Scanning electron microscope high-throughput full-process automated testing system and method

Abstract

This invention provides a high-throughput, fully automated testing system and method for scanning electron microscopy (SEM), belonging to the field of material microstructure characterization and automated detection. The task management and scheduling module receives and parses preset test plans, generates test task sequences, and issues them sequentially. The SEM control system receives basic operation commands and controls the SEM to complete vacuum breaking, vacuuming, and sample testing. The robotic arm control system receives robotic arm movement commands and controls the robotic arm. A QR code camera captures QR codes at multiple candidate points. The robotic arm control system calculates the actual position of the support base center in the SEM. The robotic arm control system controls the gripper assembly to pick up the multi-station sample stage and move it above the support base. The gripper assembly then lowers the multi-station sample stage onto the support base. This forms an automated process from automatic sample loading, vacuuming, testing to data retrieval, significantly improving testing throughput.

Description

Technical Field

[0001] This invention relates to the field of material microstructure characterization and automated testing technology, and in particular to a high-throughput fully automated testing system and method for scanning electron microscopy. Background Technology

[0002] Scanning electron microscopy (SEM), with its high resolution, large depth of field, and flexible detection modes, is a core tool for characterizing the microstructure, structure, and composition of materials science, bioscience, and nanotechnology. However, traditional SEM testing processes are highly dependent on manual operation, typically testing only one or a few samples per batch. Furthermore, operators must repeatedly perform repetitive operations such as loading and unloading samples, venting and evacuating the SEM chamber, setting parameters, and acquiring images. This is time-consuming, labor-intensive, and introduces human error, severely limiting testing efficiency. Therefore, there is an urgent need to develop an automated testing system and method that can be deeply integrated with SEM to achieve automatic sample transfer and end-to-end task scheduling.

[0003] In the prior art, Chinese invention patent application CN117949678A, entitled "A Sample Batch Detection Device and Detection Method," discloses a sample batch detection device, including a sample transport component, a sample transfer component, a scanning electron microscope (SEM) device, and a vacuum component. The sample transport component includes a transport pipe, a sample platform, and a platform moving mechanism. The sample platform has multiple sample placement stations. The platform moving mechanism is located inside the transport pipe, and its moving end is connected to the sample platform. The platform moving mechanism is adapted to drive the sample platform to move along the extension direction of the transport pipe within the transport pipe. The sample transfer component includes a transfer pipe and a transfer mechanism. The transfer pipe is connected to the transport pipe, and the transfer mechanism is located inside the transfer pipe. The sample inlet of the SEM device is connected to the transport pipe, and the transfer mechanism is adapted to transfer the samples placed at the sample placement stations to the sample detection stage at the sample inlet. The vacuum pump's suction end is connected to the transport pipe. The control component is electrically connected to the sample transport component, the sample transfer component, the SEM device, and the vacuum component. This sample batch testing device uses transport pipelines, transfer pipelines, and overall pre-vacuuming, which avoids repeatedly opening and closing the scanning electron microscope chamber door, but it has the following inherent drawbacks: (1) A transport pipeline and a transfer pipeline need to be designed for the scanning electron microscope equipment. Since the sample inlet of the scanning electron microscope equipment needs to be connected to the transport pipeline, the scanning electron microscope equipment itself needs to have an interface or be subject to destructive modification, which is not applicable to most commercial scanning electron microscope equipment.

[0004] (2) In the sample batch detection device, the transfer mechanism includes a robotic arm. The end of the robotic arm can extend and retract along the extension direction of the transfer pipe, thereby transferring the sample on the sample platform moved here to the scanning electron microscope. It can be seen that the robotic arm transfers a single sample to the sample detection stage. The accuracy depends on the absolute positioning accuracy of the robotic arm. However, after long-term operation, the repeated positioning error of the robotic arm will accumulate, reducing the sample placement accuracy. Furthermore, long-term operation requires manual calibration. Summary of the Invention

[0005] The technical problem to be solved by this invention is: how to solve the problems of low throughput, insufficient automation, and lack of overall process control in the existing technology of scanning electron microscope testing.

[0006] This invention solves the above-mentioned technical problems through the following technical solution: a high-throughput fully automated testing system for scanning electron microscopy, comprising: Multi-station sample stage, including Individual presentation The sample-bearing sites are evenly distributed in an array for simultaneous loading. One sample to be tested; The intelligent scientist platform includes a task management and scheduling module. The task management and scheduling module is used to receive and parse the preset test plan, generate test task sequences, and issue them in sequence. The test task sequences include basic operation instructions and robotic arm movement instructions. The SEM control system is used to receive basic operation instructions and control the scanning electron microscope to complete vacuum breaking, vacuuming and sample testing. After the scanning electron microscope completes a basic operation instruction, the SEM control system will feed back the execution result or status information to the intelligent scientist platform. The robotic arm control system receives and controls the robotic arm's movement commands. The end effector of the robotic arm is equipped with a QR code camera and a gripper assembly. The gripper assembly is adapted to the interface on the scanning electron microscope (SEM) and is used to open or close the SEM's hatch. The QR code camera captures QR codes on the hatch at multiple candidate points and sends them to the robotic arm control system. The robotic arm control system calculates the actual position of the center of the support base within the SEM. The robotic arm control system then controls the gripper assembly to pick up the multi-station sample stage and move it above the support base. The gripper assembly then lowers the multi-station sample stage onto the support base. Each time the robotic arm completes a movement command, the robotic arm control system feeds back the execution status to the intelligent scientist platform.

[0007] Beneficial Effects: This invention uses an intelligent scientist platform as the intelligent control hub, employing an asynchronous communication architecture of issuing commands and providing message feedback. Through a unified task description format, global task queue scheduling, and a command-feedback-stepping mechanism, it achieves seamless integration between the scanning electron microscope and the robotic arm, forming an integrated automated process from automatic sample loading, vacuuming, testing to data retrieval, fully meeting the integration needs of high-throughput unmanned laboratories. A customized multi-station sample stage enables continuous testing after a single loading. This significantly increases the testing throughput of scanning electron microscopes by 100 samples.

[0008] Preferably, the preset test plan includes the sample number, sample composition information, SEM test parameters, and test sequence corresponding to each sample carrier position in the multi-station sample stage. According to the preset test plan, the task management and scheduling module converts the test objectives of each sample carrier position into basic operation units in sequence according to the preset disassembly rules. The basic operation units include moving the stage, setting parameters, automatically adjusting brightness / contrast, automatically eliminating astigmatism, automatically focusing, and automatically acquiring images. The basic operation units corresponding to all sample carrier positions form a task queue. The robot arm opens the door, grabs the sample stage, closes the door, evacuates and breaks the vacuum, and removes the sample stage, respectively, at the beginning and end of the task queue. The basic operation instructions include the task queue, evacuation and vacuum breaking instructions, and the robot arm movement instructions include the robot arm opens the door, grabs the sample stage, closes the door, and removes the sample stage.

