Wafer reloading optimization system, mounting equipment and wafer reloading optimization method
By introducing a state storage module and an exception handling module into the placement equipment, the wafer changeover process is optimized, and a placement sequence with the minimum number of changes is generated, solving the problem of frequent wafer disk switching and improving equipment efficiency and stability.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-29
- Publication Date
- 2026-03-27
AI Technical Summary
In the existing technology, the frequent switching of wafer disks on the equipment leads to low equipment efficiency, especially the problems of ineffective switching and repeated loading and unloading of baskets when switching between multiple materials.
By introducing a state storage module, a sequence decision engine, a task queue manager, and an exception handling module into the placement equipment, the wafer changeover process is optimized, a placement sequence with the minimum number of changes is generated, and an exception local repair mechanism is combined to avoid repeated pull-out and invalid switching.
Without increasing hardware costs, the frequency of wafer disk switching was reduced, non-production time was shortened, equipment efficiency and operational stability were improved, and resource waste and production line downtime were reduced.
Smart Images
Figure CN121752092A_ABST
Abstract
Description
Technical Field
[0001] This application belongs to the field of wafer mounting technology, and particularly relates to a wafer material changing optimization system, mounting equipment and wafer material changing optimization method. Background Technology
[0002] In precision manufacturing processes such as semiconductor packaging and die bonding, equipment typically needs to mount multiple different types of wafer materials on the same substrate. To enable multi-material mounting, the equipment is equipped with a basket structure for temporarily storing wafer disks. This basket can hold multiple wafer disks, and the wafers in the disks can be of the same or different materials, in states including full disk, half disk, or empty disk after recycling.
[0003] When the equipment is operating, the target wafer disk must be pulled from the basket into the equipment's working area, and a series of preparatory steps, such as alignment, calibration, and vacuum adsorption, must be completed before the placement operation can begin. However, the time required from pulling the wafer disk out of the basket to its readiness is relatively long; similarly, when changing materials, the currently used wafer disk must be completely removed from the working area and returned to the basket before the wafer disk corresponding to the next material is pulled in, making the entire changeover process significantly time-consuming. Especially when the placement task involves multiple materials in a fixed order, if each startup ignores the current workstation status and forces a start from a specific material, it can easily lead to unnecessary repeated wafer disk loading and unloading, significantly reducing the overall efficiency of the equipment.
[0004] Therefore, on a single device, the fixed mounting sequence and lack of state awareness mechanism lead to frequent and redundant switching of wafer disks. Summary of the Invention
[0005] This application aims to address at least one of the technical problems existing in the prior art. To this end, this application proposes a wafer changing optimization system, a mounting equipment, and a wafer changing optimization method, which reduces the wafer disk switching frequency without increasing hardware costs.
[0006] In a first aspect, this application provides a wafer replacement optimization system, the system comprising: The system is applied to a mounting equipment for mounting wafers containing various material types. The mounting equipment is equipped with a basket for storing the wafers to be mounted and a work area station for mounting the wafers to be mounted. The system includes an interconnected state storage module, a sequence decision engine, a task queue manager, and an exception handling module. The status storage module is used to store the material type corresponding to the wafer disk to be mounted on the work area station as the end status information when the first mounting task is completed. The sequence decision engine is used to read the end status information when the second mounting task is started, and take the material type corresponding to the end status information as the starting point, determine the scheduling priority of each task to be executed in the second mounting task according to the material rotation order in the basket and the material demand sequence of the mounting equipment, so as to generate a mounting sequence that minimizes the number of material changes according to the scheduling priority. The task queue manager is used to group and rearrange the original mounting points according to the material type based on the mounting order, generate a task queue for the tasks to be executed, and skip the completed part according to the set of mounted materials and execute the remaining tasks if the tasks to be executed are interrupted. The exception handling module is used to detect abnormal events during the operation of the mounting equipment. If the abnormal event is a material picking failure, it triggers a material replenishment operation at the failed point and resumes the execution flow of the task queue after the material replenishment is completed.
[0007] According to one embodiment of this application, the sequence decision engine has a built-in weighted scoring model. The weighted scoring model is constructed based on the material changeover time factor, the process compatibility factor, and the order delivery urgency factor, and is used to comprehensively score the candidate mounting sequence. The material changeover time factor is calculated based on the switching distance between adjacent material types in the basket. The process compatibility factor is determined by the matching degree between the current material and the process parameters of the target station. The order delivery urgency factor is associated with the ratio of the remaining processing time to the delivery deadline of the order to which the task belongs. The weighted scoring model is used to score the material change path by linearly weighting the factors according to their weight coefficients, and the scoring result is used to adjust the scheduling priority.
