A layout method and system of a full-automatic material flow production line for luggage production

Abstract

The application discloses a layout method and system of a full-automatic material flow production line for luggage production, and the method comprises the following steps: constructing a four-section physical layout channel comprising an entrance line, a central shuttle line, a parallel station line and an exit line; receiving a material box input by an external cutting piece area through an adaptive turning module of the entrance line, performing direction correction of the material box, and generating a standard material box; detecting the real-time position of the material box in the station unit based on an infrared dual-mode sensor, and generating a state marking signal with verification; in response to the verification result of the completed area in the state marking signal, distributing a shuttle vehicle task instruction to the central shuttle line, generating a material box transfer path, and according to the material box transfer path, controlling the shuttle vehicle to send the completed material box back to the cutting piece area through the exit line, and synchronously releasing the space of the completed area in the station unit. By using the embodiment of the application, accurate orientation of material box flow and efficient reuse of station space can be realized, and the reliable traceability and real-time collaboration of the production state can be ensured.

Description

Technical Field

[0001] This invention belongs to the field of automation technology, specifically a layout method and system for a fully automated material flow production line for bag manufacturing. Background Technology

[0002] Traditional bag production lines often employ linear or circular material flow layouts, resulting in low workstation coordination efficiency, chaotic material flow, and difficulties in status tracking. Existing technologies, with frequent manual intervention in assembly line designs, lead to inconsistent material bin orientation, frequent workstation congestion, and a lack of real-time data verification mechanisms, impacting production cycle time and quality traceability. While some solutions have introduced automated conveyor equipment, issues such as ambiguous workstation zoning and delayed scheduling response remain unresolved, making it difficult to adapt to the demands of flexible production. Furthermore, information silos prevent reliable verification of production status, hindering intelligent upgrades. Summary of the Invention

[0003] The purpose of this invention is to provide a layout method and system for a fully automated material flow production line for bag manufacturing, in order to overcome the shortcomings of the prior art, achieve precise orientation of material box flow and efficient reuse of workstation space, and ensure reliable traceability and real-time collaboration of production status.

[0004] One embodiment of this application provides a layout method for a fully automated material handling production line for bag manufacturing, the method comprising:

[0005] A four-section physical layout channel is constructed, including an inbound line, a central shuttle line, parallel workstation lines, and an outbound line. The workstation lines are composed of multiple independent workstation units connected in parallel. Each workstation unit is divided into a physically isolated unstarted area, an in-process area, and a completed area.

[0006] The adaptive steering module of the inbound line receives the material box input from the external cutting area and performs material box orientation correction to generate a standard material box with uniform orientation and aligned with the workstation line.

[0007] Based on the real-time position detection of the material box in the workstation unit by the infrared dual-mode sensor, combined with the status change button command triggered by the worker, a status mark signal with verification is generated.

[0008] In response to the verification result of the completed area in the status flag signal, the MES dynamic scheduling engine is triggered to allocate shuttle car task instructions to the central shuttle line and generate a material box transfer path.

[0009] According to the material box transfer path, the shuttle car is controlled to send the completed material box back to the cutting area via the exit line, simultaneously releasing the completed area space in the workstation unit, and triggering the blockchain storage module to record the status data of the entire process.

[0010] Optionally, the construction includes a four-section physical layout channel comprising an inbound line, a central shuttle line, parallel workstation lines, and an outbound line. The workstation lines are composed of multiple independent workstation units connected in parallel. Each workstation unit is divided into physically isolated unstarted, in-process, and completed areas, including:

[0011] Analyze the bag manufacturing process flow chart, extract the material input / output relationships including sewing and assembly processes, and generate a process dependency topology graph;

[0012] Based on the process dependency topology diagram, the production line is divided into four segments: inbound buffer zone, central shuttle channel, parallel workstation cluster, and outbound recycling area, and the spatial coordinate framework of the four segments is output.

[0013] Within a workstation cluster with a four-segment spatial coordinate framework, deploy liftable partition walls and load the physical coordinate parameters of the unstarted / in-process / completed zones to generate a three-zone isolation topology map;

[0014] Input the three-zone isolation topology diagram into the logistics simulation system to verify that there is no conflict between the material box transfer path and the partition wall action, and output a four-segment layout channel with isolation parameters.

[0015] Optionally, the adaptive steering module via the inbound line receives the material bins input from the external cutting area and performs material bin orientation correction to generate standard material bins with uniform orientation and aligned with the workstation line, including:

[0016] The reference direction vector of the workstation line is extracted from the four-segment layout channel with isolation parameters and used as the correction target reference.