[0009] Beneficial effects: The mechanical motion commands and SEM parameter control share the same task sequence and are uniformly scheduled by the same upper-level platform. For example, after the door closing action is completed, the intelligent scientist platform immediately sends a vacuuming command to the scanning electron microscope; after the test is completed, the intelligent scientist platform immediately sends an opening command to the robotic arm control system, forming true full-process automation.

[0010] Preferably, the SEM control system is equipped with an SEM automation gateway service. The intelligent scientist platform communicates with both the SEM automation gateway service and the robotic arm control system via HTTP protocol. Basic operation instructions issued by the intelligent scientist platform are transmitted to the SEM automation gateway service in the form of HTTP requests. The SEM automation gateway service converts the HTTP requests into AutoScript function calls. The SEM control system controls the scanning electron microscope to complete vacuum breaking, vacuuming, and sample testing based on the function call results. The execution results or status information of the scanning electron microscope are fed back to the intelligent scientist platform via Kafka messages, and the execution status of the robotic arm is fed back to the intelligent scientist platform via the Kafka messaging mechanism.

[0011] Preferably, the intelligent scientist platform also includes: The vacuum status monitoring module acquires vacuum level data within the scanning electron microscope (SEM) chamber before the door is opened. The task management and scheduling module determines whether the chamber pressure has reached atmospheric pressure based on this data. If so, the module sends a door-opening command to the robotic arm control system; otherwise, it sends a vacuum-breaking command to the SEM control system until atmospheric pressure is reached. After the door is closed, the module checks the door locking sensor signal. Once confirmed, it sends a vacuum-evacuation command to the SEM control system, which then controls the SEM to perform vacuum evacuation. The vacuum status monitoring module receives real-time vacuum level data within the SEM chamber. When the vacuum level data is less than or equal to the vacuum threshold, the vacuum is considered adequate. The module then feeds back the result or status information to the intelligent scientist platform, and the task management and scheduling module sends a start-test signal.

[0012] Preferably, the intelligent scientist platform also includes a data retrieval module. The task management and scheduling module issues instructions from the basic operation units one by one. During the execution of each basic operation unit, the stage is moved to the coordinates of the current sample to be tested, and parameters are set, brightness / contrast is automatically adjusted, astigmatism is automatically corrected, focus is automatically adjusted, and images are automatically acquired in sequence. After image acquisition, the image and parameter data of the current sample to be tested are transmitted to the data retrieval module. The data retrieval module automatically associates the corresponding sample information according to the sample station number. After the test of the current sample to be tested is completed, the stage is moved to the coordinates of the next sample to be tested, until the test of all samples to be tested is completed.

[0013] Preferably, the interface includes a cylindrical head, a connecting cylindrical section, and an intermediate constricted shaft section located between the cylindrical head and the connecting cylindrical section. The intermediate constricted shaft section is adapted to the gripper assembly, and the end face of the connecting cylindrical section is provided with a U-shaped fork arm, which is engaged with the door handle of the scanning electron microscope.

[0014] Beneficial effects: This invention requires no hardware modification to the scanning electron microscope. By setting an interactive interface at the door of the scanning electron microscope, and adapting the interactive interface to the gripper assembly of the robotic arm, the robotic arm can not only perform sample stage loading and unloading operations, but also independently perform the automatic opening and closing of the scanning electron microscope chamber. This achieves fully automated control of the external physical interface of the scanning electron microscope without human intervention.

[0015] This invention also provides a high-throughput fully automated testing method for scanning electron microscopy, employing a testing system, characterized in that the method includes: S1. After the pressure inside the scanning electron microscope chamber reaches atmospheric pressure, the robotic arm control system controls the gripper assembly to open the scanning electron microscope chamber door. S2. The task management and scheduling module sends a sample stage grabbing command. The QR code camera captures the QR code on the hatch at multiple candidate points and sends it to the robotic arm control system. The robotic arm control system calculates the actual position of the center of the support base in the scanning electron microscope. The robotic arm control system controls the gripper assembly to grab the multi-station sample stage and move it above the support base in the scanning electron microscope. The gripper assembly slowly lowers the multi-station sample stage onto the support base. S3. The task management and scheduling module sends a door-closing command. The robotic arm control system controls the gripper assembly to close the door. The task management and scheduling module sends a vacuuming command. When the vacuum level in the chamber is less than or equal to the vacuum threshold, the task management and scheduling module sends test commands to the SEM control system in the order of the test task sequence. The SEM control system controls the scanning electron microscope to complete the test commands for each sample bearing position in sequence until all sample bearing positions are tested. The task management and scheduling module sends a vacuum-breaking command. After the pressure in the chamber reaches atmospheric pressure, the robotic arm control system controls the gripper assembly to open the door and remove the multi-station sample stage.

[0016] Preferably, the process of the robotic arm control system calculating the actual position of the center of the support base in the scanning electron microscope in step S2 includes: The robotic arm control system calculates the transformation matrix of the QR code relative to the center of the support base from each viewpoint. The final pose of the QR code is obtained through data fusion; By combining the pre-calibrated QR code with the fixed conversion relationship between the center of the multi-station sample stage support base, the actual position of the support base center is calculated.

[0017] Preferably, in step S3, the task management and scheduling module sends test instructions to the SEM control system according to the test task sequence, and the SEM control system controls the scanning electron microscope to complete the test instructions for each sample bearing position in sequence, including: The task management and scheduling module issues instructions from each basic operation unit sequentially. During the execution of each basic operation unit, the stage is moved to the coordinates of the current sample under test, and parameters are set, brightness / contrast is automatically adjusted, astigmatism is automatically corrected, focus is automatically adjusted, and images are automatically acquired. After image acquisition, the image and parameter data of the current sample under test are transmitted to the data retrieval module. The data retrieval module automatically associates the corresponding sample information according to the sample station number. After completing the test of the current sample, the stage is moved to the coordinates of the next sample under test, until all samples under test are tested.

[0018] Preferably, the method further includes step S4: after completing the test of a multi-station sample stage, the robotic arm control system controls the gripper assembly to move the completed multi-station sample stage to the unloading area, the robotic arm control system controls the gripper assembly to pick up the next multi-station sample stage to be tested, and completes the test according to steps S2 to S3 until all multi-station sample stages are tested.