[0008] According to one embodiment of this application, the state storage module is further configured to: Once the material switching is complete, update the material type on the workstation in the work area, and trigger a status snapshot to save the current material type, the completed material set, and the task queue position when the equipment stops abnormally. The task queue manager is used to skip all original mounting points corresponding to the mounted materials according to the completed material set when the device is restarted or reset, and continue to execute the remaining tasks to be executed from the task queue position. The sequence decision engine generates the mounting order with the current material type as the starting point of the new mounting task.
[0009] According to one embodiment of this application, the task queue manager is further configured to generate a preload instruction when the task queue is generated, to instruct the basket to move the next wafer to be attached to the pull-in position.
[0010] According to one embodiment of this application, the mounting equipment is provided with a visual calibration unit, and the anomaly handling module is further configured to: If the detected abnormal event is an abnormal placement offset, the vision calibration unit is invoked to perform coordinate compensation on the workstation in the work area, and the positioning parameters of the placement point are adjusted according to the calibration results. If the number of consecutive occurrences of the same type of abnormal event exceeds the threshold, the handling strategy for the corresponding material type will be updated to the historical experience database, and an alarm notification will be generated. In the event that the abnormal event is an unrecoverable hardware failure, the task queue is frozen and the execution context is saved; Upon completion of the replenishment operation, the task queue manager is triggered to re-evaluate the scheduling priority of the remaining tasks to be executed in order to update the mounting order.
[0011] According to one embodiment of this application, the state storage module is coupled to the basket control system and is used to read the material type of the wafer disk to be mounted in each slot of the basket, and to feed back the empty disk status or half disk status to the sequence decision engine before generating the mounting sequence.
[0012] According to one embodiment of this application, when material switching is performed at the work area station, the difference in material type of the wafer disk to be mounted is compared. If the material type is the same before and after the switching, the wafer disk pulling in and preparation work is skipped, and the wafer disk to be mounted at the work area station is called to continue the mounting process. When the material types are different before and after the switch, the mounting equipment is controlled to return the wafer disk to be mounted on the work area station to the basket, and pull the wafer disk of the target material type from the basket.
[0013] Secondly, this application provides a mounting device equipped with a wafer replacement optimization system as described in the first aspect.
[0014] According to one embodiment of this application, the mounting equipment is provided with a basket for storing wafer disks to be mounted, a work area station for mounting the wafer disks to be mounted, and a vision calibration unit. The work area station is used to mount the wafer disk to be mounted. The visual calibration unit is used to perform coordinate compensation on the workstation in the work area.
[0015] Thirdly, this application provides a wafer material change optimization method, applied to a mounting equipment. The mounting equipment is used to perform mounting operations on wafer disks containing multiple material types. The mounting equipment is equipped with a basket for storing the wafer disks to be mounted, and a work area station for mounting the wafer disks to be mounted. The method includes: By using the status storage module, when the first mounting task is completed, the material type corresponding to the wafer disk to be mounted on the work area station is retained as the end status information and stored. Using the sequential decision engine, when the second mounting task is started, the end status information is read, and the scheduling priority of each task to be executed in the second mounting task is determined according to the material type corresponding to the end status information, the material rotation order in the basket and the material demand sequence of the mounting equipment, so as to generate a mounting sequence that minimizes the number of material changes according to the scheduling priority. The task queue manager regroups and rearranges the original mounting points according to the material type based on the mounting order, generating a task queue for the tasks to be executed. If the tasks to be executed are interrupted, the completed part is skipped based on the set of mounted materials, and the remaining tasks are executed. The exception handling module detects abnormal events during the operation of the mounting equipment. If the abnormal event is a material picking failure, a material replenishment operation is triggered at the failed point, and the execution flow of the task queue is restored after the material replenishment is completed.
[0016] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application.
[0017] The wafer replacement optimization method, apparatus, electronic equipment, medium, and product provided in this application have the following advantages over the prior art: (1) By reusing the workstation wafer disk at the end of the previous task through state memory, combined with the round-robin scheduling and abnormal local repair mechanism, without increasing hardware costs, repeated disk pulling, invalid switching and global reordering are effectively avoided, the frequency of wafer disk switching is reduced, non-production time is shortened, the overall efficiency and operational stability of the mounting equipment are improved, and production line stagnation and resource waste caused by material changes are reduced.
