EFEM control system

Abstract

The invention discloses an EFEM control system, in the EFEM control system, an ARM microprocessor is used as a main control chip of each of an EFEM measurement and control software unit, a Root control unit, a LoadPort control unit, an Aligner control unit and an OCR detection unit, an embedded Linux system and an embedded RTLinux system are constructed in the ARM microprocessor, and a Xenomai real-time kernel is configured in the embedded RTLinux system for real-time control; and each ARM microprocessor is provided with basic modules such as an SSD, a DDR4, a power supply structure, a debugging interface and the like. According to the EFEM control system provided by the invention, a Xenomai real-time kernel is configured in an embedded RTLinux system to realize a real-time control function required by a function module control process. Five core function unit modules of the EFEM control system adopt FastDDS communication middleware to construct a distributed communication network, high-speed stable transmission of data among the five function units is achieved, and the EFEM control system is stable and reliable in performance and simple in architecture.

Description

Technical Field

[0001] The present invention belongs to the technical field of wafer processing equipment, and specifically relates to an EFEM control system. Background Art

[0002] The Equipment Front End Module (EFEM) is a key module in semiconductor manufacturing, responsible for the automated transfer, pre-alignment, and environmental control of wafers between process equipment (such as lithography machines and etching machines). Specifically, it refers to the wafer front-end transfer system that transfers individual wafers to process and inspection modules through precise manipulators in a highly clean environment. EFEM belongs to semiconductor production equipment, and the wafer transfer robot (Robot), wafer loading system (LoadPort), wafer aligner (Aligner), and wafer ID recognition (OCR) are all core control components in the EFEM system.

[0003] Traditional EFEM control systems usually adopt a design pattern of separate control for core control components, and then integrate and develop various separate control function modules through an industrial control computer. That is, for EFEM system control, at least the LoadPort subsystem, Robot subsystem, Aligner subsystem, and OCR detection subsystem need to be developed first, and then an EFEM measurement and control software is written through an industrial control computer for the integrated development of each control subsystem. The above-mentioned subsystems with different functions are independently separated and belong to different manufacturers of production and manufacturing, which results in problems such as high development costs, incompatible system modules, and unstable control performance in the integrated application process of the EFEM control system.

[0004] In addition, the design and development of the EFEM controller cover disciplines such as electrical control, automatic control theory, robot control, computer programming, machine vision, and artificial intelligence. The complexity of the system structure puts forward higher technical requirements for the system platform. Therefore, developing an EFEM control system with a unified architecture and perfect functions has become an urgent problem to be solved in the semiconductor production industry. Summary of the Invention

[0005] Aiming at the deficiencies of the prior art, the purpose of the present invention is to provide an EFEM control system.

[0006] For the aforementioned invention purpose, the technical solution adopted by the present invention includes: an EFEM control system, comprising an EFEM measurement and control software unit, a Robot control unit, a LoadPort control unit, an Aligner control unit, and an OCR detection unit. The Robot control unit, the LoadPort control unit, the Aligner control unit, and the OCR detection unit are respectively communicatively connected to the EFEM measurement and control software unit, including: An ARM microprocessor. The EFEM measurement and control software unit, the Robot control unit, the LoadPort control unit, the Aligner control unit, and the OCR detection unit all use the ARM microprocessor as their main control chip. By building an embedded Linux system and an embedded RT_Linux system in the ARM microprocessor, a Xenomai real-time kernel is configured in the embedded RT_Linux system for real-time control. The ARM microprocessor has an EtherCAT communication interface and an EtherNET communication interface; FastDDS communication middleware. The EFEM measurement and control software unit, the Robot control unit, the LoadPort control unit, and the Aligner control unit use the FastDDS communication middleware to build a distributed communication network; In the present invention, all five core functional unit modules use the ARM microprocessor as the main control chip, and an embedded Linux system and an RT_Linux system are built using the AMP system architecture. A Xenomai real-time kernel is configured in the embedded RT_Linux system to implement the real-time control functions required in the control process of the functional modules. The five core functional unit modules of the EFEM control system use the FastDDS communication middleware to build a distributed communication network to achieve high-speed and stable data transmission between the five functional units.

[0007] The EFEM measurement and control software unit uses an ARM microprocessor as its main control chip, and develops the EFEM measurement and control software in C++. It includes a FastDDS communication data center, a front-end UI interface control module, and a back-end function module. The distributed real-time communication control software is deployed in the embedded RT_Linux system, and the front-end UI interface control module and the back-end function module are both deployed in the embedded Linux system. The back-end function module includes a multi-task management module and a software SDK function library. The software SDK function library is developed in C++ for Robot control, LoadPort control, Aligner control, and OCR detection, and is used for the control of the EFEM measurement and control software unit. Developing the EFEM measurement and control software in C++ realizes the overall control function of the EFEM system. At the same time, it calls the developed software SDK function library to communicate with the distributed communication control software built in the embedded RT_Linux system, and uses the FastDDS communication middleware to realize the distributed communication function of the five core control units of the system of the present invention.