[0019] The advantages provided by this invention are: 1. The coordinates of each bearing position of the multi-station sample stage are manually pre-calibrated and entered into the system; during the test, the data recovery module automatically associates the sample information, test parameters and timestamp according to the current test station number and stores them in the database to ensure the consistency and traceability of the test data of each station sample.

[0020] 2. The system architecture of this invention has good modularity. The robotic arm and intelligent scientist platform can be adapted to different brands and models of scanning electron microscope equipment, and have high versatility and promotional value.

[0021] 3. In existing technologies, a robotic arm transfers a single sample to the sample testing stage. The accuracy depends on the absolute positioning accuracy of the robotic arm. However, after long-term operation, the repetitive positioning error of the robotic arm accumulates, reducing the sample placement accuracy. Furthermore, manual calibration is required for long-term operation. This invention obtains the center position of the support base through multi-view repositioning, using real-time measurements. The placement accuracy of the sample stage depends on the accuracy of visual measurement and the effectiveness of the data fusion algorithm, rather than the absolute positioning accuracy of the robotic arm. Therefore, the accumulation of repetitive positioning errors of the robotic arm during long-term operation will not affect the final placement accuracy. This invention automatically completes benchmark repositioning before each operation, eliminating the need for periodic shutdowns for manual calibration, significantly improving equipment utilization and experimental efficiency. Multi-view data fusion effectively suppresses random noise from single measurements, ensuring high accuracy and high repeatability of the positioning results. Attached Figure Description

[0022] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0023] Figure 1 This is a schematic diagram of the high-throughput fully automated testing system for scanning electron microscopes provided in Embodiment 1 of the present invention; Figure 2 This is a perspective view of the multi-station sample stage in the high-throughput fully automated testing system for scanning electron microscopy provided in Embodiment 1 of the present invention. Figure 3 This is a top view of the multi-station sample stage in the high-throughput fully automated testing system for scanning electron microscopy provided in Embodiment 1 of the present invention; Figure 4This is a schematic diagram of the interactive interface in the high-throughput fully automated testing system for scanning electron microscopes provided in Embodiment 1 of the present invention; Figure 5 This is a schematic diagram of the gripper assembly in the high-throughput fully automated testing system for scanning electron microscopes provided in Embodiment 1 of the present invention; Figure 6 This is a flowchart illustrating the operation of the high-throughput fully automated testing system for scanning electron microscopes provided in Embodiment 1 of the present invention. Figure 7 The raw SEM image of the sample obtained by the high-throughput fully automated testing system for scanning electron microscopy provided in Embodiment 1 of the present invention; Figure 8 The raw SEM image of the sample obtained by the high-throughput fully automated testing system for scanning electron microscopy provided in Embodiment 1 of the present invention; In the figure: 1 Intelligent Scientist Platform, 2 SEM Control System, 3 Scanning Electron Microscope, 4 Interactive Interface, 41 Cylindrical Head, 42 Intermediate Necked Shaft Section, 43 Connecting Cylindrical Section, 44 U-shaped Fork Arm, 5 Robotic Arm, 51 Robotic Arm End Effector, 511 QR Code Camera, 512 Gripper Assembly, 6 Multi-station Sample Stage.

[0024] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below in conjunction with specific embodiments and with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0026] Example 1 like Figure 1 As shown, this embodiment provides a high-throughput fully automated testing system for scanning electron microscopy, including: an intelligent scientist platform 1, a SEM control system 2, a robotic arm 5, and a multi-station sample stage 6. The multi-station sample stage 6 includes... Individual presentation The sample-bearing sites are evenly distributed in an array for simultaneous loading. One sample to be tested; Indicates a line, To represent columns, see [link / reference]. Figure 2 and Figure 3 In this embodiment, , The multi-station sample stage 6 is pre-set with 20 standardized sample-bearing positions, evenly distributed in a 4x5 grid pattern. Each position has a unique coded identifier and can simultaneously hold up to 20 samples to be tested. The 20-position sample stage has a rectangular structure, approximately 60 mm × 44 mm in size. Its bearing surface is divided into a 4x5 grid structure by evenly distributed horizontal and vertical dividing lines, forming 20 identical and independent square sample-bearing slots. The edges of the sample stage are marked with scale markings corresponding to the numbers, including start number markings for each row on the left edge and end number markings for each row on the right edge, to achieve positioning and identification of each bearing position.

[0027] It should be noted that in the multi-station sample stage 6 The number and arrangement of sample carrier positions can be adjusted as needed. Typically, the overall dimensions (length × width) of the multi-station sample stage 6 are fixed; for example, the length of the multi-station sample stage 6 is... Width is To ensure that all sample bearing positions are evenly distributed, i.e., the size of all sample bearing positions is... The sample support area is square. Let be the side length of the sample bearing position. Each sample bearing site is When the array is uniformly distributed, the number of rows Number of columns The value can be: , The first row of the multi-station sample stage 6 is marked with the starting number 1 and the ending number 1. The starting number of the second row is marked as The end number of the second line is marked as And so on, the first The starting number of the row is marked as , No. The end number of the line is marked as .

[0028] The intelligent scientist platform 1 includes a task management and scheduling module, a vacuum state monitoring module, and a data collection module. The task management and scheduling module is used to receive and parse the preset test plan, generate a test task sequence, and issue it in sequence. The test task sequence includes basic operation instructions and robotic arm movement instructions. The preset test plan includes the sample number, sample composition information, SEM test parameters, and test sequence corresponding to each sample carrier position in the multi-station sample stage 6. According to the preset test plan, the task management and scheduling module converts the test objectives of each sample carrier position into basic operation units in sequence according to the preset disassembly rules. The basic operation units include moving the stage, setting parameters, automatically adjusting brightness / contrast, automatically eliminating astigmatism, automatically focusing, and automatically acquiring images. The basic operation units corresponding to all sample carrier positions form a task queue. The robot arm opens the door, grabs the sample stage, closes the door, evacuates and breaks the vacuum, and removes the sample stage, respectively, at the beginning and end of the task queue. The basic operation instructions include the task queue, evacuation and vacuum breaking instructions, and the robot arm movement instructions include the robot arm opens the door, grabs the sample stage, closes the door, and removes the sample stage.