[0018] (2) By using multi-dimensional factor weighted scoring, the scheduling strategy is upgraded from a single objective of minimizing material changes to a multi-objective optimization that is efficient, stable, and timely, effectively improving the flexibility of the production line and the utilization rate of equipment. Attached Figure Description
[0019] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which: Figure 1 This is a schematic diagram of the structure of the wafer reloading optimization system provided in the embodiments of this application; Figure 2 This is a schematic flowchart of the wafer replacement optimization method provided in the embodiments of this application. Detailed Implementation
[0020] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.
[0021] The terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such use of data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first," "second," etc., are generally of the same class and the number of objects is not limited; for example, a first object can be one or more. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship.
[0022] The wafer replacement optimization system, mounting equipment, and wafer replacement optimization method provided in this application will be described in detail below with reference to the accompanying drawings and through specific embodiments and application scenarios.
[0023] The wafer material change optimization system (hereinafter referred to as the "system") is applied to a mounting equipment. The mounting equipment is used to perform mounting operations on wafer disks containing multiple material types. The mounting equipment is equipped with a basket for storing the wafer disks to be mounted and a work area station for mounting the wafer disks to be mounted.
[0024] like Figure 1 As shown, the wafer refueling optimization system includes an interconnected state storage module 110, a sequential decision engine 120, a task queue manager 130, and an exception handling module 140. The state storage module 110 is used to store the material type corresponding to the wafer disk to be mounted on the work station in the work area as the end state information when the first mounting task is completed. The sequence decision engine 120 is used to read the end status information when the second mounting task is started, and take the material type corresponding to the end status information as the starting point, determine the scheduling priority of each task to be executed in the second mounting task according to the material rotation order in the basket and the material demand sequence of the mounting equipment, so as to generate a mounting sequence that minimizes the number of material changes according to the scheduling priority. The task queue manager 130 is used to group and rearrange the original mounting points according to the material type based on the mounting order, generate a task queue of tasks to be executed, and skip the completed part according to the set of mounted materials and execute the remaining tasks if the tasks to be executed are interrupted. The exception handling module 140 is used to detect abnormal events during the operation of the mounting equipment. If the abnormal event is a material picking failure, it triggers a material replenishment operation at the failed point and resumes the execution flow of the task queue after the material replenishment is completed.
[0025] Understandably, the work area station is a mechanical platform in the mounting equipment used to temporarily place the wafer disks currently in use, with a vacuum adsorption device above it to fix the wafer disks; the wafer disks to be mounted are carriers loaded with wafers of the same material type, and each disk corresponds to a unique material identifier, such as A, B, or C; the end status information is used to characterize the material type actually retained at the station when the previous task ended, and is the key memory data for realizing material change optimization.
[0026] The material rotation sequence is a preset fixed cyclic sequence, such as A→B→C→A, used to constrain the scheduling direction to avoid disordered switching; the material requirement sequence is the set of materials required for this task issued by the superior operation formula, such as {A, B, C}.
[0027] The original placement point is a complete list of placement instructions imported from the CAD file, including material type, X / Y coordinates, and rotation angle; the task queue is a dynamic linked list structure organized in execution order, with each item corresponding to a schedulable placement action.
[0028] Pickup failure refers to the nozzle failing to pick up the wafer, which is determined by vacuum pressure sensor or visual pickup verification; replenishment operation is a local retry of a single failure point and does not change the overall scheduling logic.
[0029] In actual execution, after the placement equipment completes the last placement point action and confirms that the task has ended normally or terminated abnormally, the controller reads the current wafer disk ID fed back by the workstation sensor, converts it into a material type string through the material mapping table, and calls the Flash write function of the non-volatile storage driver interface to write the string into the "LastDisk" field of the preset configuration area of the state storage module 110, thus completing the state persistence.
[0030] During the task initialization phase, the sequential decision engine 120 calls read_nvm("LastDisk") to obtain the end status information. If the value is not empty, the material position is located in the rotation sequence, and subsequent materials are selected sequentially until all required materials are covered, generating a new sequence as the mounting order. For example, if LastDisk=C, then [C, A, B] will be output. This sequence ensures that the first material does not need to be pulled in from the basket, thereby reducing one complete material change operation.