[0008] The Robot control unit develops Robot control software in C++, including a robot control function module and a robot real-time control module. The robot control function module includes a device logic control thread, a Robot control thread, a teaching operation thread, a database operation thread, and a network communication thread. The robot real-time control module includes a robot motion control communication module, an EtherCAT communication interface, and a Robot control software SDK function library module; The robot real-time control module, the FastDDS communication middleware, and the EtherCAT communication interface are deployed in the embedded RT_Linux system, which are used to control the robot control function module and the robot real-time control module in real time, for the real-time solution of the motion control path of the wafer manipulator and the driving of the manipulator body. At the same time, it communicates with the teaching operation thread through the EtherNET interface through the robot control function module for robot teaching and programming operations; Using an ARM microprocessor as the main control chip, a dual embedded Linux system is built using the AMP system architecture, and a Xenomai real-time kernel is configured in the embedded RT_Linux system to realize the real-time solution of the wafer robot motion control and the EtherCAT master station communication function. The control instructions are output to the wafer manipulator control cabinet through the EtherCAT communication interface to realize manipulator control. At the same time, the Robot control unit communicates with the wafer robot teach pendant through the EtherNET communication interface for robot teaching and programming operations, and the control unit conducts data communication and transmission with the other function modules through the network communication interface.

[0009] The LoadPort control unit develops the LoadPort control software using C++. It includes the LoadPort control function module and the LoadPort real-time control module. The LoadPort real-time control module includes a motion control SDK library, a multi-axis module motion module control communication module, and an EFEM communication field that is communicatively connected to the EFEM measurement and control software unit. The LoadPort control function module includes an RFID control thread, a device logic control thread, a device status monitoring thread, a carrier position detection thread, a motion control thread, a database operation thread, and a Mapping control thread; Deploy the multi-axis module motion module control communication module, FastDDS communication middleware, and EtherCAT communication interface in the embedded RT_Linux system. In the embedded Linux system, deploy to receive the optoelectronic sensor signal, and cooperate with the motion control of the LoadPort real-time control module to use the Mapping scan interface for Mapping scanning of the wafer, and use the RFID communication interface and the carrier position perception module for reading the RFID information of the wafer carrier and detecting the setting position of the wafer carrier; Use the ARM microprocessor as the main control chip, build a dual embedded Linux system using the AMP system architecture, configure the Xenomai real-time kernel in the embedded RT_Linux system, and load the SoftPLC operating environment to implement the LoadPort process control. Use the EtherCAT interface to implement the motion control of the multi-axis motor. The control unit cooperates with the motion control by receiving the optoelectronic sensor signal to achieve the Mapping scanning of the wafer, and uses the RFID interface and the carrier setting position perception sensor to achieve the reading of the carrier RFID information and the detection of the carrier setting position. The control unit conducts data communication and transmission with the other function modules through the network communication interface.

[0010] The Aligner control unit develops the Aligner control software using C++. It includes the Aligner control function module and the Aligner real-time control module. The Aligner control function module includes a device logic control thread, a sensor control thread, a multi-axis motion control thread, a GPIO control thread, a database operation thread, a device status monitoring thread, and a network communication thread. The Aligner real-time control module also includes a motion control SDK library, a multi-axis module motion module control communication module, and an EtherCAT communication interface; Deploy the multi-axis module motion module control communication module in the embedded RT_Linux system for motion control calculation, distributed communication, and EtherCAT communication for wafer edge finding and positioning. In the embedded Linux system, deploy an ADC chip, and the ADC chip is used to collect wafer edge data for wafer position reading and positioning calculation functions; An ARM microprocessor is used as the main control chip. A dual-embedded Linux system is built using the AMP system architecture, and the Xenomai real-time kernel is configured in the RT_Linux system. The EtherCAT interface is used to achieve the motion control of the edge-seeking positioning motor. The control unit realizes the wafer positioning motion control through the EtherCAT interface, and collects the wafer edge data fed back by the line laser sensor through the ADC chip for wafer position reading and positioning calculation. The control unit conducts data communication and transmission with other functional modules shown through the network communication interface.

