A paperless system for a silicon wafer cutting workshop based on the internet of things and an application method thereof

Abstract

The application provides a paperless system for a silicon wafer cutting workshop based on the Internet of Things and an application method thereof. The system comprises a dynamic RFID binding module, an automatic work order circulation module, an exception processing and alarm module, a multi-specification silicon rod mixed line production scheduling module and a data analysis and tracing module. The automatic work order circulation module comprises a wafer bonding process module, a wafer cutting process module, a wafer debonding process module and a wafer arranging process module. The method adopts the paperless system for the silicon wafer cutting workshop based on the Internet of Things, and comprises a wafer bonding step, a wafer cutting step, a wafer debonding step and a wafer arranging step. Through dynamic RFID binding, AGV cooperative scheduling, AGV double-checking mechanism, wafer arranging priority algorithm and multi-objective optimization algorithm, the application realizes digital tracking and flexible production of the whole wafer cutting process.

Description

Technical Field

[0001] This invention relates to the fields of intelligent manufacturing and the Internet of Things, specifically to a paperless system for a silicon wafer cutting workshop and its application method. Background Technology

[0002] Traditional silicon wafer cutting workshops use a paper-based work order system: each process (die bonding → slicing → debinding → wafer stacking) requires manual verification of the paper work order against the physical information, and data updates to the MES system are completed through manual registration, transmission, and scanning. Work order circulation relies on manual transfer of physical documents, and each process node requires manual entry of key information such as the die number and any exceptions. The specific process includes: Crystal bonding process: After manually assembling the blades, print the work order and place it at the opening of the crystal holder.

[0003] Slicing process: Manually input the blade number into the slicer, and the work order moves with the degumming cart.

[0004] Degumming process: Manually register abnormal information and transfer work orders.

[0005] Film arrangement process: After manually verifying the consistency between the work order and the actual product, the information is entered into the system.

[0006] Therefore, although existing technologies (such as MES systems) have achieved partial digitization, they still have the following drawbacks: Information silos: Data updates are not synchronized across different processes, resulting in a lag between the system and actual production progress.

[0007] High human error rate: misplacement of orders leads to mixed materials (occurring an average of 1.2 times per week), and failure to process orders causes inventory deviations in the system (requiring an average of 15 manual corrections per month).

[0008] Efficiency bottleneck: Each silicon wafer requires manual verification of the work order 3 times (total time 9-12 minutes).

[0009] Traceability difficulties: Anomaly tracing relies on paper records (locating problems takes an average of 1 hour), and complete production records cannot be traced after paper documents are damaged or lost.

[0010] Insufficient flexibility: Unable to respond in real time to the mixed production needs of multiple specifications of silicon rods (8-9 types). Summary of the Invention

[0011] To address the various shortcomings of existing technologies, this invention provides a paperless silicon wafer cutting workshop system based on the Internet of Things and its application method. Through dynamic RFID binding, AGV collaborative scheduling, AGV dual verification mechanism, wafer arrangement priority algorithm and multi-objective optimization algorithm, it realizes digital tracking and flexible production of the entire silicon wafer cutting process.

[0012] The present invention achieves the above objectives through the following technical solutions: A paperless silicon wafer dicing workshop system based on the Internet of Things includes: The dynamic RFID binding module is used to temporarily and dynamically bind the tool number to the physical carrier. During the degumming process, the RFID reader on the AGV reads and transfers the tool number information to the RFID card on the degumming frame, thereby realizing dynamic tracking and management of the tool number during the degumming process. The automated work order workflow module is integrated into the intelligent warehouse management system and the manufacturing execution system. It uses automated equipment to automatically transfer and update work order information between processes, thereby eliminating the need for manual paper document transfer. The anomaly handling and alarm module adopts a three-level alarm system and work order state machine linkage control. When inconsistent tool number information, process error or other abnormal situation is detected, an alarm is immediately triggered and the relevant equipment is automatically stopped. The multi-specification silicon rod mixed production scheduling module constructs a multi-objective optimization scheduling model based on the first 9 digits of the tool number feature code to realize the concurrent processing and automated scheduling of multi-specification silicon rods. The data analysis and traceability module is used for the storage and management of the entire electronic resume process, and supports setting a data traceability period.

[0013] According to the present invention, a paperless silicon wafer cutting workshop system based on the Internet of Things is provided, wherein the automated work order transfer module includes: Crystal bonding process module: used to scan the QR code of the crystal rod to read the tool number, and transmit the tool number information to the storage location in the crystal bonding workshop through the intelligent warehouse management system, and then the AGV completes the physical feeding; Slicing process module: The knife number is verified when the AGV picks up the material, and the knife number is automatically transmitted to the intelligent warehouse management system and manufacturing execution system during the loading and unloading process of the slicing machine; Degumming process module: The RFID reader inside the degumming machine reads the knife number information on the RFID card of the degumming rack and uploads it to the manufacturing execution system for order processing; The film arrangement process module uses the RFID reader on the film arrangement machine to read the knife number and related information on the RFID card of the film retrieval cart and displays it on the screen.

[0014] According to the present invention, a paperless silicon wafer cutting workshop system based on the Internet of Things includes an AGV dual verification mechanism in the automated work order transfer module, comprising: During the material handling and feeding process, the AGV first scans the QR code information attached to the surface of the silicon rod using its built-in scanning device to obtain the corresponding tool number data of the current silicon rod. At the same time, it receives the tool number instructions for the batch of silicon rods pushed in real time by the intelligent warehouse management system through the wireless communication module. The tool number obtained by scanning is compared and verified with the tool number issued by the intelligent warehouse management system. The AGV will only perform material handling operations if the two match completely. If there is a difference in the comparison result, the abnormal interruption mechanism will be triggered immediately, the current operation will be stopped, and the abnormal event will be reported to the system.

