doping levels, contact resistance, particle contamination can all potentially affect the final performance of the device.
In order to reduce process variation, various devices in the process line are controlled according to a performance model. Commonly controlled processing equipment includes photolithography steppers, polishing equipment, etching equipment, and deposition equipment. Measurement data before and / or after processing is supplied to a process control device (controller) for the apparatus. The process control device calculates operation recipe parameters such as processing time based on the performance model and measurement information so that the result after processing is as close as possible to the target value. By reducing the deviation (variation) in this way, a result equivalent to an increase in profitability such as an increase in throughput, a reduction in cost, and an improvement in device performance is produced.

  In distributed processing environments such as factory-scale APC systems, configuration control and efficiency issues prevail. In general, there are many software developers who have written control code for configuring a process controller. Certain developers make great efforts to develop specific types of controllers. Usually, each developer has a unique programming style and relies on routines they have already developed. For example, each developer has a set of routines that interface with databases and other components within the APC framework and perform various computational and basic utility operations.

One problem with such a configuration is that there is not much consistency between process control scripts. Many custom scripts also create configuration control issues and efficiency issues. Developers may have spent considerable time replicating code that has already been developed for another type of process controller that another developer has already created. Debugging non-standardized code takes more time and further reduces efficiency.
The present invention is directed to overcoming one or more of the problems set forth above, or at least reducing the effects thereof.

Summary of the Invention

  One aspect of the present invention is a manufacturing system including a processing device, a controller agent, and a first processing agent. The processing device is configured to process a workpiece (workpiece) according to an operating recipe. The controller agent is configured to determine a control action related to processing of the workpiece in the processing device. The first processing agent is associated with at least one of the processing device and the processed product, forms an interface to the controller agent, receives a control operation, and creates an operation recipe based on the control operation. It is configured.

  Another aspect of the invention is a method for controlling a processing device configured to process a workpiece according to an operational recipe. A first processing agent associated with at least one of the processing device and the workpiece is created. A controller agent is created that is configured to determine control actions relating to processing of the workpiece in the processing device. The first processing agent is connected to the controller agent via an interface and receives a control operation. The first processing agent is connected to the processing device through an interface, and creates an operation recipe based on the control operation.

The present invention may be understood by reference to the following description in conjunction with the accompanying drawings. In the drawings, like reference numbers indicate like elements.
While the invention is susceptible to various modifications and alternative forms, specific embodiments thereof are shown by way of example in the drawings and will herein be described in detail. However, the description herein of a particular embodiment is not intended to limit the invention to the particular form disclosed, but rather to the invention as defined in the appended claims. It should be understood that it is intended to include all variations, equivalents, and alternatives that fall within the spirit and scope of this.

  Exemplary embodiments of the present invention are described below. For clarity, this specification does not describe all features of a real implementation. In the development of such real-world implementations, there are a number of practices that must be followed in order to achieve specific developer goals that vary from case to case, such as system-related compliance and business constraints. Of course, you will understand that you have to make these decisions. Further, it can be understood that such development efforts may be complex and time consuming but nevertheless are routine tasks for those skilled in the art who have benefited from the disclosure herein. Let's go.

  First, FIG. 1 is a conceptual diagram illustrating a portion of a processing flow 100 configured and executed in accordance with a specific embodiment of the present invention. The depicted portion of the process flow 100 includes two stations 105, each station 105 including a computer device 110 that communicates with a processing device 115. The two stations 105 communicate with each other over a communication link 120. In the illustrated embodiment, computing device 110 and communication link 120 include a portion of a larger computer system, such as network 125. The processing apparatus 115 shown in FIG. 1 processes a lot 130 of wafers 135 for which an integrated circuit device is being manufactured. Although the present invention has been shown as being usable to manufacture semiconductor devices, its application is not limited thereto and is applicable to other types of manufacturing processes. Accordingly, in the processing flow 100 described above, the lot 130 of the wafer 135 can be more generally referred to as a “processed product”. The processing apparatus 115 and all processing operations performed thereon need not necessarily be related to the manufacture of semiconductor devices in all embodiments. However, for the sake of clarity and for a better understanding of the invention, terms relating to semiconductor manufacturing are used when the invention is disclosed within the framework of the illustrated embodiments.