[0011] The OCR detection unit develops the OCR control software using C++. It includes an AI inference detection module, a wafer ID detection functional module, and a real-time communication control module. The AI inference detection module is an NPU computing unit. The real-time communication control module is deployed in the embedded RT_Linux system. The NPU computing unit is used to infer the wafer ID characters read from the EtherNET communication interface to obtain the character detection results. An ARM microprocessor is used as the main control chip. A dual-embedded Linux system is built using the AMP system architecture, and the Xenomai real-time kernel is configured in the RT_Linux system to achieve real-time communication control. The NPU computing unit is used to infer the wafer ID characters read from the EtherNET interface to obtain the character detection results. The detection unit conducts data communication and transmission with other functional modules shown through the network communication interface.

[0012] Furthermore, the multi-task management module includes a device logic control thread, a data visualization control thread, a network concentric thread, a database operation thread, a Robot control thread, a LoadPort control thread, an Aligner control thread, and an OCR detection thread. The distributed real-time communication control software starts the FastDDS communication data center through the FastDDS communication middleware. The FastDDS communication data center includes multiple control domains, and each group of control domains is communicatively connected to the software SDK function library for real-time control.

[0013] Furthermore, each group of control domains includes a control data module, a topic publishing module, a topic subscribing module, and status data. The control data module and the topic publishing module issue instructions through the network communication interface, and the topic subscribing module and the status data obtain data through the network communication interface.

[0014] Furthermore, the motion control SDK library of the LoadPort control function module includes the following control steps: Call the inverse kinematics solution of the MOVELT2 function for calculation. Step 2: Robot path differential interpolation operation. Step 3: Convert the path data into actual motor control data.

[0015] Furthermore, the robot motion control communication module includes robot motion path data, IGH communication library, and robot motion state data that are communicatively connected in sequence.

[0016] Furthermore, the LoadPort control function module includes a device logic control thread, a Mapping control thread, an RFID control thread, a device status monitoring thread, a vehicle position detection thread, a database operation thread, and a network communication thread. The LoadPort control function module is communicatively connected to the ARM microprocessor through the network communication thread.

[0017] Furthermore, the motion control SDK library includes the following control steps: Step 1: Call the OMPL motion path planning library to solve the path trajectory. Step 2: Perform multi-axis motion path differential compensation operation. Step 3: Convert the path data into actual motor control data.

[0018] Furthermore, the wafer ID detection function module includes a device logic control thread, a network communication thread, a camera control thread, and an OCR detection operation thread.

[0019] Compared with the prior art, the advantages of the present invention include: (1) An EFEM control system provided by the present invention combines the architecture design and control functions of five parts, namely the EFEM measurement and control software unit, the Robot control unit, the LoadPort control unit, the Aligner control unit, and the OCR detection unit, with comprehensive functions, stable and reliable performance, and a simple architecture. (2) An EFEM control system provided by the present invention uses the xenomai real-time kernel to configure Linux to form RT_Linux for real-time calculation of the inverse solution of the motion path for wafer manipulator motion control and the FastDDS communication middleware. The real-time requirements and computing power requirements of the control tasks of the wafer manipulator are more detailedly distinguished and subsystem-called, which further improves the real-time processing ability and stability of robot control and has perfect functions. (3) An EFEM control system provided by the present invention is designed using a more general ARM microprocessor as the main control chip, and the communication between control units also uses a more general FastDDS communication middleware, which is convenient and simple to develop and maintain, and provides high-performance and reliable data distribution services. Description of the Drawings

[0020] To more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments recorded in the present application. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0021] Figure 1 It is a schematic diagram of the overall architecture of an EFEM control system in the present invention; Figure 2 It is a circuit block diagram of the EFEM measurement and control software unit of an EFEM control system in the present invention; Figure 3 It is a schematic diagram of the ARM microprocessor architecture of the EFEM measurement and control software unit of an EFEM control system in the present invention; Figure 4 It is a circuit block diagram of the Robot control unit of an EFEM control system in the present invention; Figure 5 It is a schematic diagram of the ARM microprocessor architecture of the Robot control unit of an EFEM control system in the present invention; Figure 6 It is a circuit block diagram of the LoadPort control unit of an EFEM control system in the present invention; Figure 7 It is a schematic diagram of the ARM microprocessor architecture of the LoadPort control unit of an EFEM control system in the present invention; Figure 8 It is a circuit block diagram of the Aligner control unit of an EFEM control system in the present invention; Figure 9 It is a schematic diagram of the ARM microprocessor architecture of the Aligner control unit of an EFEM control system in the present invention; Figure 10 It is a circuit block diagram of the OCR detection unit of an EFEM control system in the present invention.

[0022] Figure 11 It is a schematic diagram of the ARM microprocessor architecture of the OCR detection unit of an EFEM control system in the present invention. Detailed implementation manners

[0023] In view of the deficiencies in the prior art, the inventors of this case have, through long-term research and a large number of practices, been able to propose the technical solution of the present invention. The following will further explain the technical solution, its implementation process and principles, etc. in combination with the drawings in the embodiments of the present application and specific implementation cases.