[0017] By scanning the surface of the material box in the station with 3D line laser, the corner coordinates and the center of gravity offset are extracted to generate the material box pose deviation matrix;

[0018] Based on the pose deviation matrix and the work station reference direction, calculate the pitch / yaw compensation angle of the three-axis steering mechanism and output the real-time steering control command set;

[0019] The parallel steering mechanism is driven to execute a set of control commands, and at the same time, pose correction data stream is generated through binocular visual feedback;

[0020] When the angle between the pose correction data stream and the reference direction is less than the preset angle, the pneumatic locking device is triggered and a standard material box with a direction verification code is output.

[0021] Optionally, the real-time position detection of the material bins within the workstation unit based on the infrared dual-mode sensor, combined with the status change button command triggered by the worker, generates a verified status marker signal, including:

[0022] Based on the coordinate parameters of the three-zone isolation topology, the detection area of ​​the dual-mode infrared sensor within the workstation unit is calibrated.

[0023] Scan the spatial coordinates of the standard material bin within the workstation unit and output the material bin-zone location mapping table;

[0024] Parse the worker button commands and the bin-zone location mapping table, and generate a status change permission flag when the operation conforms to the zone transfer rules;

[0025] Retrieve the historical status timestamp of the material bin, check whether the current operation interval meets the process threshold, and output the timing compliance assessment value;

[0026] It integrates the status change permission flag with the timing compliance assessment value, and adds a digital signature to generate a verified status marker signal.

[0027] Optionally, in response to the verification result of the completed area in the status flag signal, the MES dynamic scheduling engine is triggered to allocate shuttle task instructions to the central shuttle line and generate a material bin transfer path, including:

[0028] Extract the completed area bin ID, workstation coordinates, and verification timestamp from the verification status marker signal to construct a bin transfer trigger event;

[0029] Based on the material bin transfer trigger event, collect the shuttle car location, aisle occupancy status, and workstation demand priority to generate a dynamic resource status matrix;

[0030] Using a dynamic resource situation matrix as input, a dynamic window algorithm is used to solve for the optimal path and output the initial path point sequence.

[0031] The system detects spatiotemporal conflict points in the path point sequence and adjusts them through a virtual track allocation mechanism to generate the final material box transfer path.

[0032] Optionally, according to the material box transfer path, the shuttle car is controlled to send the completed material boxes back to the cutting area via the exit line, simultaneously releasing the completed area space within the workstation unit, and triggering the blockchain evidence storage module to record the entire process status data, including:

[0033] The shuttle car operation is controlled based on the final material bin transfer path, and trajectory deviation alarm events are generated in real time through the UWB positioning module.

[0034] When the trajectory deviation alarm event is empty, the outbound steering module is triggered to rotate to the outbound direction and outputs the outbound calibration execution command;

[0035] Based on the completion signal of the outgoing calibration execution command, the locking of the workstation unit partition wall is released, and a workstation space release signal is generated;

[0036] The system integrates trajectory deviation alarm events, workstation space release signals, and quality inspection data, and writes them into the blockchain for evidence storage through the BFT consensus mechanism.

[0037] Another embodiment of this application provides a layout system for a fully automated material handling production line for bag manufacturing, the system comprising:

[0038] The construction module is used to construct a four-segment physical layout channel including an inbound line, a central shuttle line, parallel workstation lines, and an outbound line. The workstation lines are composed of multiple independent workstation units connected in parallel. Each workstation unit is divided into a physically isolated unstarted area, an in-process area, and a completed area.

[0039] The receiving module is used to receive the material boxes input from the external cutting area through the adaptive steering module of the inbound line and perform material box orientation correction to generate standard material boxes with uniform orientation and aligned with the workstation line.

[0040] The detection module is used to detect the real-time position of the material box in the workstation unit based on the infrared dual-mode sensor, and generate a verified status mark signal in combination with the status change button command triggered by the worker.

[0041] The allocation module is used to respond to the verification result of the completed area in the status flag signal, trigger the MES dynamic scheduling engine, allocate shuttle car task instructions to the central shuttle line, and generate a material box transfer path.

[0042] The control module is used to control the shuttle car to send the completed material boxes back to the cutting area via the exit line according to the material box transfer path, simultaneously release the completed area space in the workstation unit, and trigger the blockchain storage module to record the status data of the entire process.

[0043] Another embodiment of this application provides a storage medium storing a computer program, wherein the computer program is configured to execute the method described in any of the preceding claims when running.

[0044] Another embodiment of this application provides an electronic device including a memory and a processor, wherein the memory stores a computer program and the processor is configured to run the computer program to perform the method described in any of the preceding claims.