[0015] An application method for an IoT-based paperless silicon wafer dicing workshop system, comprising the following steps: Die bonding step: After curing, the silicon rod is output from the die bonding line. The barcode of the silicon rod is scanned by a barcode scanner to obtain the tool number information, which is then transmitted to the intelligent warehouse management system. The intelligent warehouse management system allocates target storage locations according to the production plan. After the AGV carries the silicon rod to the designated storage location, the intelligent warehouse management system unlocks the tool number. When the slicing workshop storage location issues a material request, the intelligent warehouse management system reallocates the silicon rod from the die bonding workshop storage location to the slicing process storage location. The AGV carries the silicon rod to perform the feeding task and simultaneously transmits the tool number to the intelligent warehouse management system. The manufacturing execution system automatically updates the process status to slicing. Slicing steps: When the loading / unloading AGV retrieves materials from the storage location, the barcode scanning device verifies the consistency between the silicon rod knife number at the storage location and the system command, and executes the material retrieval or alarm termination process; when the slicing machine loads / unloads materials, the AGV action is triggered through the control terminal, and the knife number is transmitted to the manufacturing execution system; after slicing is completed, the loading / unloading AGV unloads the material and stores the knife number in the RFID card, while the manufacturing execution system automatically updates the process status to pre-cleaning; when the loading / unloading AGV transports the material to the degumming area, it binds the knife number and degumming rack RFID card information and uploads it to the intelligent warehouse management system; Degumming step: The degumming machine verifies the consistency between the knife number information in the RFID card of the degumming rack and the process status of the manufacturing execution system through the built-in RFID reader; if the process status corresponding to the knife number is not pre-cleaning, the degumming machine triggers an alarm and stops running; after degumming is completed, the knife number is obtained by scanning with a mobile terminal and bound to the RFID card of the sheet retrieval vehicle. After the material information is entered, the material is put into the warehouse, and the AGV carrying the material pulls the material to the sheet arrangement process according to the instructions of the intelligent warehouse management system. The stacking process is as follows: The intelligent warehouse management system transmits the degummed material information to the stacking buffer storage location and allocates materials to the buffer storage location according to the equipment's processing capacity; the AGV carries the materials to the stacking machine, and the stacking machine reads the knife number, precision information and material status from the RFID card of the stacking vehicle through the RFID card reader. After the operator verifies the information, the knife number is uploaded to the manufacturing execution system. In case of abnormality, manual intervention is performed.

[0016] According to the present invention, an application method for a paperless silicon wafer cutting workshop based on the Internet of Things is provided, wherein in the die bonding step: The scanning device is a fixed laser barcode scanner, which is installed on a robotic arm or fixed bracket at the end of the die bonding line. The scanning angle is adjustable to accommodate silicon rods of different specifications. When allocating target storage locations, the intelligent warehouse management system prioritizes storage locations that are physically closest to the current process and are currently available. After the AGV carries silicon rods to the target storage location, it sends a signal of arrival to the intelligent warehouse management system via RFID tags or photoelectric sensors on the target storage location. The system then unlocks the tool number and updates the storage location status to occupied.

[0017] According to the present invention, an application method for a paperless silicon wafer dicing workshop based on the Internet of Things is provided, wherein the dicing step includes: The barcode scanning device for the loading and unloading AGV is a handheld or vehicle-mounted QR code scanner; When the slicer is being loaded, a loading request is triggered via a touch screen or voice control terminal. After receiving the instruction, the loading and unloading AGV completes the loading action within a predetermined time and sends a completion signal back to the manufacturing execution system. When the AGV transports materials to the degumming cart area, the RFID card information of the degumming rack is read by the RFID card reader installed at the bottom of the AGV.

[0018] According to the present invention, in the application method of a paperless silicon wafer cutting workshop based on the Internet of Things, the degumming step includes: The degumming machine has a built-in RFID reader, which is either a high-frequency or ultra-high-frequency reader. When reading the RFID card of the degumming rack, it automatically verifies the matching between the knife number in the card and the current process status of the manufacturing execution system. If the process status is not pre-cleaning, the degumming machine will immediately stop running and issue an audible and visual alarm. The mobile terminal is a PDA device that supports NFC or QR code scanning. After scanning the edge skin QR code or the glue removal rack RFID card, the material information entry interface will automatically pop up. The content to be entered includes the warehouse location number, material status, zone and operator employee number. When the AGV pulls materials, the intelligent warehouse management system dynamically adjusts the allocation order of the buffer storage locations according to the priority of the stacking process, and materials with higher priority are allocated to the storage locations closer to the stacking machine.

[0019] According to the present invention, an application method for a paperless silicon wafer cutting workshop based on the Internet of Things is provided, wherein the wafer arrangement step includes: The RFID reader of the film sorting machine is installed next to the lifting rail of the film retrieval vehicle. After successful reading, the knife number, precision information and material status are displayed on the film sorting machine's operation interface. When verifying the edge-cutting information against the system information, if there are discrepancies after K consecutive verifications, the system will automatically lock the current retrieval vehicle and push an abnormal notification to the workshop supervisor's mobile terminal. When allocating materials in the cache storage location, the intelligent warehouse management system calculates the optimal allocation scheme based on the processing capacity of the stacking machine.