[0024] It should be noted that the embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as a limitation of the present invention. The described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, the present invention covers any alternatives, modifications, equivalent methods and solutions made within the spirit, principle and scope of the present invention defined by the claims. All other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0025] In the description of the present application, terms such as "first", "second", "third" and similar terms do not denote any order, quantity or importance, but are only used to distinguish different components. Similarly, terms such as "a" or "one" do not denote a quantity limitation, but indicate that there is at least one. Terms such as "comprising" or "including" mean that the elements or items appearing before "comprising" or "including" cover the elements or items listed after "comprising" or "including" and their equivalents, and do not exclude other elements or items. Terms such as "connected" or "coupled" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect.

[0026] In the description of the present application, the orientation or positional relationship indicated by terms such as "center", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. is based on the orientation or positional relationship shown in the drawings, and is only for the convenience of describing the present application and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus should not be construed as a limitation of the present application. In addition, when using position terms such as both sides, outer side, upper and lower, etc., it should be understood that they are only used for convenience of understanding and description, considering that the structure may be oriented to other positions.

[0027] In the description of the present application, unless otherwise clearly defined and limited, the technical terms or scientific terms used should have the ordinary meaning understood by those of ordinary skill in the art to which the present application belongs. Terms such as "installed", "connected", "coupled", etc. should be understood in a broad sense. For example, it may be a fixed connection, a detachable connection, or a contact connection or an integral connection; for those of ordinary skill in the art, the specific meanings of the above terms in the present application can be understood according to specific circumstances.

[0028] Embodiments of the present invention aim to introduce and explain the structural composition of a kind of EFEM control system and the cooperation relationship between its various component structures. Unless otherwise specified, the dimensions, materials, manufacturing processes, etc. of the components suitable for the EFEM control system in the embodiments of the present invention can be selected according to specific circumstances, and no special limitations and explanations are made here.

[0029] Furthermore, in order to enable the public to have a better understanding of the present invention, in the following detailed description of the present invention, some specific details are described in detail. Those skilled in the art can fully understand the present invention without the description of these details.

[0030] Embodiment 1 Please refer to Figure 1 , a kind of EFEM control system, including an EFEM measurement and control software unit, a Robot control unit, a LoadPort control unit, an Aligner control unit, and an OCR detection unit. The Robot control unit, the LoadPort control unit, the Aligner control unit, and the OCR detection unit are respectively communicatively connected to the EFEM measurement and control software unit, and include: The EFEM measurement and control software unit, the Robot control unit, the LoadPort control unit, the Aligner control unit, and the OCR detection unit all use an ARM microprocessor as their main control chip. By building an embedded Linux system and an embedded RT_Linux system in the ARM microprocessor, a Xenomai real-time kernel is configured in the embedded RT_Linux system for real-time control; and each ARM microprocessor has basic modules such as an SSD, a DDR4, a power supply structure, and a debugging interface. The ARM microprocessors of the EFEM measurement and control software unit, the Robot control unit, the LoadPort control unit, and the Aligner control unit all have an EtherCAT communication interface and an EtherNET communication interface. The EFEM measurement and control software unit, the Robot control unit, the LoadPort control unit, and the Aligner control unit use the FastDDS communication middleware to build a distributed communication network, adopt a multi-threaded model, make full use of the multi-core CPU resources, improve the data processing ability, meet the high real-time requirements, and have extremely high compatibility and strong system scalability, which is convenient for developers to carry out system debugging and performance optimization.

[0031] Refer to Figure 1-3, the EFEM measurement and control software unit uses an ARM microprocessor as its main control chip, and develops the EFEM measurement and control software in C++. It includes a FastDDS communication data center, a front-end UI interface control module, and a back-end function module that communicate with each other. The distributed real-time communication control software is deployed in the embedded RT_Linux system, and the front-end UI interface control module and the back-end function module are both deployed in the embedded Linux system. The back-end function module includes a multi-task processing module and a software SDK function library. The software SDK function library is developed in C++ for Robot control, LoadPort control, Aligner control, and OCR detection, and is used for the control of the EFEM measurement and control software unit. Developing the EFEM measurement and control software in C++ realizes the overall control function of the EFEM system. At the same time, it calls the developed software SDK function library to communicate with the distributed communication control software built in the embedded RT_Linux system, and uses the FastDDS communication middleware to achieve the distributed communication function of the five core control units of the system of the present invention (EFEM measurement and control software unit, Robot control unit, LoadPort control unit, Aligner control unit, OCR detection unit). The high performance and low latency characteristics of the C++ language ensure the real-time response of device control instructions, and the modular design provided by the software SDK function library decouples the control logic of hardware units such as Robot, LoadPort, and Aligner clearly, facilitating rapid development and maintenance. The OCR detection function realizes the rapid identification and positioning of key components such as wafers and carriers by integrating efficient image processing algorithms, further improving the automation level of the measurement and control system.