[0045] Compared with existing technologies, this invention provides a layout method for a fully automated material handling production line for bag manufacturing. It constructs a four-segment physical layout channel including an inbound line, a central shuttle line, parallel workstation lines, and an outbound line. Through the adaptive steering module of the inbound line, it receives material boxes input from the external cutting area and performs material box orientation correction to generate standard material boxes. Based on infrared dual-mode sensors, it detects the real-time position of material boxes within the workstation unit and generates a verified status marker signal. In response to the verification result of the completed area in the status marker signal, it assigns shuttle car task instructions to the central shuttle line, generating a material box transfer path. According to the material box transfer path, it controls the shuttle car to send the completed material boxes back to the cutting area via the outbound line, simultaneously releasing the completed area space within the workstation unit. This enables precise orientation of material box transfer and efficient reuse of workstation space, ensuring reliable traceability and real-time collaboration of production status. Attached Figure Description

[0046] Figure 1 A hardware structure block diagram of a computer terminal for a layout method of a fully automated material flow production line for bag production provided in an embodiment of the present invention;

[0047] Figure 2 This is a schematic flowchart of a layout method for a fully automated material transfer production line for bag production provided in an embodiment of the present invention;

[0048] Figure 3 This is a schematic diagram of the layout system of a fully automated material transfer production line for bag production, provided in an embodiment of the present invention. Detailed Implementation

[0049] The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0050] This invention first provides a layout method for a fully automated material flow production line for bag production. This method can be applied to electronic devices, such as computer terminals, specifically ordinary computers.

[0051] The following detailed explanation uses a computer terminal as an example. Figure 1 This is a hardware structure block diagram of a computer terminal for a layout method of a fully automated material handling production line for bag manufacturing, provided in an embodiment of the present invention. (See diagram for reference.) Figure 1 As shown, the computer device includes a processor, memory, and network interface connected via a system bus, wherein the memory may include non-volatile storage media and internal memory.

[0052] The non-volatile storage medium can store an operating system and a computer program. This computer program includes program instructions that, when executed, cause the processor to perform any layout method for a fully automated material handling production line used in bag manufacturing.

[0053] The processor provides computing and control capabilities, supporting the operation of the entire computer device.

[0054] The internal memory provides an environment for the execution of computer programs in non-volatile storage media. When the computer program is executed by the processor, it enables the processor to execute any layout method of a fully automated material flow production line for bag production.

[0055] This network interface is used for network communication, such as sending assigned tasks. Those skilled in the art will understand that... Figure 1 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0056] It should be understood that the processor can be a Central Processing Unit (CPU), but it can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among these, a general-purpose processor can be a microprocessor or any conventional processor.

[0057] See Figure 2 The present invention provides a layout method for a fully automated material handling production line for bag production, which may include the following steps:

[0058] S201, construct a four-segment physical layout channel including an inbound line, a central shuttle line, parallel workstation lines and an outbound line. The workstation lines are composed of multiple independent workstation units connected in parallel. Each workstation unit is divided into a physically isolated unstarted area, an in-process area and a completed area.

[0059] Specifically, it can analyze the bag manufacturing process flow chart, extract the material input / output relationships including sewing and assembly processes, and generate a process dependency topology diagram;

[0060] The system first imports a Bag Manufacturing Process Flowchart, drawn by process engineers using professional design software (such as AutoCAD or Tecnomatix). This flowchart details the entire sequence of bag manufacturing processes, including cutting, sewing, labeling, assembly, and quality inspection. The flowchart uses a directed graph structure to represent process relationships: nodes represent processes (e.g., "front panel sewing," "zipper installation"), and directed edges represent material flow directions (e.g., "cutting area → sewing station"). The system automatically parses the symbols and lines in the flowchart using an image recognition engine (based on the OpenCV library), and combines this with a Natural Language Processing (NLP) module to recognize text labels next to process names (e.g., "Input: front panel; Output: sewing semi-finished product"), extracting the Input Material List (IML) and Output Material List (OML) for each process. For example, the IML for the "main bag assembly" process might include "side panel cutting, bottom panel cutting, zipper," while the OML would be "assembled main bag body." The system establishes process dependency chains based on the matching relationship between IML and OML: if the output material of process A is the input of process B, then a directed edge A→B is established, ultimately forming a process dependency topology graph. This graph is a directed acyclic graph (DAG), which can intuitively show the sequential order (e.g., "sewing" must follow "cutting") and parallel relationships between processes (e.g., "handle sewing" and "bag body sewing" can be performed simultaneously).