[0031] The task queue manager 130 traverses the original mounting point list and groups the points of the same material into a continuous sub-queue according to the order in which the materials appear in the mounting order. During execution, after all points of a material are completed, its material type is added to the "CompletedMaterials" set in memory. If an interruption and restart occur, only material points not in this set are loaded into the queue to achieve accurate continuation.
[0032] The exception handling module 140 listens for the material picking result signal. If it fails, it records the material type and coordinates at that point and checks whether the wafer disk at the current workstation is the target material. If it is, it directly retryes picking and mounting. If not, it temporarily pulls the target disk to complete the material replenishment and then immediately returns it. It also notifies the task queue manager to mark the point as "completed" and continues to execute the next task in the original queue.
[0033] According to the wafer material change optimization system provided in the embodiments of this application, by reusing the workstation wafer disks at the end of the previous task through state memory, combined with the round-robin scheduling and abnormal local repair mechanism, the system effectively avoids repeated disk pulling, invalid switching and global reordering without increasing hardware costs, reduces the frequency of wafer disk switching, shortens non-production time, improves the overall efficiency and operational stability of the mounting equipment, and reduces production line stagnation and resource waste caused by material changes.
[0034] In some embodiments, the sequential decision engine has a built-in weighted scoring model, which is constructed based on the material changeover time factor, the process compatibility factor, and the order delivery urgency factor, and is used to comprehensively score the candidate mounting sequence. The material changeover time factor is calculated based on the switching distance between adjacent material types in the basket, the process compatibility factor is determined by the matching degree between the current material and the process parameters of the target station, and the order delivery urgency factor is related to the ratio of the remaining processing time to the delivery deadline of the order to which the task belongs. The weighted scoring model is used to score the material change path by linearly weighting the factors according to their weight coefficients, and the scoring result is used to adjust the scheduling priority.
[0035] Understandably, the material changeover time factor is used to reflect the mechanical preparation time required to switch from the current material to the target material; the process compatibility factor is used to characterize the degree of commonality of different materials in parameters such as nozzle type, vision template, and mounting pressure; and the order delivery urgency factor is used to quantify the urgency of the task.
[0036] Scheduling priority here is reflected in the ordering of candidate sequences.
[0037] In actual execution, the sequential decision engine first generates all legal candidate sequences that meet the material requirements (such as [C,A,B], [C,B,A]). For each sequence, it calculates three factor scores in sequence: material changeover time is obtained by looking up the distance of the basket slot; process compatibility is calculated by comparing the number of common items in the parameter configuration file; and delivery urgency is determined by the reciprocal of the delivery deadline in the work recipe and the current system time difference. Finally, a linear weighting is performed to obtain the comprehensive scheduling score.
[0038] The sequential decision engine needs to select the best among multiple candidate mounting sequences. Traditional rotation cannot take into account both delivery time and process. The embodiments of this application unify the material change time, process compatibility and delivery urgency into a calculable linear weighted model, and introduce normalization processing to eliminate dimensional differences.
[0039] For example, the overall scheduling score is: in, Candidate sequences The higher the overall scheduling score, the better. For the preset weighting coefficients, satisfy ; For sequence Total refueling time; For materials arrive Switching time, For flower basket slots and The Euclidean distance between them These are calibration coefficients; The largest among all candidate sequences , used for normalization; For sequence average process compatibility, i.e., Jaccard similarity, This is a set of process parameters for material m, such as nozzle ID, visual template ID, etc. To normalize the score for delivery urgency, For materials The remaining delivery time for the order. Guard against zero.
[0040] The system can automatically select the best scheduling sequence with short material changeover time, smooth process switching, and priority for urgent orders, avoiding reliance on manual experience and improving multi-objective collaborative optimization capabilities.
[0041] All candidate sequences are sorted in descending order of score, and the first one is taken as the output. This sequence prioritizes the reuse of workstations while also taking into account high-urgent orders and process compatibility, avoiding inefficient switching caused by simple rotation.
[0042] In this embodiment, by using multi-dimensional factor weighted scoring, the scheduling strategy is upgraded from a single objective of minimizing material changes to a multi-objective optimization that is efficient, stable, and timely, effectively improving production line flexibility and equipment utilization.