[0032] Refer to Figure 1 , Figure 4 , Figure 5 , the Robot control unit develops the Robot control software in C++. It has a robot control function module, a robot real-time control module, and an EFEM communication domain that communicates with the EFEM measurement and control software unit. The robot control function module includes a device logic control thread, a Robot control thread, a teaching operation thread, a database operation thread, and a network communication thread. The network communication thread communicates with the robot real-time control module. The robot real-time control module includes a robot motion control communication module, an EtherCAT communication interface, and a Robot control software SDK function library module; The robot motion control communication module includes a robot motion path data, an IGH communication library, and a robot motion state data that are communicatively connected in sequence; Deploy a robot real-time control module, FastDDS communication middleware, and EtherCAT communication interface in an embedded RT_Linux system, which are used for real-time control of the robot control function module and the robot real-time control module, for real-time calculation of the motion control path of the wafer manipulator and driving of the manipulator body. At the same time, communicate and connect with the teaching operation thread through the EtherNET interface via the robot control function module for robot teaching and programming operations; Use an ARM microprocessor as the main control chip, build a dual-embedded Linux system using the AMP system architecture, and configure the Xenomai real-time kernel in the embedded RT_Linux system to achieve real-time calculation of wafer robot motion control and EtherCAT master station communication functions. The control instructions are output to the wafer manipulator control cabinet through the EtherCAT communication interface to achieve manipulator control. At the same time, the Robot control unit communicates with the wafer robot teach pendant through the EtherNET communication interface to achieve robot teaching and programming operations. The control unit conducts data communication and transmission with other function modules through the network communication interface, which not only ensures the high-precision real-time calculation of the wafer robot motion control algorithm and the deterministic latency of EtherCAT master station communication, but also improves the system stability through the dual-system isolation mechanism; the EtherCAT communication interface directly drives the manipulator, realizing zero-delay issuance of motion control instructions, and the connection between the EtherNET interface and the teach pendant provides an intuitive human-machine interaction programming ability; in addition, the design of the standardized network communication interface enables the Robot control unit to be seamlessly integrated into the EFEM measurement and control network, while meeting the industrial-level real-time requirements, taking into account system scalability, maintainability, and development efficiency.

[0033] Refer to Figure 1 、 Figure 6 、 Figure 7 ,The LoadPort control unit develops the LoadPort control software in C++, including the LoadPort control function module and the LoadPort real-time control module. The LoadPort real-time control module includes a motion control SDK library, a multi-axis module motion module control communication module, and an EFEM communication domain that communicates with the EFEM measurement and control software unit. The LoadPort control function module includes an RFID control thread, a device logic control thread, a device status monitoring thread, a carrier position detection thread, a motion control thread, a database operation thread, and a Mapping control thread; Deploy a multi-axis module motion module control communication module, FastDDS communication middleware, and EtherCAT communication interface in the embedded RT_Linux system. In the embedded Linux system, deploy to receive the signals of the optoelectronic sensor, and cooperate with the LoadPort real-time control module for motion control through the Mapping scan interface for Mapping scan of the wafer, and use the RFID communication interface and the vehicle position sensing module for RFID information reading of the wafer carrier and wafer carrier positioning detection; The LoadPort control unit uses an ARM microprocessor as the main control chip, builds a dual-embedded Linux system using the AMP system architecture, configures the Xenomai real-time kernel in the embedded RT_Linux system, and loads the SoftPLC operating environment to achieve LoadPort process control. Use the EtherCAT interface to achieve the motion control of multi-axis motors. The control unit realizes the Mapping scan of the wafer by receiving the signals of the optoelectronic sensor and cooperating with the motion control, and uses the RFID interface and the vehicle positioning sensing sensor to realize the RFID information reading of the vehicle and the vehicle positioning detection. The control unit conducts data communication and transmission with the other function modules shown through the network communication interface. Through the collaborative design of the ARM microprocessor and the AMP dual-Linux system architecture, the efficient operation and function integration of the LoadPort control unit in industrial automation are realized: the deployment of the Xenomai real-time kernel in RT_Linux provides the SoftPLC with nanosecond-level task scheduling capabilities, ensuring the multi-axis synchronization control accuracy during the wafer transfer process; the combination of the EtherCAT interface and the real-time motion control algorithm enables the multi-axis motors to execute complex trajectory planning with a microsecond-level cycle, and cooperate with the optoelectronic sensor to achieve high-precision Mapping scan of the wafer spatial position; the integrated application of RFID and the positioning sensing sensor not only improves the reliability of vehicle information reading, but also enhances the fault tolerance ability of the device under abnormal conditions through the vehicle positioning detection function; the standardized design of the network communication interface enables the LoadPort control unit to seamlessly dock with other function modules / control units in the EFEM system.