[0061] To improve the accuracy of the topology diagram, the system introduces Bill of Materials (BOM) data as a supplement. The BOM, provided by the Enterprise Resource Planning System (ERP), contains the complete hierarchical relationship of materials required for each bag model (e.g., "suitcase → body → outer shell → PC sheet"). The system performs code mapping between the material codes in the BOM and the process IML / OML in the flowchart; for example, mapping "PC sheet" to the output of the "outer shell cutting" process. For complex processes (such as "inner lining assembly," which may depend on the output of multiple preceding processes), the system uses a graph traversal algorithm to verify dependencies: tracing back from the final process (e.g., "final assembly completed"), if it is found that the IML material of a certain process cannot be covered by the preceding OML, a dependency missing alert is triggered, prompting engineers to supplement the process connection. After the topology graph is generated, the system calculates the critical path (CP) and identifies the longest process chain (such as "cutting → sewing → quality inspection → packaging"), providing a basis for bottleneck analysis in subsequent production line layout. All process nodes are assigned process parameter attributes, including standard time (ST, unit: seconds), required equipment (such as "high-speed sewing machine"), and safety distance requirements (such as sewing machine spacing ≥ 1.5 meters).

[0062] The final generated process flow depends on the topology diagram, which is stored as JSON structured data (JavaScript Object Notation, a lightweight data interchange format) and contains three core elements:

[0063] Node Set: Each node contains a process ID (e.g., OP001), name (e.g., "front panel sewing"), type (sewing / assembly, etc.), and a list of IML / OML material codes.

[0064] Edge Set: Each edge contains the source node ID, the target node ID, and the dependency type (strong dependency / weak dependency). Strong dependency means that the process must be executed strictly in sequence (e.g., "applying glue" must be done before "pressing"), while weak dependency allows processes to be performed in parallel (e.g., "installing decorative buckles" and "sewing handles").

[0065] Global Attributes: These include the topology map version number (VN), generation timestamp (TS), and validation status (VS). This topology map will serve as the logical basis for subsequent physical layout design, ensuring that production line partitions and process flows are fully matched.

[0066] Based on the process dependency topology diagram, the production line is divided into four segments: inbound buffer zone, central shuttle channel, parallel workstation cluster, and outbound recycling area, and the spatial coordinate framework of the four segments is output.

[0067] The system performs production line zoning based on the process dependency topology diagram. First, an Inbound Buffer Zone (IBZ) is defined: this area, adjacent to the external cutting area, receives the boxes of materials to be processed (containing cut pieces). Based on the material input frequency of all starting processes (nodes without prior dependencies) in the topology diagram (e.g., receiving 120 boxes of cut pieces per hour), and combined with the box dimensions (length × width × height = 0.8m × 0.6m × 0.4m), the minimum area requirement (AR) of the buffer zone is calculated. For example, if 15 minutes' worth of materials (30 boxes) needs to be buffered, and each box occupies 0.48 square meters (including spacing), then the IBZ area is ≥ 14.4 square meters. An installation position for an adaptive steering module (described later) is set at the end of the buffer zone. Next, plan the Central Shuttle Corridor (CSC): This is the core logistics channel connecting all workstations. Its width is determined by the size of the Automated Guided Vehicle (AGV) (e.g., 1.2 meters wide) and the need for two-way traffic, and is usually set at 3 meters. The length of the corridor covers the entire workstation cluster, and a turning radius (Turning Radius, TR ≥ 1.5 meters) must be reserved.

[0068] The Parallel Workstation Cluster (PWC) is the core work area. The system determines the number of workstation units (e.g., 12) based on the parallel process chains (such as three independent sewing lines) in the topology diagram. The area of ​​each workstation unit (WU) is determined by the equipment layout (sewing machine + workbench) and the division into three zones (not started / in progress / completed), with a typical size of 4 meters × 3 meters. Workstation units are arranged in a parallel topology: multiple units share the access point of the central shuttle aisle, forming a "fishbone" layout. The total cluster length (CL) is calculated as follows: CL = (single workstation length × number of workstations per column) × number of columns + column spacing × (number of columns - 1).

[0069] For example, with 3 columns × 4 workstations (each workstation is 4 meters long) and a column spacing of 2 meters, then CL = (4 × 4) × 3 + 2 × 2 = 52 meters. Finally, define the Outbound Recycling Zone (ORZ): located at the end of the production line, used to temporarily store finished material bins (waiting to be returned to the cutting area). Its area calculation is similar to the inbound buffer zone, but it needs to consider the switching between empty and full states of the material bins (empty bins have half the volume when folded).