[0029] The vacuum status monitoring module acquires vacuum level data within the scanning electron microscope (SEM) chamber before the door is opened. The task management and scheduling module determines whether the chamber pressure has reached atmospheric pressure based on this data. If so, the module sends a door-opening command to the robotic arm control system; otherwise, it sends a vacuum-breaking command to the SEM control system 2 until the chamber pressure reaches atmospheric pressure. After the door is closed, the module checks the door locking sensor signal. Once the door is confirmed closed, the module sends a vacuum-evacuation command to the SEM control system 2. The SEM control system 2 then controls the scanning electron microscope 3 to perform vacuum evacuation. The vacuum status monitoring module receives real-time vacuum level data within the SEM chamber. When the vacuum level data is less than or equal to the vacuum threshold, the vacuum is considered adequate. The module then feeds back the result or status information to the intelligent scientist platform, and the task management and scheduling module sends a start-test signal.

[0030] This invention establishes a closed-loop interlocking control mechanism based on the hatch status and vacuum status. By acquiring the hatch locking status and vacuum parameters in real time, the next stage operation is triggered only when safety conditions are met, realizing full-process status interlocking control in the automatic SEM sample changing process. It solves the problem of equipment damage or test failure caused by hatch not closing or insufficient vacuum during automatic sample changing, and ensures the safety and reliability of the automatic testing process under unattended conditions.

[0031] The task management and scheduling module issues instructions from each basic operation unit sequentially. During the execution of each basic operation unit, the stage is moved to the coordinates of the current sample under test, and parameters are set, brightness / contrast is automatically adjusted, astigmatism is automatically corrected, focus is automatically adjusted, and images are automatically acquired. After image acquisition, the image and parameter data of the current sample under test are transmitted to the data retrieval module. The data retrieval module automatically associates the corresponding sample information according to the sample station number. After completing the test of the current sample, the stage is moved to the coordinates of the next sample under test, until all samples under test are tested.

[0032] SEM control system 2 receives basic operation commands and controls scanning electron microscope 3 to complete vacuum breaking, vacuuming, and sample testing. After each basic operation command is completed, SEM control system 2 feeds back the execution result or status information to intelligent scientist platform 1. Status information includes vacuum level data, acquired images, and door locking signals. SEM control system 2 is equipped with an SEM automation gateway service. Intelligent scientist platform 1 communicates with both the SEM automation gateway service and the robotic arm control system via HTTP. Basic operation commands issued by intelligent scientist platform 1 are transmitted to the SEM automation gateway service in the form of HTTP requests. The SEM automation gateway service converts the HTTP requests into AutoScript function calls. SEM control system 2 controls scanning electron microscope 3 to complete vacuum breaking, vacuuming, and sample testing based on the function call results. The execution result or status information of scanning electron microscope 3 is fed back to intelligent scientist platform 1 via Kafka messages. The execution status of robotic arm 5 is also fed back to intelligent scientist platform 1 via Kafka messaging.

[0033] The Intelligent Scientist Platform 1, acting as the system control hub, establishes a command transmission channel with the robotic arm control system via the HTTP protocol. After executing the commands, the robotic arm control system provides feedback on the execution status via Kafka messages. Simultaneously, the Intelligent Scientist Platform 1 establishes a communication connection with the SEM Automation Gateway service via the HTTP protocol. The SEM Automation Gateway service is deployed in the network environment where the SEM control system 2 resides. Internally, it encapsulates the AutoScript application programming interface provided by the SEM vendor, which is used to convert the platform's HTTP requests into AutoScript function calls and feed back the execution results or status data to the Intelligent Scientist Platform 1 via Kafka messages. The vacuum status monitoring module receives data from the vacuum gauge forwarded by the SEM Automation Gateway service in real time via Kafka messages, obtains the chamber vacuum level, and performs threshold judgments within the Intelligent Scientist Platform 1 to control the start and stop of processes such as vacuuming and testing.

[0034] In this invention, the SEM control system adopts the original embedded control system of the ThermoFisher Apreo field emission scanning electron microscope without any hardware modifications. To achieve networked remote control, this invention deploys a self-developed automated gateway service in the network environment of the SEM control system. This gateway service is written in Python, provides an HTTP interface externally, and internally encapsulates the AutoScript interface provided by ThermoFisher. When the intelligent scientist platform sends an HTTP request to the gateway, the gateway converts it into a corresponding AutoScript function call, encapsulates the return value in JSON format, and returns it to the platform via the Kafka service. The specific implementation of this gateway service is based on the AutoScript Python package, and those skilled in the art can develop it themselves by referring to the official AutoScript documentation.

[0035] The vacuum status monitoring module is a software submodule within the platform. It uses the SEM automation gateway service to call AutoScript functions to obtain the actual vacuum level, reading it cyclically. Before the robotic arm performs the door opening operation, the module checks the most recently received vacuum level value. If this value is ≥101.3 kPa (atmospheres), it determines that the door can be safely opened and sends a "door opening condition met" signal to the task management and scheduling module via the Kafka message queue. After the robotic arm completes the door closing operation, the module first queries the door locking sensor signal through the gateway. After confirming the locking, it sends a vacuuming command to SEM. Subsequently, the module receives vacuum level data from the Kafka service in real time. When the value is ≤1×10⁻⁶ kPa, it checks the vacuum level. -3 When Pa is reached, the vacuum level is determined to be met. Then, a message is sent to the message server via Kafka. The executor sends the next instruction, the scanning electron microscope configures the software parameters and test environment, and begins to execute the test task sequence.

[0036] The data retrieval module, also a software module running within the Intelligent Scientist platform, uses Python's PIL library to implement image format conversion. This module provides an HTTP upload interface. After each image acquisition, the SEM automation gateway service proactively calls this interface to upload the raw image data (TIFF format) along with metadata such as the current workstation number and test parameters to the Intelligent Scientist platform's data retrieval module. Users can view experimental results and corresponding experimental parameters at any time through the platform's task list interface.

[0037] Users configure experimental parameters through the graphical user interface provided by the intelligent scientist platform. After submission, the platform generates a global test task sequence. The task management and scheduling module acts as the executor, employing an instruction-feedback-stepping mechanism. It sequentially retrieves an instruction from the task queue and sends it to the corresponding execution unit (scanning electron microscope and robotic arm) via an HTTP interface. Then, it listens for the completion message of the instruction through the Kafka service. After completing the current instruction, each execution unit sends a message to Kafka containing the execution status (success / failure) and result data (such as image data and vacuum degree values). Throughout the entire process, users can view the task execution status, experimental results, and parameters in real time through the user interface, achieving a fully automated closed-loop process from instruction issuance and execution feedback to data collection.