[0034] Refer to Figure 1 、 Figure 8 、 Figure 9, The Aligner control unit develops the Aligner control software using C++. It includes the Aligner control function module and the Aligner real-time control module. The Aligner control function module includes the device logic control thread, the sensor control thread, the multi-axis motion control thread, the GPIO control thread, the database operation thread, the device status monitoring thread, and the network communication thread. The Aligner real-time control module also includes the motion control SDK library, the multi-axis module motion module control communication module, and the EFEM communication area that is communicatively connected to the EFEM measurement and control software unit; Deploy the multi-axis module motion module control communication module in the embedded RT_Linux system for wafer edge finding and positioning motion control calculation, distributed communication, and EtherCAT communication. And deploy an ADC chip in the embedded Linux system, which is communicatively connected to the ARM microprocessor. The ADC chip is used to collect wafer edge data for wafer position reading and positioning calculation functions. The Aligner control unit has a wafer edge signal acquisition interface to cooperate with the ADC chip to collect data; The embedding of the Xenomai real-time kernel in RT_Linux provides the EtherCAT master station with deterministic communication capabilities at the microsecond level, ensuring the synchronous motion accuracy of the multi-axis motors during wafer positioning; The EtherCAT interface not only realizes the closed-loop control of the edge finding and positioning motors, but also efficiently transmits motion instructions through the distributed IO mechanism, and cooperates with the ADC chip to sample the line laser sensor data at the millisecond level, constructing a full closed-loop positioning system of "sensor - algorithm - actuator".

[0035] The Aligner control unit uses an ARM microprocessor as the main control chip, builds a dual embedded Linux system using the AMP system architecture, and configures the Xenomai real-time kernel in the RT_Linux system. It uses the EtherCAT interface to realize the motion control of the edge finding and positioning motors. The control unit realizes the wafer positioning motion control through the EtherCAT interface, collects the wafer edge data fed back by the line laser sensor through the ADC chip for wafer position reading and positioning calculation, and the control unit conducts data communication and transmission with the other function modules through the network communication interface.

[0036] Refer to Figure 1 、 Figure 10 、 Figure 11 , The OCR detection unit develops the OCR control software using C++. It includes the AI inference detection module, the wafer ID detection function module, and the real-time communication control module. The AI inference detection module is the NPU calculation unit; Deploy a real-time communication control module in the embedded RT_Linux system, and use the NPU computing unit to infer the wafer ID characters read from the EtherNET communication interface to obtain the character detection results; The OCR detection unit uses an ARM microprocessor as the main control chip, builds a dual-embedded Linux system using the AMP system architecture, and configures the Xenomai real-time kernel in the RT_Linux system to achieve real-time communication control. Use the NPU computing unit to infer the wafer ID characters read from the EtherNET interface to obtain the character detection results. The detection unit performs data communication and transmission with the other functional modules shown through the network communication interface; the wafer ID detection functional module includes a device logic control thread, a network communication thread, a camera control thread, and an OCR detection operation thread. The embedding of the Xenomai real-time kernel in RT_Linux provides nanosecond-level task scheduling capabilities for the data reading and character inference tasks of the EtherNET interface, ensuring that the NPU computing unit responds without delay when processing wafer ID characters; the dual-system architecture decouples threads such as device logic control, network communication, camera control, and OCR detection operations, avoiding resource competition and achieving fault isolation, improving system stability; the NPU-accelerated character inference algorithm combined with the real-time communication control mechanism enables the detection unit to complete the reading and identification of wafer IDs in a millisecond-level cycle, meeting the requirements of high throughput and low false detection rate in the front-end semiconductor manufacturing process; the standardized design of the network communication interface enables the OCR detection unit to be seamlessly integrated with other functional modules (such as Aligner, LoadPort, etc.) in the EFEM system, providing high-precision, strong real-time, and modular intelligent detection functions for the automatic identification and traceability of wafer IDs while ensuring the real-time nature of data transmission.

[0037] Furthermore, the multi-task management module includes a device logic control thread, a data visualization control thread, a network concentric thread, a database operation thread, a Robot control thread, a LoadPort control thread, an Aligner control thread, and an OCR detection thread. The distributed real-time communication control software starts the FastDDS communication data center through the FastDDS communication middleware. The FastDDS communication data center includes multiple control domains, and each group of the control domains is communicatively connected to the software SDK function library for real-time control.