[0070] The system integrates the geometric parameters (length, width, height, origin, and orientation angle) of the four regions into a four-segment spatial coordinate frame (FSSCF). It uses an absolute coordinate system (ACS), with the origin (0,0,0) at the corner of the production line entrance. The X-axis runs along the length of the production line, the Y-axis along the width, and the Z-axis points vertically upwards. For example:

[0071] IBZ coordinate range: X[0,5], Y[0,3] (unit: meters);

[0072] CSC coordinate range: X[5,55], Y[1.5,4.5];

[0073] PWC coordinate range: X[5,55], Y[0,1.5] and Y[4.5,6] (both sides of the channel);

[0074] ORZ coordinate range: X[55,60],Y[0,3].

[0075] The framework outputs in DXF file format (Drawing Exchange Format, a CAD data exchange standard), which can be directly imported into layout design software for 3D visualization.

[0076] Within a workstation cluster with a four-segment spatial coordinate framework, deploy liftable partition walls and load the physical coordinate parameters of the unstarted / in-process / completed zones to generate a three-zone isolation topology map;

[0077] Physical isolation is achieved by deploying retractable partition walls (RPWs) within each workstation unit (WU). The partition walls utilize pneumatic actuation or electric linear actuators (ELA), with a lifting speed ≥0.2 m / s and a noise level ≤65 dB. Three sets of partition walls are installed in each workstation unit.

[0078] Partition Wall 1 (PW1): Separates the Not Started Zone (NSZ) from the In-Progress Zone (IPZ), with a height of 1.2 meters (semi-isolation, allowing workers to observe);

[0079] Second partition wall (PW2): Separates the IPZ from the completed zone (CZ), with a height of 2 meters (fully isolated to prevent accidental operation).

[0080] Side Partition Wall (SPW): Separates adjacent workstations and is made of fire-resistant acrylic panel. A magnetic seal strip (MSS) is embedded at the bottom of the partition wall to ensure a tight fit with the floor track when lowered.

[0081] The system defines the physical coordinate parameters (PCP) for three zones for each workstation unit:

[0082] Unstarted Zone (NSZ): Located on the side of the workstation entrance, with coordinates X[WU_Xmin,WU_Xmin+1.2],Y[WU_Ymin,WU_Ymax] (e.g., X[0,1.2],Y[0,3]), used to store material bins to be processed, and marked with yellow marker tape (YMT) on the ground.

[0083] In-process Zone (IPZ): Core work area, coordinate range X[WU_Xmin+1.2,WU_Xmin+2.8],Y[WU_Ymin,WU_Ymax] (e.g., X[1.2,2.8],Y[0,3]), includes sewing machines and workbenches, and the floor is marked with green marker tape (GMT).

[0084] Completed Zone (CZ): Exit side, coordinate range X[WU_Xmin+2.8,WU_Xmax], Y[WU_Ymin,WU_Ymax] (e.g., X[2.8,4.0],Y[0,3]), with blue marker tape (BMT) on the ground. The coordinate parameters are written to the local storage unit (LSU) of the workstation controller.

[0085] Integrate the coordinates of the three zones and the location information of partition walls for all workstations to generate a Three-Zone Isolation Topology Map (TZITM). This map is a hierarchical structure graph.

[0086] Top layer: Overall boundary coordinates of the workstation cluster;

[0087] Middle layer: The ID of each workstation unit (e.g., WS001) and its three-zone coordinate range;

[0088] Bottom layer: Installation coordinates of each partition wall (e.g., PW1's coordinates in WS001 (X=1.2, Y=1.5)), lifting status (0=lowered / 1=raised), and safety sensor ID (e.g., infrared beam sensor IRS001). The topology graph is stored as a Topological Adjacency Matrix (TAM) for easy and quick lookup of region connectivity (e.g., whether NSZ and IPZ are isolated).

[0089] Input the three-zone isolation topology diagram into the logistics simulation system to verify that there is no conflict between the material box transfer path and the partition wall action, and output a four-segment layout channel with isolation parameters.

[0090] The system imports the Three-Zone Isolation Topology Map (TZITM) into a Logistics Simulation System (LSS), such as FlexSim or AnyLogic software. A Digital Twin Model (DTM) is then constructed within the simulation environment.

[0091] Static model: The three-dimensional geometry of the entrance line, shuttle passage, work station cluster and exit line is established according to the four-segment spatial coordinate framework, and the material properties (such as the ground friction coefficient is set to 0.6) are consistent with the actual situation.

[0092] Dynamic model:

[0093] AGV (Automated Guided Vehicle): Dimensions 1.2m × 0.8m, maximum speed 1.5m / s, equipped with LIDAR obstacle avoidance algorithm (LIDAR: Light Detection and Ranging, lidar);

[0094] Material Tote (MT): Dimensions 0.8m × 0.6m × 0.4m, weight available in empty (5kg) and full (20kg) containers;

[0095] Partition wall: Add lifting animation (lifting time ≤ 2 seconds) and collision volume.