[0020] According to the present invention, an application method for a paperless silicon wafer dicing workshop based on the Internet of Things is provided, wherein the intelligent warehouse management system dynamically adjusts the allocation order of buffer storage locations according to the priority of the wafer dicing process, including: The intelligent warehouse management system embeds a priority calculation algorithm to generate a priority index based on at least one of the following factors: processing urgency, material correlation, and equipment load balancing rate. The processing urgency is calculated by the difference between the process deadline pushed by the manufacturing execution system in real time and the current time; the smaller the difference, the higher the priority. Material correlation is determined based on the upstream and downstream dependencies of materials with the same tool number in the scheduling process; materials with higher dependencies have higher priority. The equipment load balancing rate is calculated by the ratio of the real-time processing queue length of the scheduling machine to the theoretical maximum queue length; when the ratio is below a threshold, the priority of the corresponding cache location is increased. The intelligent warehouse management system pre-stores 3D map data of the workshop, marking the coordinates of all cache locations and the coordinates of the stacking machine, and uses Euclidean distance algorithm or A... The path planning algorithm calculates the physical distance between each cache location and the chip setter, and generates a distance-sorted list; The system sorts materials in descending order according to priority index and generates a priority queue; for each material in the priority queue, it selects the nearest available storage location to the stacker from the distance sorting list as the target storage location; The intelligent warehouse management system packages target storage location information, material tool numbers, and transportation route planning data into task instructions, which are then sent to the central controller carrying the AGV via a wireless communication protocol. After receiving the instruction, the AGV carrying the vehicle travels to the target buffer storage location using laser navigation or vision navigation technology, grabs the material using a robotic arm or pallet lifting device, and transports it to the designated loading position of the stacking machine.

[0021] According to the present invention, an application method for a paperless silicon wafer cutting workshop based on the Internet of Things is provided. When allocating materials in the buffer storage location, the intelligent warehouse management system calculates the optimal allocation scheme based on the processing capacity of the wafer stacking machine, including: The intelligent warehouse management system obtains the processing capacity parameters of the stacking machine in real time from the manufacturing execution system or the local controller of the stacking machine, including at least: the maximum number of stacks N processed in a single operation. max Processing speed V process Current processing queue length Q current Theoretical maximum queue length Qmax ; Based on the production plan for the film arrangement process and combined with historical processing data, predict the material demand D within the future time window ΔT. predict ; Scan the real-time status of all cache locations and obtain the data: remaining capacity C available Material compatibility ij Physical distance Dist i ; An optimization model is constructed based on linear programming or genetic algorithms, with the objective function being to minimize the total transportation time T. total And maximize the load balancing rate of the film setter L balance The constraints must include at least: processing capacity constraints, storage space capacity constraints, and time window constraints. After solving the optimization model, the optimal allocation scheme is generated, including: Target storage location list: The optimal cache storage location number corresponding to each material; Allocation quantity: The quantity of materials that need to be stored in each storage location; Transportation sequence: the priority sequence of tasks performed by the AGVs carrying the AGVs; When carrying AGVs to perform transportation tasks, the system provides real-time feedback on changes in warehouse location status. If there is a sudden change in the processing speed of the chip stacking machine or the queue length, the optimal allocation scheme is immediately recalculated, and an update instruction is pushed to the AGV via wireless communication protocol.

[0022] Therefore, compared with the existing technology, the paperless silicon wafer cutting workshop system and its application method based on the Internet of Things proposed in this invention have the following beneficial effects: 1. This invention utilizes dynamic RFID binding and AGV collaborative scheduling to replace manual paper-based work orders, achieving fully automated material flow and data updates between processes and eliminating manual operation waiting time. The system uses the Internet of Things to collect real-time equipment status and material information, dynamically adjusts production plans, and quickly responds to the needs of multi-specification silicon rod mixed-line production, reducing equipment idleness and process downtime.

[0023] 2. This invention employs a dual-verification error-proofing mechanism. In critical stages such as AGV material handling and wafer stacker loading, it verifies the silicon rod's QR code against the system's tool number, combined with RFID carrier binding, ensuring strict matching of materials and processes and eliminating the risk of material mixing. This invention constructs a three-tiered alarm system: equipment layer (AGV / degummer), system layer (IWMS), and management layer (production supervisor). Abnormal events automatically trigger shutdown, alarm, and handling procedures, blocking the error propagation chain.

[0024] 3. This invention supports the concurrent processing of 8-9 silicon rods with different diameters and cutting thicknesses. Through a multi-objective optimization scheduling model, it dynamically allocates storage space and equipment resources to meet the flexible production needs of small batches and multiple varieties. The temporary binding relationship between the cutter number and physical carriers such as the wafer tray and debinding rack is automatically updated as the process flows, without the need for manual parameter adjustment, adapting to scenarios of rapid switching of production specifications.

[0025] 4. This invention integrates barcode scanning during the die bonding process with uploading data to the MES system during the wafer stacking process. All operation records (tool number, time, equipment, and anomaly information) are stored in real-time on an industrial server, forming an immutable electronic traceability chain. A 3D model of the workshop is constructed using digital twin technology, mapping equipment status, material location, and production progress in real time. Managers can remotely monitor and make decisions, improving production transparency.

[0026] 5. This invention replaces paper printing, transmission, and verification with electronic work orders, reducing reliance on manpower and the rate of human error. The three-level alarm system intercepts over 90% of abnormal events, requiring only a small number of technical personnel, freeing up human resources for high-value processes.

[0027] 6. Automated order processing and material flow provide data interfaces for "lights-out factories," supporting full-process automation upgrades; multi-objective optimization scheduling models balance equipment load and improve overall equipment efficiency.

[0028] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments. Attached Figure Description

[0029] Figure 1 This is a schematic diagram of an embodiment of a paperless silicon wafer cutting workshop system based on the Internet of Things according to the present invention.

[0030] Figure 2 This is a flowchart of an embodiment of an application method for a paperless silicon wafer cutting workshop based on the Internet of Things according to the present invention.