[0038] The robotic arm control system receives motion commands from the robotic arm and controls robotic arm 5. In this embodiment, robotic arm 5 is a six-axis articulated industrial robotic arm. Robotic arm 5 is mounted on an independent base outside the scanning electron microscope chamber. The workspace of robotic arm 5 covers the sample stage storage area, the scanning electron microscope chamber door, and the support base position inside the scanning electron microscope chamber. See also... Figure 5 The end effector 51 of the robotic arm 5 is equipped with a QR code camera 511 and a gripper assembly 512. The gripper assembly 512 is adapted to the interactive interface 4 on the scanning electron microscope and is used to open or close the hatch of the scanning electron microscope. The QR code camera 511 captures the QR code on the hatch of the scanning electron microscope 3 at multiple candidate points and sends it to the robotic arm control system. The robotic arm control system calculates the actual position of the center of the support base in the scanning electron microscope. The robotic arm control system controls the gripper assembly 512 to grasp the multi-station sample stage 6 and move it above the support base. The gripper assembly 512 lowers the multi-station sample stage 6 onto the support base. After the robotic arm 5 completes a robotic arm movement command, the robotic arm control system feeds back the execution status to the intelligent scientist platform 1. The execution status includes execution success and execution failure.

[0039] This invention employs QR code multi-view repositioning, enabling the robotic arm to overcome the limitations of its own repeatability accuracy in placement. In existing technologies, the commonly used positioning method for robotic arms is single-shot positioning, where the robotic arm captures an image of the target object in a fixed posture and calculates the target object's pose using visual algorithms. This method is highly susceptible to environmental influences, and random errors cannot be eliminated, making it impossible to verify the reliability of the results. This invention, through multi-view, multi-point adaptive sampling and data fusion, obtains target pose information with higher accuracy and stability. The results are verifiable and highly automated.

[0040] In this invention, the robotic arm control system utilizes the Universal Robots UR5e six-axis collaborative robot, with a repeatability accuracy of ±0.03mm. The system supports the RTDE (Real-Time Data Exchange) real-time control protocol, enabling real-time motion control and status data synchronization via a standard TCP / IP connection, and accepts URScript script strings to execute tasks. The intelligent scientist platform directly issues motion commands via HTTP. After the robotic arm control system completes execution, it sends feedback on the execution status (success / failure and current pose) to the platform via Kafka messages.

[0041] This invention requires no modification to the shape of the scanning electron microscope; it only requires the installation of an interface 4 at the original manual handle of the scanning electron microscope. (See [link]) Figure 4 The interaction interface 4 includes a cylindrical head 41, a connecting cylindrical section 43, and an intermediate constricted shaft section 42 located between the cylindrical head 41 and the connecting cylindrical section 43. The intermediate constricted shaft section 42 is adapted to the gripper assembly 512. A U-shaped fork arm 44 is provided on the end face of the connecting cylindrical section 43, and the U-shaped fork arm 44 is engaged with the door handle of the scanning electron microscope. The gripper assembly 512 and the interaction interface 4 form a detachable mechanical coupling. By applying a translational pulling force or a pushing force along the opening and closing direction of the door, the opening and closing actions of the door are automatically executed. The gripper assembly 512 is equipped with a force sensor to monitor the clamping force value in real time. When the clamping force reaches a preset threshold, it is determined that the gripping is reliable, preventing the sample stage from slipping due to insufficient clamping force or deforming due to excessive clamping force. The gripper assembly also has a degree of rotational freedom to adjust the gripper posture to adapt to different operating requirements.

[0042] This invention enables the robotic arm to automatically take over the external operating mechanism of the scanning electron microscope through the interactive interface 4. This allows the robotic arm to not only transport multi-station sample stages, but also perform equipment operations that were originally done manually, such as opening and closing the hatch, thus realizing the automated control of the external physical interface of the scanning electron microscope.

[0043] The QR code camera 511 is used to capture the QR code fixed on the scanning electron microscope door. Through multi-view sampling and data fusion repositioning method, the precise pose of the door and the center of the internal support base is obtained to ensure that the robotic arm can complete the door opening and closing and the placement of the multi-station sample stage into the designated position inside the scanning electron microscope door with high precision.

[0044] The system also includes an automated material handling device (mobile robot) to automatically transport multi-station sample stages from the material storage area to a storage station accessible to the robotic arm, and to automatically return the tested sample stages to the storage area after testing, thereby automating the sample handling process.

[0045] Working principle: See Figure 6Taking a multi-station sample stage with 6 stations and 20 stations as an example, the high-throughput fully automated testing system for scanning electron microscopes achieves a closed-loop automated workflow from sample loading, automatic sample feeding, automatic testing, automatic sample unloading to data recording as follows: (1) Task Generation and Issuance: Users pre-set the test plan for 20 sample stages on the intelligent scientist platform interface, including the sample number, sample composition information, SEM test parameters (accelerating voltage, electron beam current, working distance, magnification, detector type, etc.) and test sequence for each stage. The task management and scheduling module, based on the preset plan, breaks down the test objectives of each sample into basic operation units (such as moving the stage, setting the voltage, autofocusing, and image acquisition), and automatically inserts global operation units such as robotic arm opening, grabbing the sample stage, closing the door, vacuuming, breaking the vacuum, and removing the sample stage at the beginning and end of the sequence, ultimately generating a complete global test task sequence. The platform adopts an instruction-feedback-stepping mechanism. After issuing each instruction, it waits for the completion message of the execution unit via Kafka before continuing with the next instruction. This generates a complete test task sequence.

[0046] (2) Pre-opening status confirmation: The vacuum status monitoring module receives the chamber vacuum data forwarded by the SEM automated gateway service via Kafka messages. If the current chamber is in a vacuum state, the platform first sends a venting command through the SEM automated gateway service. Only after the pressure returns to atmospheric pressure is the opening operation allowed. The QR code camera at the end of the robotic arm photographs the QR code fixed on the door. Through multi-view sampling and data fusion repositioning, the precise pose of the door and the center of the internal support base is obtained.

[0047] (3) Door opening: The robotic arm moves to the scanning electron microscope door, and the gripper assembly mechanically couples with the door interface fixed to the original manual handle. A preset translational pulling force is applied along the door opening direction, and the door is slowly pulled open to the predetermined position. The gripper has a built-in force sensor that monitors the gripping force in real time. When the preset threshold is reached, the gripping is considered reliable. After the door is opened, the robotic arm control system feeds back the status to the platform via Kafka.