  FIG. 2 illustrates selected portions of the hardware and software architecture of computing device 110 programmed and operating in accordance with the present invention. Some elements of the hardware and software architecture (eg, individual cards, basic input / output system (BIOS), input / output drivers, etc.) are not shown. These elements are omitted for clarity, but do not obscure the present invention. However, one of ordinary skill in the art having the benefit of this disclosure will appreciate that the software and hardware architecture of computing device 110 includes many such common features.

  In the illustrated embodiment, computing device 110 is a workstation employing a UNIX-based operating system, although the invention is not so limited. The computing device 110 can be implemented in virtually any type of electronic computing device, such as a laptop computer, desktop computer, minicomputer, mainframe computer, or supercomputer. The computing device 110 may be an alternative embodiment, a processor or controller embedded in the processing device 115. The present invention is also not limited to UNIX-based operating systems. Alternative operating systems (eg, WINDOWS-based or disk operating system (DOS) -based) can also be employed. The invention is not limited by the specific implementation of computing device 110.

  The computing device 110 further includes a processor 205 that communicates with a storage device 210 that is on the bus system 215. The storage device 210 generally includes at least a hard disk (not shown) and a random access memory (RAM) (not shown). In some embodiments, the computer device 110 further includes a removable storage device such as an optical disk (not shown), a floppy electro-magnetic disk (not shown), or a magnetic tape (not shown) or a ZIP disk (not shown). And other forms of removable storage devices. The processor 205 may be any suitable processor known in the art. For example, the processor may be a general purpose microprocessor or a digital signal processor (DSP). In the illustrated embodiment, the processor 205 is an Athlon ™ 64-bit processor commercially available from Advanced Micro Devices Inc. (AMD), but the invention is not so limited. 64-bit UltraSPARC (TM) or 32-bit microSPARC (TM) from Sun Microsystems, Itanium (TM) or Pentium (TM) class processors from Intel Corporation, Compaq An Alpha ™ processor provided by Computer Corporation may be employed instead. The computing device 110 includes a monitor 240, a keyboard 245, and a mouse 250, which together constitute a user interface 260 with user interface software 255 associated therewith. The user interface 260 in the illustrated embodiment is a graphical user interface (GUI), but this is not necessary for the practice of the present invention.

  Selected portions of the software architecture of computing device 110 are further illustrated in FIG. In the illustrated embodiment, each computing device 110 has a software agent 265 resident in the storage device 210. Note that software agent 265 may reside in process flow 100 at a location other than computer device 110. The location of the software agent 265 is not essential in the practice of the present invention. The location of the software agent 265 is not essential, so some computer devices 110 have multiple software agents 265 residing on them, while other computer devices 110 have nothing. I want you to be more careful. An automated manufacturing execution system (MES) 270 such as WORKSTREAM ™ provided by Consilium, Inc. of Mountain View, Calif. Resides in at least one computing device 110.

  As described above, in FIG. 1, the computer device 110 can be part of a larger computer system 125 by being connected over the communication link 120. Examples of computer systems in such an implementation include a local area network (LAN), a wide area network (WAN), a system area network (SNA), an intranet, or even the Internet. The computer system 125 employs a networked client / server architecture, although alternative embodiments may employ a peer-to-peer architecture. Thus, in an alternative embodiment, computing devices 110 can communicate directly with each other. Communication link 120 may be wireless, coaxial cable, optical fiber, or twisted pair. In embodiments that employ it, the computer system 125 and communication link 120 will vary depending on the implementation and may be provided in any suitable manner known in the art. The computer system 125 may employ any appropriate communication protocol known in the art, for example, TCP / IP (Transmission Control Protocol / Internet Protocol).

  With reference to FIGS. 1 and 2, software agents 265 are jointly responsible for efficiently monitoring lots 130 of wafers 135 throughout the manufacturing process. Each processing device 115 is one of the resources that can be employed for this purpose. For example, the processing apparatus 115 may be a manufacturing apparatus used for manufacturing a part of the wafer 135, for example, for the layer, pattern, doping, heat treatment and the like of the wafer 135. Alternatively, the processing device 115 may be a measurement device used to evaluate the capabilities of various parts of the processing flow 100. Accordingly, software agent 265 assesses a plurality of resources for subsequent processing of lots 130 of wafers 135, allocates resources represented by processing unit 115, and subsequently processes lots 130 of wafers 135. The allocation of these resources can be negotiated with each other.