[0031] Figure 3 This is a flowchart illustrating an embodiment of an application method for a paperless silicon wafer cutting workshop system based on the Internet of Things according to the present invention. Detailed Implementation

[0032] 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 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.

[0033] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0034] An embodiment of a paperless system for a silicon wafer dicing workshop based on the Internet of Things See Figure 1 This embodiment provides a paperless silicon wafer cutting workshop system based on the Internet of Things, including: The dynamic RFID binding module is used to temporarily and dynamically bind the tool number to the physical carrier. During the degumming process, the RFID reader on the AGV reads and transfers the tool number information to the RFID card on the degumming frame, thereby realizing dynamic tracking and management of the tool number during the degumming process. The automated work order workflow module is integrated into the Intelligent Warehouse Management System (IWMS) and Manufacturing Execution System (MES). It utilizes automated equipment to automatically transfer and update work order information between processes, thereby eliminating the need for manual paper document transfer. The anomaly handling and alarm module adopts a three-level alarm system and work order state machine linkage control. When inconsistent tool number information, process error or other abnormal situation is detected, an alarm is immediately triggered and the relevant equipment is automatically stopped. The multi-specification silicon rod mixed production scheduling module constructs a multi-objective optimization scheduling model based on the first 9 digits of the tool number feature code to realize the concurrent processing and automated scheduling of multi-specification silicon rods. The data analysis and traceability module is used for the storage and management of the entire electronic resume process, and supports setting a data traceability period.

[0035] In this embodiment, the automated work order workflow module includes: Crystal bonding process module: used to scan the QR code of the crystal rod to read the tool number, and transmit the tool number information to the storage location in the crystal bonding workshop through the intelligent warehouse management system, and then the AGV completes the physical feeding; Slicing process module: The knife number is verified when the AGV picks up the material, and the knife number is automatically transmitted to the intelligent warehouse management system and manufacturing execution system during the loading and unloading process of the slicing machine; Degumming process module: The RFID reader inside the degumming machine reads the knife number information on the RFID card of the degumming rack and uploads it to the manufacturing execution system for order processing; The film arrangement process module uses the RFID reader on the film arrangement machine to read the knife number and related information on the RFID card of the film retrieval cart and displays it on the screen.

[0036] In this embodiment, the AGV dual verification mechanism in the automated work order workflow module includes: During the material handling and feeding process, the AGV first scans the QR code information attached to the surface of the silicon rod using its built-in scanning device to obtain the corresponding tool number data of the current silicon rod. At the same time, it receives the tool number instructions for the batch of silicon rods pushed in real time by the intelligent warehouse management system through the wireless communication module. The tool number obtained by scanning is compared and verified with the tool number issued by the intelligent warehouse management system. The AGV will only perform material handling operations if the two match completely. If there is a difference in the comparison result, the abnormal interruption mechanism will be triggered immediately, the current operation will be stopped, and the abnormal event will be reported to the system.

[0037] Furthermore, during the degumming process, the RFID card information of the degumming rack is read by the RFID card reader on the AGV, and the currently processed knife number is bound to the RFID card, realizing seamless transfer and tracking of knife number information in the degumming process.

[0038] Furthermore, the multi-specification silicon rod mixed-line production scheduling module dynamically adjusts the material allocation order of the wafer racking buffer location based on the processing capacity of the wafer racking machine, the current inventory status, and the production plan, thereby optimizing the production process and improving overall production efficiency.

[0039] An Example of an Application Method for a Paperless System in a Silicon Wafer Cutting Workshop Based on the Internet of Things like Figure 2 and Figure 3 As shown in this embodiment, an application method for an IoT-based paperless silicon wafer cutting workshop system is provided. This method utilizes the aforementioned IoT-based paperless silicon wafer cutting workshop system and includes: Die bonding step: After curing, the silicon rod is output from the die bonding line. The barcode of the silicon rod is scanned by a barcode scanner to obtain the tool number information, which is then transmitted to the intelligent warehouse management system. The intelligent warehouse management system allocates target storage locations according to the production plan. After the AGV carries the silicon rod to the designated storage location, the intelligent warehouse management system unlocks the tool number. When the slicing workshop storage location issues a material request, the intelligent warehouse management system reallocates the silicon rod from the die bonding workshop storage location to the slicing process storage location. The AGV carries the silicon rod to perform the feeding task and simultaneously transmits the tool number to the intelligent warehouse management system. The manufacturing execution system automatically updates the process status to slicing. Slicing steps: When the loading / unloading AGV retrieves materials from the storage location, the barcode scanning device verifies the consistency between the silicon rod knife number at the storage location and the system command, and executes the material retrieval or alarm termination process; when the slicing machine loads / unloads materials, the AGV action is triggered through the control terminal, and the knife number is transmitted to the manufacturing execution system; after slicing is completed, the loading / unloading AGV unloads the material and stores the knife number in the RFID card, while the manufacturing execution system automatically updates the process status to pre-cleaning; when the loading / unloading AGV transports the material to the degumming area, it binds the knife number and degumming rack RFID card information and uploads it to the intelligent warehouse management system; Degumming step: The degumming machine verifies the consistency between the knife number information in the RFID card of the degumming rack and the process status of the manufacturing execution system through the built-in RFID reader; if the process status corresponding to the knife number is not pre-cleaning, the degumming machine triggers an alarm and stops running; after degumming is completed, the knife number is obtained by scanning with a mobile terminal and bound to the RFID card of the sheet retrieval vehicle. After the material information is entered, the material is put into the warehouse, and the AGV carrying the material pulls the material to the sheet arrangement process according to the instructions of the intelligent warehouse management system. The stacking process is as follows: The intelligent warehouse management system transmits the degummed material information to the stacking buffer storage location and allocates materials to the buffer storage location according to the equipment's processing capacity; the AGV carries the materials to the stacking machine, and the stacking machine reads the knife number, precision information and material status from the RFID card of the stacking vehicle through the RFID card reader. After the operator verifies the information, the knife number is uploaded to the manufacturing execution system. In case of abnormality, manual intervention is performed.