[0048] (4) Sample grabbing stage The robotic arm moves to the storage position of the multi-station sample stage, and the gripper grasps the 20 sample stages pre-loaded with samples using a preset threshold force. Once the force sensor reaches the threshold, it determines that the gripping is reliable, preventing slippage due to insufficient force or deformation of the sample stage due to excessive force. Feedback is sent via Kafka after the grasp is complete. The robotic arm's motion path is a fixed trajectory pre-calibrated and stored in the robotic arm control system through a teaching method. Therefore, the intelligent scientist platform does not need to calculate path planning instructions in real time. Instead, it selects the corresponding pre-stored path identifier based on the current task type (such as opening a door, retrieving a sample stage, placing it into a chamber, etc.) and sends this identifier to the robotic arm control system via an HTTP interface. After receiving the identifier, the robotic arm control system directly calls the corresponding internally stored trajectory to execute the movement, completing the corresponding spatial position movement and attitude adjustment.

[0049] (5) Sample stage placement in the chamber: The end-effector QR code camera is used again to scan the QR code on the chamber door. The precise pose of the center of the support base is obtained in real time through repositioning calculation, which is used as the target position for sample stage placement. The robotic arm carries the sample stage to the predetermined docking position outside the scanning electron microscope chamber. The robotic arm extends the sample stage into the chamber and uses the vision positioning system to precisely align the positioning reference post at the bottom of the sample stage with the positioning groove of the built-in support base of the scanning electron microscope, completing the loading. The robotic arm releases the gripper, exits the chamber, and provides feedback via Kafka.

[0050] (6) Door Closure and Status Confirmation: After the robotic arm retracts to the safe position, the gripper assembly clamps the door interface again and applies a translational thrust along the closing direction to close the door. The platform queries the door locking sensor signal and the chamber vacuum data through the SEM gateway to confirm that the door has been closed. The robotic arm returns to the initial standby position. Feedback is provided via Kafka after each step is completed.

[0051] (7) Vacuuming and Testing: The intelligent scientist platform sends a vacuuming start command to the SEM control system via the SEM gateway (HTTP to AutoScript). The SEM starts the vacuum pump group to evacuate the chamber. The platform receives vacuum level data pushed by the SEM gateway in real time via Kafka. When the vacuum level reaches the preset test threshold (≤1×10⁻⁶), the platform will initiate the vacuuming process. -3When the vacuum status monitoring module reaches (Pa), it automatically triggers the task management and scheduling module to start the test. Following the task sequence generated in step 1, the task management and scheduling module issues commands one by one through the SEM gateway: first, it moves the stage to the coordinates of the first sample; then, it sequentially sets parameters such as accelerating voltage, electron beam current, detector type, and magnification; and performs automatic focusing, astigmatism correction, contrast and brightness adjustment; finally, it captures a high-resolution image under the corresponding conditions. Each time the SEM completes a command, it returns the execution result via Kafka. The image data is transmitted to the platform's data recovery module via the SEM gateway, where it undergoes format conversion (e.g., from TIFF to JPG) and is stored in the local database, while also being associated with the corresponding sample information and workstation number. After completing all the scheduled tests for the first sample, the stage moves to the coordinates of the second sample, and the above process is repeated until all tests for the 20th sample are completed.

[0052] (8) Sample unloading and recovery: After all sample tests are completed, the intelligent scientist platform sends a vacuum release command through the SEM gateway. After the chamber pressure is restored to atmospheric pressure, the robotic arm performs the door opening operation (same as step 3), removes the tested sample stage, and transfers it to the designated unloading area. Then the robotic arm closes the chamber door (same as step 6), and the platform sends a vacuum pumping command to the SEM control system to restore the chamber to a vacuum state for standby.

[0053] (9) Data Archiving: All images and parameter data generated by SEM are automatically uploaded to the data retrieval module of the Intelligent Scientist Platform. The platform archives the data based on the workstation number, sample number, test parameters, timestamp, and other information, and automatically generates a test report. The report includes a sample information table, images of each workstation, and statistics of key parameters.

[0054] If a batch of experiments includes multiple multi-station sample stages, the intelligent scientist platform automatically triggers the unloading process after all condition tests are completed on one sample stage. Specifically, the platform sends a vacuum breaking command through the SEM automation gateway. After the Kafka feedback indicates that the chamber pressure has returned to atmospheric pressure, the platform sends an opening command to the robotic arm control system via HTTP and waits for a Kafka completion message. After the robotic arm performs the translational opening action, it removes the tested sample stage and moves it to the designated unloading area, and sends a completion status feedback via Kafka. Subsequently, the robotic arm automatically loads and executes the tests according to steps 4 to 7 in Example 2. Each step follows a command-feedback-step mechanism, that is, after each command is issued, it waits for a Kafka completion message from the corresponding execution unit (robotic arm or SEM gateway) before continuing. This cycle repeats automatically until all sample stages under test have completed the tests. The entire process requires no manual intervention, achieving high-throughput continuous testing.

[0055] Compared with the prior art, the advantages provided by the present invention are as follows: (1) This invention enables continuous testing with a single loading by using a customized multi-station sample stage. This significantly increases the throughput of scanning electron microscopy by providing a single sample, overcoming the efficiency bottleneck of traditional single-sample or small-batch testing.

[0056] (2) The robotic arm can not only perform the sample stage loading and unloading operations, but also independently perform the automatic opening and closing of the scanning electron microscope chamber without any hardware modification to the scanning electron microscope. It realizes fully automated control of the external physical interface of the scanning electron microscope without manual intervention.

[0057] (3) Using the intelligent scientist platform as the intelligent control center, an asynchronous communication architecture of HTTP command issuance + Kafka message feedback is adopted. Through a unified task description format, global task queue scheduling and command-feedback-stepping mechanism, seamless connection between the scanning electron microscope and the robotic arm is achieved, forming an integrated automated process from automatic sample loading, vacuuming, testing to data recovery, which fully meets the integration needs of high-throughput unmanned laboratories.

[0058] The mechanical motion commands and SEM parameter control share the same task sequence and are uniformly scheduled by the same upper-level platform. For example, after the door closing action is completed, the intelligent scientist platform immediately sends a vacuuming command to the scanning electron microscope; after the test is completed, the intelligent scientist platform immediately sends an opening command to the robotic arm control system, forming true full-process automation.

[0059] (4) The coordinates of each bearing position of the multi-station sample stage are manually pre-calibrated and entered into the system; during the test, the data recovery module automatically associates the sample information, test parameters and timestamp according to the current test station number and stores them in the database to ensure the consistency and traceability of the test data of each station sample.