[0043] In some embodiments, the state storage module is further configured to: Once the material switching is complete, update the material type on the workstation in the work area, and trigger a status snapshot to save the current material type, the completed material set, and the task queue position when the equipment stops abnormally. The task queue manager is used to skip all original mounting points corresponding to the mounted materials according to the completed material set when the device is restarted or reset, and continue to execute the remaining tasks to be executed from the task queue position. The sequence decision engine generates the mounting order with the current material type as the starting point of the new mounting task.
[0044] Understandably, the task queue position refers to the index value of which patch point has been executed so far.
[0045] In actual execution, each time the wafer disk is successfully pulled into the workstation, the state storage module 110 updates the memory variable current_material string variable; at the same time, a watchdog interrupt service routine is registered. When an emergency stop signal or communication timeout is detected, the snapshot_save() function is immediately called to pack the current_material string variable, the completed_set collection data structure, and the queue_index=158 integer variable into the Flash backup sector.
[0046] After the system starts its self-test, the task queue manager calls the load_snapshot() function to read the snapshot data; it traverses the original mounting point list, adds only the points whose materials are not in the completed_set collection data structure to the execution queue, and starts scheduling from the corresponding position of queue_index; at the same time, the sequence decision engine treats the current_material string variable as the last disk information, which is used as the starting point for the next round of task scheduling.
[0047] The string variable current_material represents the material type of the wafer disk actually loaded at the current workstation.
[0048] `completed_set` is a collection of material types that have been fully mounted. For example, `{"A", "C"}`.
[0049] queue_index=158 indicates the index position of the next patch point to be executed in the task queue, for example, counting from 0 or 1.
[0050] The snapshot_save() function is a subroutine that saves the system snapshot state.
[0051] The load_snapshot() function is a subroutine that loads a snapshot state from non-volatile memory.
[0052] In existing technologies, after an abnormal restart of the equipment, it is necessary to accurately skip the materials that have already been completed, but the number of original mounting points is huge, and judging point by point is inefficient.
[0053] By introducing material-based mask vector and bitwise operations to accelerate the process, O(1) level skip judgment is achieved. The skip judgment function is: in, The original mounting point to be judged; The material type corresponding to point p; This is a mapping from material type to index, for example, A→1, B→2, C→3; This is a 3-bit binary mask vector, generated from the completed material set. For example, if A and C are already labeled, then... ; For right shift operation, & is for bitwise AND operation; if If the condition is met, skip that point; otherwise, add it to the execution queue.
[0054] By compressing the original O(n) judgment that required traversing the set into a single CPU instruction, it is suitable for scenarios with millions of mounting points and can improve recovery speed and memory efficiency. O(n) is a method for algorithm time complexity, used to describe the trend of algorithm running time or resource consumption as the input size increases in the worst case.
[0055] In this embodiment, millisecond-level power outage recovery capability is achieved, ensuring seamless reconnection to the original plan after an anomaly is resolved, eliminating duplicate mounting and redundant material changes, and significantly improving production reliability.
[0056] In some embodiments, the task queue manager is further configured to, in the case of generating the task queue, generate a preload instruction to instruct the basket to move the next wafer to be attached to the pull-in position.
[0057] Understandably, the position to be pulled in is the pre-positioned slot in the basket near the pull mechanism, which can shorten the subsequent pull stroke.
[0058] In actual execution, when the current material placement progress reaches 90%, the task queue manager parses the next material type in the task queue and queries its slot index in the basket. If the tray is not in the position to be pulled in, a movement command is sent to the basket motor controller to drive the conveyor belt to move the wafer tray to the designated pre-positioning slot, completing the physical pre-positioning.
[0059] In this embodiment, by moving the wafer disk in advance, the material change waiting time can be reduced, and the equipment cycle efficiency can be improved.
[0060] In some embodiments, the mounting equipment is provided with a visual calibration unit, and the anomaly handling module is further configured to: If the detected abnormal event is an abnormal placement offset, the vision calibration unit is invoked to perform coordinate compensation on the workstation in the work area, and the positioning parameters of the placement point are adjusted according to the calibration results. If the number of consecutive occurrences of the same type of abnormal event exceeds the threshold, the handling strategy for the corresponding material type will be updated to the historical experience database, and an alarm notification will be generated. In the event that the abnormal event is an unrecoverable hardware failure, the task queue is frozen and the execution context is saved; Upon completion of the replenishment operation, the task queue manager is triggered to re-evaluate the scheduling priority of the remaining tasks to be executed in order to update the mounting order.