  In the illustrated embodiment, the software agents 265 are self-configured at startup (startup), intelligent, state-aware, and specific to which they begin to operate autonomously to achieve that The goal is soaked. The software agent 265 further performs self-adjustment to these environmental changes. Although software agent 265 is implemented as an object in an object-oriented programming (OOP) environment, the present invention can also be implemented using techniques that are not object-oriented. Their operation is relatively simple and script or rule based. Its actions include selected goals, such as achieving the assigned lot deadline, achieving a predetermined quality level, maximizing equipment availability, and conveniently proactively scheduling maintenance, etc. Designed to achieve. To accomplish these objectives, software agent 265 interfaces with MES 270 and integrates with an existing factory control system (not shown). As will be apparent to those skilled in the art having the benefit of this disclosure, the manner in which this interface and integration is performed varies from implementation to implementation, depending on the characteristics of the MES 270 and the factory control system.

  As described further below, software agent 265 can be specialized to several different levels. One level represents “type”, ie, whether the software agent 265 is “consumer” or “supplier” in the processing flow 100. More specifically, whether the software agent 265 is a consumer or a supplier is determined by the type of element it represents in the overall processing flow 100. For example, software agent 265 may represent lot 130 of wafer 135 (ie, “lot agent”), may represent processing device 115 (ie, “device agent”), or represent process material. (Ie “resource agents”). The software agent 265 can also be specialized by function, ie what function the software agent 265 performs in the processing flow. For example, the software agent 265 need not necessarily represent a particular element and is configured to perform a particular function. For example, the software agent 265 can be configured to determine a control action (ie, a “control agent”) for a particular lot 130 of wafers 135 that are processed in the processing apparatus 115. The operations performed by this control agent vary depending on both the characteristics of the lot 130 and the characteristics of the processing device 115. The control agent is called by the lot agent or by the device agent, depending on the specific implementation. In the illustrated embodiment, the control agent maintains a relationship with a particular processing device 115 throughout its life cycle. However, other forms can be used. Other functions that can be performed by the software agent 265 include scheduling.

  One embodiment described more fully below is, for example, a scheduling agent, a processing agent, and a controller agent. It should be noted that the software agent 265 does not necessarily have a one-to-one relationship with manufacturing domain elements such as the lot 130 and the processing device 115. Instead, most domain elements are each represented by a group of software agents 265. For example, a lot 130 or processing device 115 can have both a scheduling agent and a processing agent. A controller agent is associated with a particular processing device 115 and invoked by one of the processing agents (ie, associated with lot 130 or processing device 115). This configuration facilitates the design of specific objects that exhibit specific actions that support one aspect of domain element functionality.

  In this particular embodiment, software agent 265 is implemented using object-oriented programming techniques. In the term object-oriented computer, a software “agent” is an autonomous, active object. Given that set of actions, the software agent 265 can operate independently in response to local conditions, thereby generating an adaptive system action. The present invention defines an autonomous and portable "software agent" that mimics and improves the functionality of a "real world" agent in a semiconductor manufacturing plant such as factory workers, materials, equipment, processes, etc. Provide an agent extension system to configure and deploy. One skilled in the art will appreciate that an agent or other software object can include one or more software objects. In this specification, the term “object” should be understood that a software object may be composed of a plurality of other software objects. Conversely, one skilled in the art will understand that the function of one object can be combined with another function. It will be appreciated that functions described as being associated with separate objects can be combined as functions associated with a single object.

  Further, in this particular embodiment, software agent 265 is “setting changeable”. In this particular embodiment, software agents can change the settings by modifying the scripts that govern their operation or by modifying a “control knob”. The control knob is an externally configurable parameter or software agent property that provides flexibility and / or system adjustment. For example, certain characteristics specify curves that govern system judgment. More specifically, the processing and scheduling operations of software agent 265 are scripted such that certain predetermined events cause script segments to execute. In one particular implementation, the script is implemented as JPG, an open source Java programming language that provides a bi-directional Java interface. More information, including information about downloading Yahoo, can be obtained from the Internet at http://www.jpython.org/.

In some portions of the detailed description of the invention, the result is described as a software-implemented process that includes symbolic representations of operations on data bits in the memory of a computer system or computing device. These descriptions and representations are intended to enable those skilled in the art to efficiently communicate their work to others skilled in the art. These processes and operations require physical manipulations of physical quantities. Usually, though not necessarily, these physical quantities take the form of electrical, magnetic, or optical signals that can be stored, transferred, combined, compared, and otherwise manipulated. For reasons of general usage only, it is sometimes convenient to refer to these signals as bits, values, elements, symbols, characters, terms, numbers, etc.
It should be borne in mind, however, that all of these terms or similar terms are to be associated with the appropriate physical quantities and are merely convenient labels applied to those quantities. Unless otherwise noted, or as is clear from previous discussions, these descriptions represent data expressed as physical (electrical, magnetic, or optical) quantities in the storage of an electronic device. It is about the operation or processing of an electronic device that manipulates or converts to other data that is also represented as a physical quantity in a storage device, transfer device or display device. Examples of terms that mean such explanation include, but are not limited to, terms such as “process”, “computing”, “calculation”, “decision”, “display”, etc. Absent.