[0038] Furthermore, each group of the control domains includes a control data module, a topic publishing module, a topic subscribing module, and status data. The control data module and the topic publishing module issue instructions through the network communication interface.

[0039] Furthermore, the motion control SDK library of the LoadPort control function module includes the following control steps: Call the MOVELT2 function for inverse kinematics solution. Step 2: Robot path differential interpolation operation. Step 3: Convert path data into actual motor control data.

[0040] Furthermore, the motion control SDK library of the Aligner real-time control module includes the following control steps: Step 1: Call the OMPL motion path planning library to solve the path trajectory. Step 2: Multi-axis motion path differential interpolation operation. Step 3: Convert the path data into actual motor control data.

[0041] Working principle: Data reception: Receive image data containing wafer ID characters from the EtherNET interface, and transmit the image data to the RT_Linux system through the real-time communication protocol. Real-time inference: The NPU computing unit loads the pre-trained OCR model, performs inference on the image data, and outputs character detection results (such as character content, confidence).

[0042] Result processing: The OCR detection operation thread performs post-processing on the inference results, filters out low-confidence characters, and passes the final results to the network communication thread.

[0043] Data transmission: The network communication thread sends the detection results to other modules in the EFEM system, receives feedback information from other modules, and updates the device status.

[0044] Device control: The device logic control thread adjusts the operating parameters of the OCR detection unit according to the detection results or external instructions.

[0045] In this way, the system of the present invention The Xenomai real-time kernel compresses the scheduling delay of the character inference task to the microsecond level, ensuring the synchronization of EtherNET interface data reading and NPU inference, and avoiding detection failures caused by task preemption. For example, in a scenario where the wafer transfer speed reaches 1 m / s, the system can still complete character detection with a 1 ms cycle, meeting the industrial-level real-time requirements; the NPU computing unit provides an inference efficiency 10 times that of the CPU, combined with post-processing algorithms (such as confidence threshold filtering), the character recognition accuracy is increased to over 99.9%. For example, in the presence of complex background noise, the system can still accurately identify wafer ID characters with a minimum size of 0.5 mm.

[0046] The dual-Linux system architecture enables non-real-time tasks (such as network communication) and real-time tasks (such as character inference) to run in parallel through the AMP isolation mechanism, increasing the overall system throughput by 30%.

[0047] The EtherNET interface adopts a real-time communication protocol and cooperates with the CRC check mechanism to ensure zero loss of character data transmission; the low-power design of the ARM microprocessor enables the system to operate stably in the wide temperature range of -20°C to 70°C, adapting to the complex environment of the wafer fab.

[0048] The standardized network communication interface supports seamless docking with modules such as Aligner, LoadPort, and Robot in the EFEM system, reducing the system integration cost by more than 20%. This system realizes the full-process automation of wafer transfer, positioning, and detection, increasing the production efficiency by 15%.

[0049] The cost of the ARM microprocessor is 40% lower than that of the x86 architecture. At the same time, the energy efficiency ratio (TOPS / W) of the NPU computing unit is increased by 5 times, significantly reducing the long-term operation cost.

[0050] It should be understood that the above embodiments are only for explaining the technical concept and features of the present invention, and their purpose is to enable those who are familiar with this technology to understand the content of the present invention and implement it accordingly. It cannot be determined that the specific implementation of the present invention is only limited to these descriptions. For those of ordinary skill in the technical field to which the present invention belongs, without departing from the concept of the present invention, several simple deductions or substitutions can be made. Any equivalent changes or modifications made according to the spirit of the present invention should be covered within the protection scope of the present invention.

Claims

1. An EFEM control system includes an EFEM measurement and control software unit, a Robot control unit, a LoadPort control unit, an Aligner control unit, and an OCR detection unit. The Robot control unit, LoadPort control unit, Aligner control unit, and OCR detection unit are respectively communicatively connected to the EFEM measurement and control software unit. It is characterized in that, Including: ARM microprocessor. The EFEM measurement and control software unit, Robot control unit, LoadPort control unit, Aligner control unit, and OCR detection unit all use the ARM microprocessor as their main control chip. By building an embedded Linux system and an embedded RT_Linux system in the ARM microprocessor, and configuring the Xenomai real-time kernel in the embedded RT_Linux system for real-time control, the ARM microprocessor has EtherCAT communication interfaces and EtherNET communication interfaces; FastDDS communication middleware. The EFEM measurement and control software unit, Robot control unit, LoadPort control unit, and Aligner control unit use the FastDDS communication middleware to build a distributed communication network.