[0040] It is evident that the above method, through the collaboration of IoT devices and systems, achieves paperless, automated, and traceable management of the entire silicon wafer cutting process.

[0041] In this embodiment, the target storage location refers to the final storage location pre-planned or dynamically allocated by the system when the AGV (Automated Guided Vehicle) is performing material handling tasks. This location is automatically generated by the IWMS (Intelligent Warehouse Management System) based on factors such as production plans, material attributes, and process requirements, and is synchronized in real time to the AGV's control system via wireless communication technologies (such as Wi-Fi and 5G).

[0042] Assuming a silicon ingot is completed in the die bonding process, the IWMS system assigns it the target storage location as "Die Bonding Workshop - Shelf A - Layer 2 - Position 12". A carrying AGV navigates to this location, places the silicon ingot, and the system automatically records "Tool No. X has been stored in A-2-12" and unlocks the tool number for use in the next process. If the slicing workshop needs material, the IWMS system updates the target storage location to "Slicing Workshop - Shelf B - Layer 3 - Position 8", and the AGV replans its route to complete the material delivery.

[0043] In this embodiment, in the above-mentioned crystal bonding step: The scanning device is a fixed laser barcode scanner, which is installed on a robotic arm or fixed bracket at the end of the die bonding line. The scanning angle is adjustable to accommodate silicon rods of different specifications. When allocating target storage locations, the intelligent warehouse management system prioritizes storage locations that are physically closest to the current process and are currently available. After the AGV carries silicon rods to the target storage location, it sends a signal of arrival to the intelligent warehouse management system via RFID tags or photoelectric sensors on the target storage location. The system then unlocks the tool number and updates the storage location status to occupied.

[0044] In this embodiment, in the above slicing step: The barcode scanning device for the loading and unloading AGV is a handheld or vehicle-mounted QR code scanner; When the slicer is being loaded, a loading request is triggered via a touch screen or voice control terminal. After receiving the instruction, the loading and unloading AGV completes the loading action within a predetermined time and sends a completion signal back to the manufacturing execution system. When the AGV transports materials to the degumming cart area, the RFID card information of the degumming rack is read by the RFID card reader installed at the bottom of the AGV.

[0045] In this embodiment, in the above degumming step: The degumming machine has a built-in RFID reader, which is either a high-frequency or ultra-high-frequency reader. When reading the RFID card of the degumming rack, it automatically verifies the matching between the knife number in the card and the current process status of the manufacturing execution system. If the process status is not pre-cleaning, the degumming machine will immediately stop running and issue an audible and visual alarm. The mobile terminal is a PDA device that supports NFC or QR code scanning. After scanning the edge skin QR code or the glue removal rack RFID card, the material information entry interface will automatically pop up. The content to be entered includes the warehouse location number, material status, zone and operator employee number. When the AGV pulls materials, the intelligent warehouse management system dynamically adjusts the allocation order of the buffer storage locations according to the priority of the stacking process, and materials with higher priority are allocated to the storage locations closer to the stacking machine.

[0046] In this embodiment, in the above-mentioned film arrangement step: The RFID reader of the film sorting machine is installed next to the lifting rail of the film retrieval vehicle. After successful reading, the knife number, precision information and material status are displayed on the film sorting machine's operation interface. When verifying the edge-cutting information against the system information, if there are discrepancies after K consecutive verifications, the system will automatically lock the current retrieval vehicle and push an abnormal notification to the workshop supervisor's mobile terminal. When allocating materials in the cache storage location, the intelligent warehouse management system calculates the optimal allocation scheme based on the processing capacity of the stacking machine.

[0047] In practical applications, the method for implementing a paperless silicon wafer cutting workshop system based on the Internet of Things provided in this embodiment includes the following steps: Die bonding process: After the cured silicon ingots come off the die bonding line, a barcode scanner at the end of the line scans the QR code on the ingot to read the tool number, which is then transmitted in the IWMS system to the die bonding workshop storage location to be delivered by the AGV. After the AGV delivers the ingot to the storage location, the IWMS system unlocks the tool number. When the wafer slicing workshop requests ingots, the IWMS allocates the silicon ingots from the die bonding workshop to the wafer slicing material request location. The AGV delivers the ingot and transmits the tool number, while the MES automatically processes the order and transfers the process to the wafer slicing location.

[0048] Slicing Process: The loading / unloading AGV retrieves materials from the storage location, scans the silicon rod's QR code to read the blade number, and verifies it against the blade number sent by the IWMS system. If they match, the material is retrieved; otherwise, an alarm is triggered. When the slicing machine needs to load a rod, the employee clicks the "Request Loading" button, and the loading / unloading AGV loads the rod and transmits the blade number. After slicing is complete, the employee clicks the "Request Unloading" button, and the loading / unloading AGV unloads the material and stores the blade number. The MES automatically processes the order and moves the process to pre-cleaning (degumming). The loading / unloading AGV then unloads the material from the degumming cart area. The RFID reader on the AGV reads the RFID card on the degumming rack, transmits the card number and blade number to the IWMS system, and stores the blade number in the RFID card.

[0049] De-adhesion process: De-adhesion workers push the de-adhesion cart into the de-adhesion machine. The RFID reader inside the de-adhesion machine reads the RFID card and uploads the bound cutter number to the MES for order processing. If the read cutter number is not in the pre-cleaning process in the MES, the de-adhesion machine alarms and stops working. After the silicon wafers are de-adheded, the wafer retrieval workers use a PDA to scan the edge skin QR code or the de-adhesion rack RFID card to read the cutter number, bind it to the wafer retrieval cart RFID card, and fill in information such as storage location, normal material / abnormal material, and A / B zone for warehousing and material issuance, waiting for the AGV to pick up the materials and move them to the wafer rack.