  The parts implemented by the software of the invention are usually encoded in the form of a program storage medium or implemented in the form of a transmission medium. The program storage medium may be a magnetic medium (eg, floppy disk or hard disk) or an optical medium (eg, compact disk read-only memory, or “CD-ROM”). May be. Similarly, the transmission medium may be a twisted pair, coaxial cable, optical fiber, or other suitable transmission medium known in the art. The present invention is not limited to these mounting formats.

  The processing flow 100 can be implemented according to an APC (Advanced Process Control) framework. The APC system includes a distributed software system consisting of interchangeable, standardized software components that enables run-to-run control and fault detection / classification in the process flow 300. The software agent 265 and the operation of the present invention, which is the upper layer of the APC system, realizes a semiconductor manufacturing process flow that operates almost autonomously.

  The software component of the APC system implements an architecture standard based on a system technology and an APC (Advanced Process Control) framework based on the SEMI (Semiconductor Equipment and Materials International) CIM (Computer Integrated Manufacturing) framework. CIM (provisional specification for SEMI E81-0699-CIM framework domain architecture) and APC (provisional specification for SEMI E93-0999-CIM framework advanced process control components) specifications are generally available from SEMI It is. SEMI contact information is Austin, Texas, 78731, Suite 290, Texas Highway North, Capital 8310, Telephone 512-349-2422, Fax 512-349-2442, Internet address: http://www.semi.org.

  This particular architecture relies heavily on software that uses object-oriented programming and employs the Object Management Group (OMG) CORBA (Common Object Request Broker Architecture) and CORBA service specifications for distributed object systems. Information about OMG's CORBA architecture and its specifications are also generally readily available.

  An example of a software system that can be adapted to perform the functions of the APC system described herein is the Catalyst system sold by KLA-Tencor, Inc. Contact KLA-Tencor is 95134 San Jose, California, Rio Robles 160, telephone 408-875-6000, fax 408-875-3030, http://www.kla-tencor.com. Additional information about the Catalyst system is available from the KLA-Tencor website, http://www.kla-tencor.com/contolsolutions/catalystapc.html.

  FIG. 3 is a simplified block diagram of an example embodiment of a processing flow 300 implemented using a software agent 265, according to another embodiment of the invention. An exemplary lot 130 is represented by a lot schedule creation agent 310 and a lot processing agent 320. The example processing device 115 is represented by a device schedule creation agent 330 and a device processing agent 340. The lot schedule creation agent 310 negotiates with various elements in the processing flow 300, for example, an apparatus schedule creation agent 330 for securing resources necessary for completing a lot. For example, the lot schedule creation agent 310 and the device schedule creation agent 330 may use a “contract net negotiation protocol” approach to create a reservation schedule for the processing device 115 for the lot 130.

  Lot processing agent 320 and / or device processing agent 340 is invoked when processing device 115 receives lot 130 for processing. Lot processing agent 320 and device processing agent 340 are interfaced with controller agent 350 to determine the control operations required to set the operation recipe of processing device 115 for processing lot 130. Either the lot processing agent 320 or the device processing agent 340 calls the controller agent 350. In one embodiment, the lot processing agent 320 sends to the controller agent 350 data relating to the characteristics of the lot 130 necessary to determine the required control action (eg, measurement data relating to dimensions, electrical measurement data, etc.). Can do. Similarly, device processing agent 350 can pass data regarding processing device 115 (eg, idle time, time since previous cleaning, device health, etc.) to controller agent 360. In an alternative embodiment, data regarding the characteristics of lot 130 and / or processing device 115 is stored in a centralized data store 360 accessible to controller agent 350. The data storage device 360 can be configured to store data including pre-processing and post-processing measurement data, device status, lot priority, and the like.