[0061] Understandably, the abnormal placement offset was detected by visual inspection after placement, which manifested as the deviation between the actual placement position and the target coordinates exceeding a threshold.
[0062] The historical experience database is a local SQLite table that records the frequency of anomalies and corresponding response strategies for each material.
[0063] Unrecoverable faults include programmable logic controller (PLC) safety signals such as motor stall and vacuum failure.
[0064] In actual execution, the exception handling module triggers the high-precision camera to image the workstation reference Mark point, calculates the current workstation coordinate system offset ΔX, ΔY, and Δθ, and superimposes this compensation value into the X / Y / θ parameters of the subsequent mounting points to achieve dynamic correction.
[0065] The counter accumulates consecutive offset events for the same material. If the number of offset events exceeds 3, the "Enable High-Precision Positioning Mode" strategy is written to the database, and the operator is prompted to check the flatness of the wafer disk of the material through a pop-up window in the Human Machine Interface (HMI).
[0066] The exception handling module immediately pauses motion control, writes the current queue status, workstation materials, and completed sets to the snapshot area, and locks the task queue to prevent external modification.
[0067] After successful material replenishment, if there are high-urgency materials in the remaining tasks that are delayed, the task queue manager notifies the sequence decision engine to re-evaluate based on the latest delivery date data and fine-tune the order of subsequent materials if necessary.
[0068] In this embodiment, a closed-loop anomaly management system is implemented, encompassing perception, compensation, learning, and alarming, thereby enhancing the system's robustness and self-healing capabilities.
[0069] In some embodiments, the state storage module is coupled to the basket control system to read the material type of the wafer disk to be mounted in each slot of the basket, and to feed back the empty disk status or half disk status to the sequence decision engine before generating the mounting sequence.
[0070] It is understandable that an empty disk refers to a number of wafers ≤ 5; a half disk refers to ≤ 50% capacity.
[0071] In actual execution, the state storage module periodically reads the slot material register array of the basket PLC through the Modbus TCP communication protocol; for each material type, it counts the total remaining quantity; if a material has less than half a tray left, it marks it with the "low_stock" tag in the material demand sequence passed to the sequential decision engine.
[0072] The weighted scoring model adds an extra urgency weight to "low_stock" materials, making them more likely to be pre-scheduled and ensuring that all mounting can be completed in a single pull.
[0073] In this embodiment, it effectively prevents mid-process shutdowns and restocking due to material depletion, ensuring task continuity.
[0074] In some embodiments, when material switching is performed at the work area station, the difference in material type of the wafer disk to be mounted is compared. If the material type is the same before and after the switching, the wafer disk pulling-in and preparation work is skipped, and the wafer disk to be mounted at the work area station is called to continue the mounting process. When the material types are different before and after the switch, the mounting equipment is controlled to return the wafer disk to be mounted on the work area station to the basket, and pull the wafer disk of the target material type from the basket.
[0075] Understandably, the switching request is issued by the task queue manager and includes the target material type.
[0076] In actual execution, the controller obtains the material type of the current workstation and the target material type, and performs a string comparison.
[0077] If the comparison results are consistent, the basket pull command will not be triggered. The wafer disk pulling and preparation work will be skipped, and the material picking process will be directly entered. The wafer disk of the current workstation will be directly called to continue to perform the placement, saving 3-5 seconds of preparation time.
[0078] If there is a discrepancy, the tray return operation is performed first, the workstation is raised and lowered and the conveyor belt is reversed to return the current wafer tray to the basket, and then a new tray is pulled in from the designated slot in the basket. After vacuum adsorption and calibration are completed, the process continues.
[0079] When adjacent tasks in the task queue contain the same material, the traditional process will trigger a redundant "return-pull" action. This application achieves zero-changeover placement by defining a material continuity flag and combining it with the workstation status.
[0080] The continuity indicator for materials is: in, The material type of the wafer disk at the current workstation, sourced from the state storage module; The material type for the next task to be executed, sourced from the head of the task queue; This is a Boolean function that returns whether a valid wafer disk containing material m has been loaded at the workstation; if... If the material changing subroutine is skipped, the material picking stage will proceed directly.
[0081] In this embodiment, erroneous switching in scenarios with multiple trays of the same material is effectively identified and eliminated, unnecessary mechanical actions are avoided, and non-production time is further optimized.