[0060] (5) The system architecture of the present invention has good modularity. The robotic arm and intelligent scientist platform can be adapted to different brands and models of scanning electron microscope equipment, and have high versatility and promotion value.

[0061] Example 2 The difference between this embodiment and Embodiment 1 is that this embodiment provides a fully automated high-throughput scanning electron microscope testing method, using the fully automated high-throughput scanning electron microscope testing system from Embodiment 1. The method includes the following steps: S1. Once the pressure inside the scanning electron microscope chamber reaches atmospheric pressure, the robotic arm control system controls the gripper assembly to open the scanning electron microscope chamber door.

[0062] S2. The task management and scheduling module sends a sample stage grabbing command. The QR code camera captures the QR code on the hatch at multiple candidate points and sends it to the robotic arm control system. The robotic arm control system calculates the actual position of the center of the support base in the scanning electron microscope. The robotic arm control system controls the gripper assembly to grab the multi-station sample stage and move it above the support base in the scanning electron microscope. The gripper assembly lowers the multi-station sample stage onto the support base.

[0063] The process by which the robotic arm control system calculates the actual position of the center of the support base in the scanning electron microscope includes: The robotic arm control system calculates the transformation matrix of the QR code relative to the center of the support base from each viewpoint. The final pose of the QR code is obtained through data fusion; By combining the pre-calibrated QR code with the fixed conversion relationship between the center of the sample stage support base, the actual position of the center of the support base is calculated.

[0064] This invention employs a multi-view visual repositioning and pose fusion method based on QR code identification to establish a unified coordinate relationship between the hatch interaction interface, the internal support base, and the robotic arm, thereby achieving high-precision repositioning of the SEM hatch and sample stage. It solves the problems of installation errors, sample stage placement errors, and cumulative errors caused by long-term operation in open environments, and ensures high repeatability and long-term stability of the automatic hatch opening and closing and sample loading and unloading processes.

[0065] S3. The task management and scheduling module sends a door-closing command, and the robotic arm control system controls the gripper assembly to close the door. The task management and scheduling module then sends a vacuuming command. When the vacuum level inside the chamber is less than or equal to the vacuum threshold, the task management and scheduling module sends test commands to the SEM control system according to the test task sequence. The SEM control system controls the scanning electron microscope to sequentially complete the test commands for each sample-bearing position until all sample-bearing positions are tested. Finally, the task management and scheduling module sends a vacuum-breaking command. Once the pressure inside the chamber reaches atmospheric pressure, the robotic arm control system controls the gripper assembly to open the door and remove the multi-station sample stage. The process of the task management and scheduling module sending test commands to the SEM control system according to the test task sequence, and the SEM control system controlling the scanning electron microscope to sequentially complete the test commands for each sample-bearing position, includes: The task management and scheduling module issues instructions from each basic operation unit sequentially. During the execution of each basic operation unit, the stage is moved to the coordinates of the current sample under test, and parameters are set, brightness / contrast is automatically adjusted, astigmatism is automatically corrected, focus is automatically adjusted, and images are automatically acquired. After image acquisition, the image and parameter data of the current sample under test are transmitted to the data retrieval module. The data retrieval module automatically associates the corresponding sample information according to the sample station number. After completing the test of the current sample, the stage is moved to the coordinates of the next sample under test, until all samples under test are tested.

[0066] If the same batch of experiments includes multiple multi-station sample stages, the method also includes step S4: after the test of a multi-station sample stage is completed, the robotic arm control system controls the gripper assembly to move the completed multi-station sample stage to the unloading area, the robotic arm control system controls the gripper assembly to pick up the next multi-station sample stage to be tested, and completes the test according to steps S2 to S3 until all multi-station sample stages are tested.

[0067] This invention establishes a non-intrusive automated testing architecture for existing commercial SEM equipment. Through the collaborative work of a robotic arm, a vision positioning system, and an AutoScript gateway, it achieves automatic sample loading and unloading, automatic testing, and automatic data retrieval without altering the original vacuum chamber structure of the SEM, adding vacuum transport pipelines, or rewriting the underlying SEM control program. It solves the problems of high implementation costs and difficulty in promotion in the automation transformation of existing SEMs, which require structural modifications to the equipment and reconstruction of the control system. It enables existing SEM equipment to have high-throughput unmanned testing capabilities without special modifications.

[0068] This invention automatically associates sample station codes, test tasks, SEM images, and parameter data to form a fully digital management mechanism from sample loading to data archiving; it solves the problem of mismatch between sample codes and test results and the need for manual sorting during high-throughput testing; and it enables automatic traceability and automatic archiving of test data under unattended conditions.

[0069] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A high-throughput, fully automated testing system for scanning electron microscopy, characterized in that: include: Multi-station sample stage, including Individual presentation The sample-bearing sites are evenly distributed in an array for simultaneous loading. One sample to be tested; The intelligent scientist platform includes a task management and scheduling module. The task management and scheduling module is used to receive and parse the preset test plan, generate test task sequences, and issue them in sequence order. The test task sequences include basic operation instructions and robotic arm movement instructions. The SEM control system is used to receive basic operation instructions and control the scanning electron microscope to complete vacuum breaking, vacuuming and sample testing. After the scanning electron microscope completes a basic operation instruction, the SEM control system will feed back the execution result or status information to the intelligent scientist platform. The robotic arm control system receives and controls the robotic arm's movement commands. The end effector of the robotic arm is equipped with a QR code camera and a gripper assembly. The gripper assembly is adapted to the interface on the scanning electron microscope (SEM) and is used to open or close the SEM's hatch. The QR code camera captures QR codes on the hatch at multiple candidate points and sends them to the robotic arm control system. The robotic arm control system calculates the actual position of the center of the support base within the SEM. The robotic arm control system then controls the gripper assembly to pick up the multi-station sample stage and move it above the support base. The gripper assembly then lowers the multi-station sample stage onto the support base. Each time the robotic arm completes a movement command, the robotic arm control system feeds back the execution status to the intelligent scientist platform.

2. The high-throughput fully automated testing system for scanning electron microscopy according to claim 1, characterized in that: The preset test plan includes the sample number, sample composition information, SEM test parameters, and test sequence corresponding to each sample carrier position in the multi-station sample stage. According to the preset test plan, the task management and scheduling module converts the test objectives of each sample carrier position into basic operation units in sequence according to the preset disassembly rules. The basic operation units include moving the stage, setting parameters, automatically adjusting brightness / contrast, automatically eliminating astigmatism, automatically focusing, and automatically acquiring images. The basic operation units corresponding to all sample carrier positions form a task queue. The robot arm opens the door, grabs the sample stage, closes the door, evacuates and breaks the vacuum, and removes the sample stage, respectively, at the beginning and end of the task queue. The basic operation instructions include the task queue, evacuation and vacuum breaking instructions, and the robot arm movement instructions include the robot arm opens the door, grabs the sample stage, closes the door, and removes the sample stage.