  The operation of the controller / agent 350 will be described in more detail. The specific control operations performed by controller agent 350 will vary depending on the implementation. For example, if the processing device 115 is a CMP device, the controller agent 350 receives a thickness measurement before polishing (eg, high structure thickness, low structure thickness) and targets after polishing. Predict the polishing time required to achieve the desired thickness. If the processing device 115 is an etching device, the controller agent 350 can model the etching performance of the processing device 115 based on thickness measurements before and / or after etching. The controller / agent 350 can make a prediction using the control model of the processing device 115. The control model is created empirically using commonly known linear or non-linear techniques. This control model can be a relatively simple equation-based model (eg linear, exponential, weighted average, etc.) or more than a neural network model, PCA (Principal component analysis) model or PLS (projection to latent structure) model It can be a complex model. The specific implementation of this model will vary depending on the modeling technique chosen. Using the control model, the controller agent 350 can determine operating recipe parameters such as etching time, plasma intensity, temperature, pressure, and reactive gas concentration to reduce post-etch thickness variations. Other control scenarios hold for other types of processing devices 115.

  When configuring the controller agent 350 (ie when it is first instantiated), global configuration variables utilized by the controller agent 350 are determined. In the illustrated embodiment, these global variables are variables and contexts from a recipe management system (RMS) (ie, a global database of recipe settings implemented by the processor 115 that stores default recipes for the current process). Contains the value of a context variable. The context variable value defines a control thread and is generally composed of variable values such as a device identification code, a lot number, and an operation number. In addition, any required baseline variable is a given value. Examples include an email list for error notification, a timeout time value, the maximum number of wafers that can be handled as a “child” lot, and the previous layer information used by the controller agent 350 ( Feed forward information).

  The controller agent 350 uses the lot number and wafer number values set in the global configuration variables to determine the lot number, family name, parent name, factory, wafer number and status (that is, the lot is the parent Whether it is a lot or a child lot). The previously determined context and RMS information can be used to set values that the controller agent 350 uses to calculate the control movement. For example, the controller agent 350 can use context information (or “thread” designation) to set the value of the control model parameter according to the value defined in the RMS. As a specific example, setting a value for the etch rate used in the control model according to the context of a particular etch chamber and a value for the etch rate for that etch chamber as defined in the RMS is there.

  Further, the controller agent 350 can retrieve feedforward measurement information from the data storage device 360 using a queue (query) by lot number and layer name. The controller agent 350 checks the elements in the data array for a given lot and fills in missing values with default values. For example, the previous process target value is used to set the missing measurements required as part of the feedforward information used by the controller agent 350. Instead of using default values, other methods of setting missing feedforward information can be implemented.

  The controller agent 350 sets the key and state structure values necessary to retrieve the data store 360 to retrieve the control state associated with the current control thread. This key is used to read thread state data from the data store 360. Controller agent 350 retrieves thread state data in the requested stack of data for the latest lot processed in this thread context. If such values are found, they are passed to the user-defined function that contains the control model. The user-defined function performs the calculation and returns a value for that thread state. When a value is not found in the stack, the controller agent 350 searches up the hierarchical structure (hierarchy) and retrieves data from the first hierarchical level that has a value for that thread state. The stack and all hierarchy levels are similar but are assumed to contain data with different accuracy.

  The controller agent 350 performs a jeopardy check by performing a match in the data store 360 to create a dangerous stack (ie, a stack of lots processed in the thread given from the last measurement operation). Read the value for the number of lots. This value is compared to a threshold for the number of lots in this risk category, usually the value specified in the RMS. If this risk threshold is not exceeded, the controller agent 350 continues to operate. If this threshold is exceeded, the controller agent 350 will abort and abort the metering event for a lot from the list of lots in the hazard stack before determining the control action for the current lot. Sends a pop-up display that instructs to execute

  The controller agent 350 calculates the controller input (process recipe update) from the previously determined state and target information as described above. The controller agent 350 then sends the result of the control operation to the device processing agent 340, thereby updating the operation recipe of the processing device 115 before processing the lot 130.

  In some embodiments, two or more controller agents 350 are created for processing lot 130. For example, if processing unit 115 is a photolithography stepper, one controller agent 350 is called for overlay control and the other controller agent 350 is called to control the critical dimension (CD). The plurality of controller agents 350 can use feedback from processed wafers to adjust various stepper parameters such as exposure dose, exposure time, focus, and the like. In another example, a deposition apparatus, such as an apparatus for forming a polysilicon layer, also includes a plurality of controllers to control parameters such as polysilicon grain size (grain size) and polysilicon layer thickness. Can have an agent 350.