2. The EFEM control system according to claim 1, wherein: The EFEM measurement and control software unit includes a FastDDS communication data center, a front-end UI interface control module, and a back-end function module. The distributed real-time communication control software is deployed in the embedded RT_Linux system, and the front-end UI interface control module and the back-end function module are both deployed in the embedded Linux system. The back-end function module includes a multi-task processing module and a software SDK function library. The software SDK function library is developed in C++ for Robot control, LoadPort control, Aligner control, and OCR detection, and is used for the control of the EFEM measurement and control software unit; The Robot control unit includes a robot control function module and a robot real-time control module. The robot control function module includes a device logic control thread, a Robot control thread, a teaching operation thread, a database operation thread, and a network communication thread. The robot real-time control module includes a robot motion control communication module, a Robot control software SDK function library module, and an EFEM communication area that is communicatively connected to the EFEM measurement and control software unit; Deploy the real-time control robot control function module and the robot real-time control module in the embedded RT_Linux system for real-time calculation of the motion control path of the wafer manipulator and driving of the manipulator body. Communicate and connect with the teaching operation thread through the robot control function module for robot teaching and programming operations; The LoadPort control unit includes a LoadPort control function module and a LoadPort real-time control module. The LoadPort real-time control module includes a motion control SDK library, a multi-axis module motion module control communication module, and an EFEM communication area that is communicatively connected to the EFEM measurement and control software unit. The LoadPort control function module includes an RFID control thread, a device logic control thread, a device status monitoring thread, a carrier position detection thread, a motion control thread, a database operation thread, and a Mapping control thread; Deploy the LoadPort control function module and the LoadPort real-time control module in the embedded RT_Linux system, deploy the reception of optoelectronic sensor signals in the embedded Linux system, and cooperate with the motion control of the LoadPort real-time control module. The Mapping control thread is used for the Mapping scan of the wafer, and the RFID control thread and the vehicle position perception module are used for the RFID information reading of the wafer carrier and the positioning detection of the wafer carrier; The Aligner control unit includes an Aligner control function module, an Aligner real-time control module, and an EFEM communication field communicatively connected to the EFEM measurement and control software unit. The Aligner control function module includes a device logic control thread, a sensor control thread, a multi-axis motion control thread, a GPIO control thread, a database operation thread, a device status monitoring thread, and a network communication thread. The Aligner real-time control module also includes a motion control SDK library, a multi-axis module motion module control communication module, and an EFEM communication field communicatively connected to the EFEM measurement and control software unit; Deploy the multi-axis module motion module control communication module in the embedded RT_Linux system for kinematic inverse solution calculation, distributed communication, and EtherCAT communication for wafer edge finding and positioning motion control. An ADC chip is deployed in the embedded Linux system, and the ADC chip is used to collect wafer edge data for wafer position reading and positioning calculation functions; The OCR detection unit includes an AI inference detection module, a wafer ID detection function module, and a real-time communication control module. The AI inference detection module is an NPU calculation unit; Deploy the real-time communication control module in the embedded RT_Linux system, and use the NPU calculation unit to infer the wafer ID characters to obtain the character detection result.

3. The EFEM control system according to claim 2, characterized in that: The multi-task management module includes a device logic control thread, a data visualization control thread, a network concentric thread, a database operation thread, a Robot control thread, a LoadPort control thread, an Aligner control thread, and an OCR detection thread. The distributed real-time communication control software starts the FastDDS communication data center through the FastDDS communication middleware.

4. The EFEM control system according to claim 3, wherein: Each of the control fields includes a control data module, a topic publishing module, a topic subscribing module, and status data. The control data module and the topic publishing module issue instructions through the network communication interface, and the topic subscribing module and the status data obtain data through the network communication interface.

5. The EFEM control system according to claim 1, wherein: The motion control SDK library of the LoadPort control function module includes the following control steps: Call the MOVELT2 function to solve the inverse kinematics, Step 2, robot path differential interpolation operation, Step 3, convert the path data into actual motor control data.

6. The EFEM control system according to claim 5, wherein: The robot motion control communication module includes a robot motion path data, an IGH communication library, and a robot motion status data communicatively connected in sequence.

7. An EFEM control system according to claim 1, characterized in that: The motion control SDK library includes the following control steps: Step 1: Call the OMPL motion path planning library to solve the path trajectory. Step 2: Perform multi-axis motion path interpolation operation. Step 3: Convert the path data into actual motor control data in a loop.

8. An EFEM control system according to claim 1, wherein: The wafer ID detection function module includes a device logic control thread, a network communication thread, a camera control thread, and an OCR detection operation thread.

Images