[0050] Wafer stacking process: After the debonding staff receives and issues the materials, the AGV carrying the wafers receives the task and pulls the materials to the wafer stacking buffer. The IWMS system transmits the debonding and warehousing information to the wafer stacking buffer location. After the wafer retrieval cart of the wafer stacking machine is emptied, the IWMS allocates materials to the wafer stacking buffer location according to the processing capacity of the wafer stacking machine. The AGV carrying the wafers delivers the corresponding materials to the wafer stacking machine. The RFID reader on the wafer stacking machine reads the RFID card of the retrieval cart and displays the cutter number, precision marking, normal / abnormal information, etc. on the screen. The wafer stacking staff checks whether the edge precision marking matches the system. If they match, the cutter number is uploaded to MES; otherwise, it is handled manually.

[0051] Specifically, the intelligent warehouse management system dynamically adjusts the allocation order of cache locations based on the priority of the scheduling process, including: The intelligent warehouse management system has an embedded priority calculation algorithm that generates a priority index based on at least one of the following factors: the urgency of the scheduling process, material correlation, and equipment load balancing rate. Among them, the processing urgency is calculated by the difference between the process deadline pushed by the manufacturing execution system in real time and the current time. The smaller the difference, the higher the priority. The material correlation is determined by the upstream and downstream dependencies of the same tool number material in the sheet-laying process. Materials with high dependencies have higher priority. The equipment load balancing rate is calculated by the ratio of the real-time processing queue length of the sheet-laying machine to the theoretical maximum queue length. When the ratio is lower than the threshold, the priority of the corresponding cache location is increased. The intelligent warehouse management system pre-stores 3D map data of the workshop, marking the coordinates of all cache locations and the coordinates of the stacking machine, and uses Euclidean distance algorithm or A... The path planning algorithm calculates the physical distance between each cache location and the chip setter, and generates a distance-sorted list; The system sorts materials in descending order according to priority index and generates a priority queue. For each material in the priority queue, the system selects the nearest available storage location to the stacker from the distance sorting list as the target storage location. If the material with the highest priority corresponds to multiple equally spaced available storage locations, the system further performs secondary screening based on environmental parameters such as temperature, humidity, and vibration of the storage location or its historical usage frequency. The intelligent warehouse management system packages target storage location information, material tool numbers, and transportation route planning data into task instructions, and sends them to the central controller carrying the AGV via wireless communication protocols (such as Wi-Fi 6 and 5G). After receiving the instruction, the AGV carrying the backpack travels to the target buffer storage location using laser navigation or vision navigation technology, grabs the material using a robotic arm or pallet lifting device, and transports it to the designated loading position of the stacking machine; During transportation, the AGV feeds back its location information to the intelligent warehouse management system in real time. The system then dynamically updates the status of the cache storage location (occupied / idle) and the processing queue progress of the chip stacking machine.

[0052] As can be seen, the above method achieves precise matching of material allocation with equipment load and process urgency through multi-dimensional priority calculation and dynamic path planning, which significantly improves the efficiency and flexible production capacity of the film arrangement process.

[0053] Specifically, when allocating materials to buffer storage locations, the intelligent warehouse management system calculates the optimal allocation scheme based on the processing capacity of the stacking machine, including: The intelligent warehouse management system obtains the processing capacity parameters of the stacking machine in real time from the Manufacturing Execution System (MES) or the local controller of the stacking machine, including: Maximum number of pieces processed in a single batch N max The maximum number of silicon wafers that the wafer stacker can process in a single feeding operation; Processing speed V process The number of scheduling operations that can be completed per unit of time (e.g., per hour); Current processing queue length Q current : The number of material batches currently awaiting processing by the sheet feeder; Theoretical maximum queue length Q max The maximum load queue length of the slicer while ensuring processing quality; Based on the production plan for the film arrangement process and combined with historical processing data, predict the material demand D within the future time window ΔT. predict The prediction algorithm uses either a weighted moving average method or a long short-term memory network (LSTM), with the weight coefficients dynamically adjusted based on the material type and the urgency of the process. Scan the real-time status of all cache locations to obtain the following data: Remaining capacity C available The current quantity of materials that each cache location can hold; Material compatibility ij : Compatibility indicator between storage location i and material j (1=compatible, 0=incompatible), determined based on material size, material, and process requirements; Physical distance Dist i The straight-line distance or path planning distance between storage location i and the stacking machine; An optimization model is constructed based on linear programming or genetic algorithms, with the objective function being to minimize the total transportation time T. total And maximize the load balancing rate of the film setter L balance The constraints include: Processing capacity constraint: The total amount of material ΣMj allocated to the stacking machine shall not exceed the product of its maximum number of stacks processed in a single run and the remaining queue capacity, expressed as:

[0054] Storage location capacity constraint: The quantity M of materials allocated to storage location i i Not exceeding its remaining capacity, expressed as:

[0055] Time window constraint: All materials must be allocated within the predicted time window ΔT to avoid the stacking machine stopping to wait for materials; After solving the optimization model, the optimal allocation scheme is generated, including: Target storage location list: The optimal cache storage location number corresponding to each material; Allocation quantity: The quantity of materials that need to be stored in each storage location; Transportation sequence: the priority sequence of tasks performed by the AGVs carrying the AGVs; When carrying AGVs to perform transportation tasks, the system should provide real-time feedback on the changes in the storage location status (such as material loading / unloading completion). If the processing speed of the stacking machine or the queue length changes abruptly (such as equipment failure or urgent order insertion), the system immediately recalculates the optimal allocation scheme and pushes update instructions to the AGV via wireless communication protocols (such as MQTT, OPC UA). The system records data such as the actual transportation time and stacking machine utilization rate for each allocation to optimize the parameters of the subsequent prediction model.