  When the controller agent 350 is invoked, it checks the context variables to determine which individual controller agent 350 needs to determine the control action. The operation ID indicates the process to be performed (eg, poly gate mask vs. second interlayer dielectric mask (ILD)). Each controller agent 350 is only needed in certain context states and will only operate when all those context conditions are met. For example, the CD controller agent 350 operates for a poly gate mask, but not the second ILD mask process. On the other hand, the overlay controller agent 350 can operate on both mask events.

  As described herein, the use of an agent-based control architecture improves overall plant efficiency and provides more comprehensive automation. More particularly, the autonomous, active software agent 265 schedules and initiates lot scheduling and processing execution. In general, improved process control reduces product variability, thereby increasing production capacity and profitability.

  The particular embodiments disclosed so far are only examples. As will be apparent to those skilled in the art having the benefit of the teachings herein, the present invention can be modified and implemented in different but equivalent ways. Furthermore, there is no intention to limit the invention to the details of construction or design disclosed herein, except as described in the appended claims. Accordingly, the particular embodiments described above can be modified or altered and such variations are considered to be within the scope and spirit of the invention. Accordingly, the scope of protection sought is as set forth in the appended claims.

The conceptual diagram which shows a part of one Embodiment of the 1st processing flow comprised and operated according to this invention. FIG. 2 is a conceptual partial block diagram illustrating selected portions of the hardware and software architecture of the computer apparatus of FIG. FIG. 9 is a conceptual partial block diagram of specialized agents for scheduling agents, processing agents, and controller agents in a second processing flow configured and operated in accordance with the present invention.

Claims (8)

A processing device (115) configured to process the workpiece (130) according to the operation recipe;
Receiving processed product data obtained by measuring the processed product (130), and relating to processing of the processed product (130) in the processing device (115) based on a control model and the received processed product data A controller agent (350) configured to determine a control action to be performed;
In connection with at least one of the processing device (115) and the work piece (130), interfaced to the controller agent (350) to receive the control action, and an action recipe based on the control action A production system (100, 300) comprising a first processing agent (320, 340) configured to change at least one of the parameters.   A second processing agent (320, 340) related to at least one of the processing device (115) and the processed product (130) and not related to the first processing agent (320, 340); The production system (100) of claim 1, wherein at least one of the one processing agent (320, 340) and the second processing agent (320, 340) is configured to interface with the controller agent (350). 300).   The first processing agent (320) is associated with the workpiece (130), is interfaced with the controller agent (350), receives the control action, and is associated with the processing device (115). Configured to pass the control action to a second processing agent (340), wherein the second processing agent (340) is configured to create the action recipe based on the control action; The production system (100, 300) according to claim 2. The first processing agent (320) is configured to supply the processed product data to the controller agent (350), and the second processing agent (340) receives processing device data related to the processing device (115). Configured to supply a controller agent (350), wherein the controller agent (350) is configured to determine the control action based on the control model, the workpiece data, and the processing device data. The production system (100, 300) according to claim 2, wherein: A method for controlling a processing device (115) configured to process a workpiece (130) according to an operating recipe comprising:
Creating a first processing agent (320, 340) associated with at least one of the processing device (115) and the workpiece (130);
Receiving processed product data obtained by measuring the processed product (130), and relating to processing of the processed product (130) in the processing device (115) based on a control model and the received processed product data Creating a controller agent (350) configured to determine a control action to perform;
Connecting the first processing agent (320, 340) to the controller agent (350) through an interface and receiving the control operation;
Connecting the first processing agent (320, 340) and the processing device (115) with an interface and changing at least one parameter of the operation recipe based on the control operation. Creating a second processing agent (320, 340) related to at least one of the processing device (115) and the processed product (130) but not related to the first processing agent (320, 340); ,
6. The method of claim 5, further comprising: interfacing at least one of the first processing agent (320, 340) and the second processing agent (320, 340) to the controller agent (350). Passing the control action from the first processing agent (320) to the second processing agent (340);
7. The method of claim 6, further comprising: connecting the second processing agent (340) and the processing device (115) with an interface to create the operation recipe based on the control operation. Connecting the first processing agent (320) and the controller agent (350) with an interface to provide the workpiece data;
Interfacing the second processing agent (340) to the controller agent (350) to provide processing device data for the processing device (115);
The method of claim 6, further comprising determining the control action based on the control model, the workpiece data, and the processing device data .

Images