[0082] This application also provides a mounting apparatus equipped with a wafer swapping optimization system as described in any of the above embodiments.
[0083] According to the mounting equipment provided in the embodiments of this application, by reusing the workstation wafer disks at the end of the previous task through state memory, combined with the round-robin scheduling and abnormal local repair mechanism, the repeated pulling of disks, invalid switching and global reordering are effectively avoided without increasing hardware costs, reducing the frequency of wafer disk switching, shortening non-production time, improving the overall efficiency and operational stability of the mounting equipment, and reducing production line stagnation and resource waste caused by material changes.
[0084] In some embodiments, the mounting equipment includes a basket for storing wafer disks to be mounted, a work area station for mounting the wafer disks to be mounted, and a vision calibration unit. The work area station is used to mount the wafer disk to be mounted. The visual calibration unit is used to perform coordinate compensation on the workstation in the work area.
[0085] This application also provides a wafer material change optimization method, applied to a mounting equipment. The mounting equipment is used to perform mounting operations on wafer disks containing multiple material types. The mounting equipment is equipped with a basket for storing the wafer disks to be mounted and a work area station for mounting the wafer disks to be mounted.
[0086] like Figure 2 As shown, the wafer swapping optimization method includes: Step 210: Using the status storage module, when the first mounting task is completed, the material type corresponding to the wafer disk to be mounted on the work station in the work area is retained as the end status information and stored. Step 220: Using the sequential decision engine, when the second mounting task is started, read the end status information, and take the material type corresponding to the end status information as the starting point, determine the scheduling priority of each task to be executed in the second mounting task according to the material rotation order in the basket and the material demand sequence of the mounting equipment, so as to generate a mounting sequence that minimizes the number of material changes according to the scheduling priority. Step 230: Using the task queue manager, the original mounting points are grouped and rearranged according to the material type based on the mounting order to generate the task queue to be executed. If the task to be executed is interrupted, the completed part is skipped according to the set of mounted materials, and the remaining tasks are executed. Step 240: During the operation of the mounting equipment, the abnormal event is detected by the abnormal event handling module. If the abnormal event is a material picking failure, a material replenishment operation is triggered at the failed point, and the execution flow of the task queue is restored after the material replenishment is completed.
[0087] According to the wafer material change optimization method provided in the embodiments of this application, by reusing the workstation wafer disk at the end of the previous task through state memory, combined with the round-robin scheduling and abnormal local repair mechanism, the repeated pulling of disks, invalid switching and global reordering are effectively avoided without increasing hardware costs, reducing the frequency of wafer disk switching, shortening non-production time, improving the overall efficiency and operational stability of the mounting equipment, and reducing production line stagnation and resource waste caused by material change.
[0088] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.
[0089] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.
[0090] Although embodiments of this application have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of this application, the scope of which is defined by the claims and their equivalents.
Claims
1. A wafer refueling optimization system, characterized in that, The system is applied to a mounting equipment for mounting wafers containing various material types. The mounting equipment is equipped with a basket for storing the wafers to be mounted and a work area station for mounting the wafers to be mounted. The system includes an interconnected state storage module, a sequence decision engine, a task queue manager, and an exception handling module. The status storage module is used to store the material type corresponding to the wafer disk to be mounted on the work area station as the end status information when the first mounting task is completed. The sequence decision engine is used to read the end status information when the second mounting task is started, and take the material type corresponding to the end status information as the starting point, determine the scheduling priority of each task to be executed in the second mounting task according to the material rotation order in the basket and the material demand sequence of the mounting equipment, so as to generate a mounting sequence that minimizes the number of material changes according to the scheduling priority. The task queue manager is used to group and rearrange the original mounting points according to the material type based on the mounting order, generate a task queue for the tasks to be executed, and skip the completed part according to the set of mounted materials and execute the remaining tasks if the tasks to be executed are interrupted. The exception handling module is used to detect abnormal events during the operation of the mounting equipment. If the abnormal event is a material picking failure, it triggers a material replenishment operation at the failed point and resumes the execution flow of the task queue after the material replenishment is completed.