[0056] As can be seen, this embodiment achieves efficient matching between cache storage locations and the stacking process by collecting real-time processing capacity parameters of the stacking machine, predicting material demand, dynamically evaluating the status of storage locations, and generating allocation schemes based on optimization algorithms, thereby significantly improving the overall production efficiency of the workshop.

[0057] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0058] The above embodiments are merely preferred embodiments of the present invention and should not be construed as limiting the scope of protection of the present invention. Any non-substantial changes and substitutions made by those skilled in the art based on the present invention shall fall within the scope of protection claimed by the present invention.

Claims

1. An application method for a paperless system in a silicon wafer cutting workshop based on the Internet of Things, characterized in that, The system includes: The dynamic RFID binding module is used to temporarily and dynamically bind the tool number to the physical carrier. During the degumming process, the RFID reader on the AGV reads and transfers the tool number information to the RFID card on the degumming frame, thereby realizing dynamic tracking and management of the tool number during the degumming process. The automated work order workflow module is integrated into the intelligent warehouse management system and the manufacturing execution system. It uses automated equipment to automatically transfer and update work order information between processes, thereby eliminating the need for manual paper document transfer. The anomaly handling and alarm module adopts a three-level alarm system and work order state machine linkage control. When inconsistent tool number information, process error or other abnormal situation is detected, an alarm is immediately triggered and the relevant equipment is automatically stopped. The multi-specification silicon rod mixed production scheduling module constructs a multi-objective optimization scheduling model based on the first 9 digits of the tool number feature code to realize the concurrent processing and automated scheduling of multi-specification silicon rods. The data analysis and traceability module is used for the storage and management of the entire electronic resume, and supports setting a data traceability period. The application method includes: Die bonding step: After curing, the silicon rod is output from the die bonding line. The barcode of the silicon rod is scanned by a barcode scanner to obtain the tool number information, which is then transmitted to the intelligent warehouse management system. The intelligent warehouse management system allocates target storage locations according to the production plan. After the AGV carries the silicon rod to the designated storage location, the intelligent warehouse management system unlocks the tool number. When the slicing workshop storage location issues a material request, the intelligent warehouse management system reallocates the silicon rod from the die bonding workshop storage location to the slicing process storage location. The AGV carries the silicon rod to perform the feeding task and simultaneously transmits the tool number to the intelligent warehouse management system. The manufacturing execution system automatically updates the process status to slicing. Slicing steps: When the loading / unloading AGV retrieves materials from the storage location, the barcode scanning device verifies the consistency between the silicon rod knife number at the storage location and the system command, and executes the material retrieval or alarm termination process; when the slicing machine loads / unloads materials, the AGV action is triggered through the control terminal, and the knife number is transmitted to the manufacturing execution system; after slicing is completed, the loading / unloading AGV unloads the material and stores the knife number in the RFID card, while the manufacturing execution system automatically updates the process status to pre-cleaning; when the loading / unloading AGV transports the material to the degumming area, it binds the knife number and degumming rack RFID card information and uploads it to the intelligent warehouse management system; Degumming Step: The degumming machine verifies the consistency between the knife number information in the RFID card of the degumming rack and the process status of the manufacturing execution system through the built-in RFID reader; if the process status corresponding to the knife number is not pre-cleaning, the degumming machine triggers an alarm and stops running; after degumming is completed, the knife number is obtained by scanning with a mobile terminal and bound to the RFID card of the sheet-collecting cart. After the material information is entered, it is put into storage. The AGV carrying the material pulls the material to the sheet-laying process according to the instructions of the intelligent warehouse management system; when the AGV carrying the material pulls the material, the intelligent warehouse management system dynamically adjusts the allocation order of the buffer storage location according to the priority of the sheet-laying process, and materials with higher priority are allocated to the storage location closer to the sheet-laying machine first; The stacking process is as follows: The intelligent warehouse management system transmits the degummed material information to the stacking buffer storage location and allocates materials to the buffer storage location according to the processing capacity of the equipment; the AGV carries the materials to the stacking machine, and the stacking machine reads the knife number, precision information and material status from the RFID card of the stacking vehicle through the RFID card reader. After the operator verifies the information, the knife number is uploaded to the manufacturing execution system. In case of abnormality, manual intervention is performed. The intelligent warehouse management system dynamically adjusts the allocation order of cache storage locations based on the priority of the warehousing process, including: The intelligent warehouse management system embeds a priority calculation algorithm to generate a priority index based on at least one of the following factors: processing urgency, material correlation, and equipment load balancing rate. The processing urgency is calculated by the difference between the process deadline pushed by the manufacturing execution system in real time and the current time; the smaller the difference, the higher the priority. Material correlation is determined based on the upstream and downstream dependencies of materials with the same tool number in the scheduling process; materials with higher dependencies have higher priority. The equipment load balancing rate is calculated by the ratio of the real-time processing queue length of the scheduling machine to the theoretical maximum queue length; when the ratio is below a threshold, the priority of the corresponding cache location is increased. The intelligent warehouse management system pre-stores 3D map data of the workshop, marking the coordinates of all cache locations and the coordinates of the stacking machine, and uses Euclidean distance algorithm or A... The path planning algorithm calculates the physical distance between each cache location and the chip setter, and generates a distance-sorted list; The system sorts materials in descending order according to priority index and generates a priority queue; for each material in the priority queue, it selects the nearest available storage location to the stacker from the distance sorting list as the target storage location; The intelligent warehouse management system packages target storage location information, material tool numbers, and transportation route planning data into task instructions, which are then sent to the central controller carrying the AGV via a wireless communication protocol. After receiving the instruction, the AGV carrying the vehicle travels to the target buffer storage location using laser navigation or vision navigation technology, grabs the material using a robotic arm or pallet lifting device, and transports it to the designated loading position of the stacking machine.