3. The high-throughput fully automated testing system for scanning electron microscopy according to claim 1, characterized in that: The SEM control system is equipped with an SEM automation gateway service. The intelligent scientist platform communicates with both the SEM automation gateway service and the robotic arm control system via HTTP protocol. Basic operation instructions issued by the intelligent scientist platform are transmitted to the SEM automation gateway service in the form of HTTP requests. The SEM automation gateway service converts the HTTP requests into AutoScript function calls. Based on the function call results, the SEM control system controls the scanning electron microscope to complete vacuum breaking, vacuuming, and sample testing. The execution results or status information of the scanning electron microscope are fed back to the intelligent scientist platform via Kafka messages. The execution status of the robotic arm is also fed back to the intelligent scientist platform via the Kafka messaging mechanism.

4. The high-throughput fully automated testing system for scanning electron microscopy according to claim 1, characterized in that: The intelligent scientist platform also includes: The vacuum status monitoring module acquires vacuum level data within the scanning electron microscope (SEM) chamber before the door is opened. The task management and scheduling module determines whether the chamber pressure has reached atmospheric pressure based on this data. If so, the module sends a door-opening command to the robotic arm control system; otherwise, it sends a vacuum-breaking command to the SEM control system until atmospheric pressure is reached. After the door is closed, the module checks the door locking sensor signal. Once confirmed, it sends a vacuum-evacuation command to the SEM control system, which then controls the SEM to perform vacuum evacuation. The vacuum status monitoring module receives real-time vacuum level data within the SEM chamber. When the vacuum level data is less than or equal to the vacuum threshold, the vacuum is considered adequate. The module then feeds back the result or status information to the intelligent scientist platform, and the task management and scheduling module sends a start-test signal.

5. The high-throughput fully automated testing system for scanning electron microscopy according to claim 2, characterized in that: The intelligent scientist platform also includes a data retrieval module and a task management and scheduling module that issues instructions from the basic operation units one by one. During the execution of each basic operation unit, the stage is moved to the coordinates of the current sample to be tested, and parameters are set, brightness / contrast is automatically adjusted, astigmatism is automatically corrected, focus is automatically adjusted, and images are automatically acquired in sequence. After image acquisition, the image and parameter data of the current sample to be tested are transmitted to the data retrieval module. The data retrieval module automatically associates the corresponding sample information according to the sample station number. After the test of the current sample to be tested is completed, the stage is moved to the coordinates of the next sample to be tested, until the test of all samples to be tested is completed.

6. The high-throughput fully automated testing system for scanning electron microscopy according to claim 1, characterized in that: The interface includes a cylindrical head, a connecting cylindrical section, and an intermediate constricted shaft section located between the cylindrical head and the connecting cylindrical section. The intermediate constricted shaft section is adapted to the gripper assembly. The end face of the connecting cylindrical section is provided with a U-shaped fork arm, which is engaged with the door handle of the scanning electron microscope.

7. A high-throughput, fully automated testing method for scanning electron microscopy, employing the testing system described in any one of claims 1-6, characterized in that: The methods include: S1. After the pressure inside the scanning electron microscope chamber reaches atmospheric pressure, the robotic arm control system controls the gripper assembly to open the scanning electron microscope chamber door. S2. The task management and scheduling module sends a sample stage grabbing command. The QR code camera captures the QR code on the hatch at multiple candidate points and sends it to the robotic arm control system. The robotic arm control system calculates the actual position of the center of the support base in the scanning electron microscope. The robotic arm control system controls the gripper assembly to grab the multi-station sample stage and move it above the support base in the scanning electron microscope. The gripper assembly slowly lowers the multi-station sample stage onto the support base. S3. The task management and scheduling module sends a door-closing command. The robotic arm control system controls the gripper assembly to close the door. The task management and scheduling module sends a vacuuming command. When the vacuum level in the chamber is less than or equal to the vacuum threshold, the task management and scheduling module sends test commands to the SEM control system in the order of the test task sequence. The SEM control system controls the scanning electron microscope to complete the test commands for each sample bearing position in sequence until all sample bearing positions are tested. The task management and scheduling module sends a vacuum-breaking command. After the pressure in the chamber reaches atmospheric pressure, the robotic arm control system controls the gripper assembly to open the door and remove the multi-station sample stage.

8. The high-throughput fully automated testing method for scanning electron microscopy according to claim 7, characterized in that: The process in step S2 where the robotic arm control system calculates the actual position of the center of the support base in the scanning electron microscope includes: The robotic arm control system calculates the transformation matrix of the QR code relative to the center of the support base from each viewpoint. The final pose of the QR code is obtained through data fusion; By combining the pre-calibrated QR code with the fixed conversion relationship between the center of the multi-station sample stage support base, the actual position of the support base center is calculated.

9. The high-throughput fully automated testing method for scanning electron microscopy according to claim 7, characterized in that: In step S3, the task management and scheduling module sends test instructions to the SEM control system according to the test task sequence. The process by which the SEM control system controls the scanning electron microscope to complete the test instructions for each sample bearing position in sequence includes: The task management and scheduling module issues instructions from the basic operation units one by one. During the execution of each basic operation unit, the stage is moved to the coordinates of the current sample to be tested, and the parameters are set, brightness / contrast is automatically adjusted, astigmatism is automatically corrected, focus is automatically adjusted, and images are automatically acquired in sequence. After the images are acquired, the image and parameter data of the current sample to be tested are transmitted to the data retrieval module. The data retrieval module automatically associates the corresponding sample information according to the sample station number. After the test of the current sample to be tested is completed, the stage is moved to the coordinates of the next sample to be tested, until the test of all samples to be tested is completed.

10. The high-throughput fully automated testing method for scanning electron microscopy according to claim 7, characterized in that: It also includes step S4: After completing the test of a multi-station sample stage, the robotic arm control system controls the gripper assembly to move the completed multi-station sample stage to the unloading area, the robotic arm control system controls the gripper assembly to pick up the next multi-station sample stage to be tested, and completes the test according to steps S2 to S3 until all multi-station sample stages are tested.

Images