[0096] The simulation system executes the Conflict Detection (CD) process:

[0097] Step 1: Path Verification

[0098] The simulated shuttle vehicle picks up a material bin from the inbound line and transports it along the central aisle to the target workstation (NSZ). As the AGV approaches the workstation, the system checks:

[0099] Has the partition wall (PW1) of the target NSZ been raised (to ensure that the AGV can enter)?

[0100] Does the AGV path overlap with the SPW projection of adjacent workstations (a safe distance of ≥0.3 meters must be maintained)?

[0101] Step 2: State Coordination Verification

[0102] Simulated worker operation: When the hopper is moved from NSZ to IPZ, the system detects:

[0103] Should PW1 be degraded (to isolate NSZ and IPZ)?

[0104] After the worker triggers the "Start" button, does PW1 rise within 1 second (allowing the next hopper to enter)?

[0105] Step 3: Recycling Path Verification

[0106] After the material bin is processed and moved into the CZ (Customer Zone), a simulated shuttle car enters the CZ to retrieve the goods.

[0107] Check if PW2 is raised in time (to prevent the AGV from hitting the partition wall)?

[0108] Is the exit passage occupied by other AGVs (dynamic avoidance required)?

[0109] If the simulation detects a conflict (such as a collision between the AGV and a PW2 that has not yet descended), the system records a conflict event (CE), which includes:

[0110] Collision Type (CT: Spatial / Temporal);

[0111] Conflict Location (CL: e.g., CZ entrance of WS003);

[0112] Timestamp (TS).

[0113] The system automatically adjusts parameters (e.g., advancing the PW2's rise action by 0.5 seconds) and resimulates until all conflicts are eliminated. The final output is a four-segment layout channel with isolation parameters (FSLIP), including:

[0114] The final version of the Spatial Coordinate Frame (FSSCF);

[0115] Isolation Parameters (IP): The timing rules for the actions of each partition wall (e.g., "When the shuttle is 3 meters away from the workstation, raise PW2");

[0116] Logistics Rule Set (LRS): such as "A maximum of 3 bins can be stored in the unstarted area".

[0117] This result serves as a blueprint for the production line construction, ensuring complete consistency between the physical layout and the digital design.

[0118] This method achieves modular zoning of the bag production line through a four-segment physical layout design. The inbound line handles raw material input, the central shuttle line serves as the main logistics artery, parallel workstations provide parallel operation capabilities, and the outbound line handles finished product recycling. Each workstation unit is physically divided into three clearly defined functional areas, ensuring physical isolation and visual control of the production status. This layout significantly improves the flexibility and scalability of the production line. The parallel design of workstation units allows for flexible adjustment of production capacity according to order demand, while the physical isolation of the three zones effectively avoids the mixing of semi-finished products, laying a physical foundation for subsequent intelligent scheduling.

[0119] S202, through the adaptive steering module of the inbound line, receives the material box input from the external cutting area and performs material box orientation correction to generate a standard material box with uniform orientation and aligned with the workstation line.

[0120] Specifically, the reference direction vector of the workstation line can be extracted from the four-segment layout channel with isolation parameters and used as the correction target reference.

[0121] After the physical layout of the production line is determined, the control system first needs to establish an absolute reference standard for orientation correction. The core of this step is accessing the "four-segment layout channel with isolation parameters" model data stored in the production line's digital twin system. This model contains precise spatial coordinate information verified by logistics simulation, especially the spatial arrangement data of parallel workstation lines. The system extracts the Workstation Line Reference Direction Vector (WLRDV) from this model. This vector is essentially a directional arrow in three-dimensional space, for example, in the factory coordinate system it might be represented as (X = 1.2 meters, Y = 0 meters, Z = 0) pointing to (X = 1.2 meters, Y = 15 meters, Z = 0), and its direction (here along the positive Y-axis) represents the standard orientation of all workstation unit entrances. This vector is set as the Calibration Target Benchmark (CTB) for the entire adaptive steering process. This reference orientation ensures that regardless of the angle or posture at which the material tote (MT) is conveyed in the external cutting area, the final calibrated material tote will be arranged strictly parallel to the workstation line, laying the directional foundation for precise positioning and automated flow within subsequent workstation units. The reference orientation vector data is typically stored in a central control database and synchronized in real time to the local controller (Programmable Logic Controller, PLC) of the steering module via industrial Ethernet.