2. The method according to claim 1, characterized in that, The automated work order processing module includes: Crystal bonding process module: used to scan the QR code of the crystal rod to read the tool number, and transmit the tool number information to the storage location in the crystal bonding workshop through the intelligent warehouse management system, and then the AGV completes the physical feeding; Slicing process module: The knife number is verified when the AGV picks up the material, and the knife number is automatically transmitted to the intelligent warehouse management system and manufacturing execution system during the loading and unloading process of the slicing machine; Degumming process module: The RFID reader inside the degumming machine reads the knife number information on the RFID card of the degumming rack and uploads it to the manufacturing execution system for order processing; The film arrangement process module uses the RFID reader on the film arrangement machine to read the knife number and related information on the RFID card of the film retrieval cart and displays it on the screen.

3. The method according to claim 1, characterized in that, The AGV dual verification mechanism in the automated work order flow module includes: During the material handling and feeding process, the AGV first scans the QR code information attached to the surface of the silicon rod using its built-in scanning device to obtain the corresponding tool number data of the current silicon rod. At the same time, it receives the tool number instructions for the batch of silicon rods pushed in real time by the intelligent warehouse management system through the wireless communication module. The tool number obtained by scanning is compared and verified with the tool number issued by the intelligent warehouse management system. The AGV will only perform material handling operations if the two match completely. If there is a difference in the comparison result, the abnormal interruption mechanism will be triggered immediately, the current operation will be stopped, and the abnormal event will be reported to the system.

4. The method according to claim 1, characterized in that, In the crystal bonding step: The scanning device is a fixed laser barcode scanner, which is installed on a robotic arm or fixed bracket at the end of the die bonding line. The scanning angle is adjustable to accommodate silicon rods of different specifications. When allocating target storage locations, the intelligent warehouse management system prioritizes storage locations that are physically closest to the current process and are currently available. After the AGV carries silicon rods to the target storage location, it sends a signal of arrival to the intelligent warehouse management system via RFID tags or photoelectric sensors on the target storage location. The system then unlocks the tool number and updates the storage location status to occupied.

5. The method according to claim 1, characterized in that, In the slicing step: The barcode scanning device for the loading and unloading AGV is a handheld or vehicle-mounted QR code scanner; When the slicer is being loaded, a loading request is triggered via a touch screen or voice control terminal. After receiving the instruction, the loading and unloading AGV completes the loading action within a predetermined time and sends a completion signal back to the manufacturing execution system. When the AGV transports materials to the degumming cart area, the RFID card information of the degumming rack is read by the RFID card reader installed at the bottom of the AGV.

6. The method according to claim 1, characterized in that, In the degumming step: The degumming machine has a built-in RFID reader, which is either a high-frequency or ultra-high-frequency reader. When reading the RFID card of the degumming rack, it automatically verifies the matching between the knife number in the card and the current process status of the manufacturing execution system. If the process status is not pre-cleaning, the degumming machine will immediately stop running and issue an audible and visual alarm. The mobile terminal is a PDA device that supports NFC or QR code scanning. After scanning the edge skin QR code or the de-adhesive rack RFID card, a material information entry interface will automatically pop up. The content to be entered includes the storage location number, material status, zone, and operator's employee number.

7. The method according to claim 1, characterized in that, In the film arrangement step: The RFID reader of the film sorting machine is installed next to the lifting rail of the film retrieval vehicle. After successful reading, the knife number, precision information and material status are displayed on the film sorting machine's operation interface. When verifying the edge-cutting information against the system information, if there are discrepancies after K consecutive verifications, the system will automatically lock the current retrieval vehicle and push an abnormal notification to the workshop supervisor's mobile terminal. When allocating materials in the cache storage location, the intelligent warehouse management system calculates the optimal allocation scheme based on the processing capacity of the stacking machine.

8. The method according to claim 7, characterized in that, When allocating materials to buffer storage locations, the intelligent warehouse management system calculates the optimal allocation scheme based on the processing capacity of the stacking machine, including: The intelligent warehouse management system obtains the processing capacity parameters of the stacking machine in real time from the manufacturing execution system or the local controller of the stacking machine, including at least: the maximum number of stacks N processed in a single operation. max Processing speed V process Current processing queue length Q current Theoretical maximum queue length Q max ; Based on the production plan for the film arrangement process and combined with historical processing data, predict the material demand D within the future time window ΔT. predict ; Scan the real-time status of all cache locations and obtain the data: remaining capacity C available Material compatibility ij Physical distance Dist i ; An optimization model is constructed based on linear programming or genetic algorithms, with the objective function being to minimize the total transportation time T. total And maximize the load balancing rate of the film setter L balance The constraints must include at least: processing capacity constraints, storage space capacity constraints, and time window constraints. After solving the optimization model, the optimal allocation scheme is generated, including: Target storage location list: The optimal cache storage location number corresponding to each material; Allocation quantity: The quantity of materials that need to be stored in each storage location; Transportation sequence: the priority sequence of tasks performed by the AGVs carrying the AGVs; When carrying AGVs to perform transportation tasks, the system provides real-time feedback on changes in warehouse location status. If there is a sudden change in the processing speed of the chip stacking machine or the queue length, the optimal allocation scheme is immediately recalculated, and an update instruction is pushed to the AGV via wireless communication protocol.

Images