2. The wafer replacement optimization system according to claim 1, characterized in that, The sequential decision engine has a built-in weighted scoring model, which is constructed based on the material changeover time factor, process compatibility factor, and order delivery urgency factor. It is used to comprehensively score the candidate mounting sequence. The material changeover time factor is calculated based on the switching distance between adjacent material types in the basket. The process compatibility factor is determined by the matching degree between the current material and the process parameters of the target station. The order delivery urgency factor is related to the ratio of the remaining processing time to the delivery deadline of the order to which the task belongs. The weighted scoring model is used to score the material change path by linearly weighting the factors according to their weight coefficients, and the scoring result is used to adjust the scheduling priority.
3. The wafer replacement optimization system according to claim 1, characterized in that, The state storage module is further used for: Once the material switching is complete, update the material type on the workstation in the work area, and trigger a status snapshot to save the current material type, the completed material set, and the task queue position when the equipment stops abnormally. The task queue manager is used to skip all original mounting points corresponding to the mounted materials according to the completed material set when the device is restarted or reset, and continue to execute the remaining tasks to be executed from the task queue position. The sequence decision engine generates the mounting order with the current material type as the starting point of the new mounting task.
4. The wafer replacement optimization system according to claim 1, characterized in that, The task queue manager is further configured to, in the case of generating the task queue, generate a preload instruction to instruct the basket to move the next wafer to be attached to the pull-in position.
5. The wafer replacement optimization system according to claim 1, characterized in that, The mounting equipment is equipped with a visual calibration unit, and the anomaly handling module is further used for: If the detected abnormal event is an abnormal placement offset, the vision calibration unit is invoked to perform coordinate compensation on the workstation in the work area, and the positioning parameters of the placement point are adjusted according to the calibration results. If the number of consecutive occurrences of the same type of abnormal event exceeds the threshold, the handling strategy for the corresponding material type will be updated to the historical experience database, and an alarm notification will be generated. In the event that the abnormal event is an unrecoverable hardware failure, the task queue is frozen and the execution context is saved; Upon completion of the replenishment operation, the task queue manager is triggered to re-evaluate the scheduling priority of the remaining tasks to be executed in order to update the mounting order.
6. The wafer replacement optimization system according to claim 1, characterized in that, The state storage module is coupled to the basket control system and is used to read the material type of the wafer disk to be mounted in each slot of the basket, and to feed back the empty disk status or half disk status to the sequence decision engine before generating the mounting sequence.
7. The wafer replacement optimization system according to claim 1, characterized in that, When material switching occurs at the workstation in the work area, the difference in material type of the wafer disk to be mounted is compared. If the material type is the same before and after the switching, the wafer disk pulling in and preparation work is skipped, and the wafer disk to be mounted at the workstation in the work area is called to continue the mounting process. When the material types are different before and after the switch, the mounting equipment is controlled to return the wafer disk to be mounted on the work area station to the basket, and pull the wafer disk of the target material type from the basket.
8. A mounting device, characterized in that, It is equipped with a wafer refueling optimization system as described in any one of claims 1-7.
9. The mounting equipment according to claim 8, characterized in that, The mounting equipment is equipped with a basket for storing wafer disks to be mounted, a work area station for mounting the wafer disks to be mounted, and a vision calibration unit. The work area station is used to mount the wafer disk to be mounted. The visual calibration unit is used to perform coordinate compensation on the workstation in the work area.
10. A wafer replacement optimization method, characterized in that, This is applied to a mounting equipment, which is used to mount wafer disks containing various material types. The mounting equipment is equipped with a basket for storing the wafer disks to be mounted and a work area station for mounting the wafer disks to be mounted. The method includes: By using the status storage module, when the first mounting task is completed, the material type corresponding to the wafer disk to be mounted on the work area station is retained as the end status information and stored. Using the sequential decision engine, when the second mounting task is started, the end status information is read, and the scheduling priority of each task to be executed in the second mounting task is determined according to the material type corresponding to the end status information, the material rotation order in the basket and the material demand sequence of the mounting equipment, so as to generate a mounting sequence that minimizes the number of material changes according to the scheduling priority. The task queue manager regroups and rearranges the original mounting points according to the material type based on the mounting order, generating a task queue for the tasks to be executed. If the tasks to be executed are interrupted, the completed part is skipped based on the set of mounted materials, and the remaining tasks are executed. The exception handling module detects abnormal events during the operation of the mounting equipment. If the abnormal event is a material picking failure, a material replenishment operation is triggered at the failed point, and the execution flow of the task queue is restored after the material replenishment is completed.