[0122] To ensure the accuracy and robustness of the reference orientation, the system employs a multi-verification mechanism. First, during the initial commissioning phase of the production line, a high-precision total station (TS) is used to survey the actually installed workstation lines, obtaining their actual coordinate data for physical position and orientation. This measured data is compared and calibrated with the theoretical coordinates in the digital twin model, generating a small Spatial Offset Compensation Matrix (SOCM). Second, after each system startup or periodic maintenance, an automatic reference calibration process is executed: a standard calibration tote (CT) with known dimensions and feature points is fed into the line, its ideal position and orientation defined by WLRDV. The standard tote is scanned by the steering module's 3D Line Laser Scanner (3DLLS), and the scan results are compared with the expected values, dynamically updating the SOCM. Ultimately, the reference direction used for real-time correction is actually the result of theoretical WLRDV after SOCM compensation. This effectively eliminates the long-term drift effects caused by installation errors, thermal deformation, or foundation settlement, ensuring the long-term stability and reliability of the reference.

[0123] The extraction and application of the Workstation Reference Direction Vector (WLRDV) is not done in isolation, but is tightly embedded in the spatial coordination framework of the entire production line. This vector defines the standard interface direction of the workstation line relative to the Central Shuttle Line (CSL) and Infeed Line (IFL). Based on the WLRDV, the control system automatically calculates the ideal parking coordinates (IPC) and orientation of the outfeed bins in the "unstarted zone" of the workstation unit. Simultaneously, this reference direction also guides the setting of the outfeed line (OFL) bin rotation direction, ensuring that the outfeed bin orientation matches the requirements for piece recycling in the cutting area. The navigation systems of all related logistics equipment (such as shuttles and AGVs) also use this unified spatial reference for their path planning algorithms. Therefore, establishing this accurate and unified Correction Target Reference (CTB) is the cornerstone for achieving spatial consistency and efficient collaborative operation of the entire material flow system.

[0124] By scanning the surface of the material box in the station with 3D line laser, the corner coordinates and the center of gravity offset are extracted to generate the material box pose deviation matrix;

[0125] As the material bin (MT) enters the inbound line (IFL) from the external cutting area and reaches the turning station, the 3D line laser scanner (3DLLS) is immediately activated. This scanner, typically mounted above the turning platform, consists of a high-speed rotating or vibrating line laser emitter (LLE) and one or more high-resolution area scan cameras (ASC). The laser line (LL) is projected onto the moving surface of the material bin at a specific angle (e.g., 45 degrees). The camera captures the deformation image of the laser line on the material bin surface. Based on the Laser Triangulation Principle (LTP), the system analyzes the degree and position of the laser line's deformation (DFM) in the image to calculate in real-time a precise 3D point cloud (3DPC) of the area scanned by the laser line on the material bin surface. A complete scan is typically completed as the material bin passes through the scanned area, generating dense point cloud data covering the top and multiple sides of the material bin.

[0126] The system's built-in Geometry Model Recognition Algorithm (GMRA) begins processing the point cloud data. The algorithm first performs Point Cloud Filtering (PCF) to remove noise and irrelevant background points. Then, it uses a Feature Extraction Algorithm (FEA) to identify the key geometric features of the hopper. The most important feature is the coordinates of the hopper's eight corner points (CPC). The algorithm accurately calculates the three-dimensional coordinates (X, Y, Z) of these eight corner points in the Factory Coordinate System (FCS) by using methods such as plane fitting (PF), edge line detection (ELD), and polyhedron intersection points (PIP). Simultaneously, the system calculates the center of gravity offset (CGO) of the hopper based on the point cloud data. This is typically done in two ways: 1) If the contents of the bin are known and uniformly distributed, the center of gravity is approximated as the geometric center of the bin; 2) If the contents may be non-uniform, the volume distribution of the bin is estimated by point cloud, and combined with the preset empty weight of the bin and the density of typical materials, the offset of the actual center of gravity (ACG) relative to the geometric center (GC) in the X, Y, and Z directions (ΔX, ΔY, ΔZ) is estimated.

[0127] After obtaining the coordinates of the eight corner points (CPC) and the center of gravity offset (CGO), the system compares this information with the standard tote model (STM) of this type of bin stored in the database. The standard model contains the corner coordinates and the center of gravity position (usually the geometric center) of the bin in an ideal state (correct orientation, no tilt, no offset). Through spatial coordinate transformation calculations (mainly rigid body transformation), the system can solve for the deviation of the current bin from the ideal pose (i.e., aligned with the workstation reference direction vector WLRDV). This deviation is usually represented by six degrees of freedom (6DOF): three translational components (offsets along the X, Y, and Z axes: Offset_X, Offset_Y, Offset_Z) and three rotational components (rotation angles around the X, Y, and Z axes: Roll, Pitch, Yaw). These six deviation parameters are systematically organized into a mathematical structure, namely the bin pose deviation matrix (TPDM). This matrix precisely quantifies all deviations of the current bin from the target reference (CTB) in terms of position and orientation, providing accurate input data for subsequent steering compensation. The scanning and calculation process must be completed within a short time (e.g., within 0.5 seconds) of the bin passing through the scanning area to meet production line cycle time requirements.

[0128] Based on the pose deviation matrix and the work station reference direction, calculate the pitch / yaw compensation angle of the three-axis steering mechanism and output the real-time steering control command set;

[0129] After receiving the precise hopper pose deviation matrix (TPDM), the core task of the control system is to convert this spatial deviation data into specific instructions to drive the three-axis steering mechanism (TASM). The three-axis steering mechanism is the key actuator for achieving multi-degree-of-freedom adjustment of the hopper, typically consisting of three independent precision motion axes: 1) Horizontal rotation axis (Yaw Axis, YA): Drives the entire steering platform (or hopper clamp) to rotate around the Z-axis perpendicular to the ground, correcting the hopper's horizontal orientation (Yaw angle deviation). 2) Forward and backward tilt axis (Pitch Axis, PA): Drives the platform to rotate around the horizontal Y-axis (perpendicular to the conveying direction), correcting the height difference between the front and rear ends of the hopper (Pitch angle deviation, i.e., "head up" or "head down"). 3) Left and right tilt axis (Roll Axis, RA): Drives the platform to rotate around the horizontal X-axis (parallel to the conveying direction), correcting the height difference between the left and right sides of the hopper (Roll angle deviation, i.e., "left tilt" or "right tilt"). These three axes are typically composed of transmission components such as high-precision servo motors (SM), harmonic drives (HD), or precision ball screws (BS).

[0130] The core algorithm for calculating the compensation angle is the Inverse Kinematics Solution (IKS). The algorithm input is the rotational angle deviation (Roll_Err, Pitch_Err, Yaw_Err) of the current bin relative to the target reference, contained in the TPDM. The goal is to adjust the bin's orientation to be completely parallel and horizontal with the workstation reference direction vector (WLRDV). Since the three rotational axes of the three-axis steering mechanism (TASM) are usually orthogonal and kinematically decoupled (or approximately decoupled), the calculation process is relatively straightforward:

[0131] Yaw Compensation Angle (YCA) = -Yaw_Err (the angle that the horizontal rotation axis YA needs to rotate, the negative sign indicates reverse compensation);

[0132] Pitch Compensation Angle (PCA) = -Pitch_Err (the angle that the forward and backward tilt axis PA needs to rotate);

[0133] Roll Compensation Angle (RCA) = -Roll_Err (the angle that the left and right tilt axes RA need to rotate).

[0134] In addition to angle compensation, positional offsets (Offset_X, Offset_Y, Offset_Z) in the TPDM also need to be handled. Offset_X (offset along the conveyor direction) and Offset_Y (offset perpendicular to the conveyor direction) are typically compensated for by a small-range translation of the steering mechanism platform in the horizontal plane (achieved via a translation slide attached to or separate from the YA axis). Offset_Z (height offset) may be ignored (handled by subsequent conveyor lines) or compensated for by a height adjustment mechanism (if integrated with TASM). The calculated compensation angles (YCA, PCA, RCA) and possible translations, combined with the maximum velocity (Max Velocity, MV), maximum acceleration (Max Acceleration, MA) of each motion axis and motion smoothness requirements (such as S-curve planning), are used by the Motion Control Card (MCC) to generate a specific real-time steering control command set (RSCCS). This instruction set includes parameters such as target position (TP), target velocity (TV), and acceleration profile (AP) for each servo motor.

[0135] In actual calculations, the physical limitations and dynamic characteristics of the mechanism must also be considered. For example, each rotation axis has its own angular range limit (e.g., Yaw axis ±180°, Pitch / Roll axis ±10°). The algorithm will immediately check whether the limit check (LC) is exceeded after calculating the compensation angle. If it is exceeded, it will execute according to the limit value and trigger an alarm. At the same time, the system will evaluate the time required for the compensation action (Motion Time Estimate, MTE) to ensure that it is completed within the cycle time allowed by the production line (e.g., no more than 1.5 seconds). In addition, due to the existence of the bin's center of gravity offset (CGO), adjusting the Pitch and Roll angles may cause the material inside the bin to slide or even tip over. The system will combine the CGO data and the characteristics of the bin's contents (e.g., whether there are anti-slip pads, whether it is full) to calculate the stability margin (SM). If the stability margin is lower than the safety threshold (e.g., SM < 1.5), the system will limit the compensation angle or reduce the adjustment speed, and trigger the clamping mechanism to act in advance if necessary. The calculated real-time steering control command set (RSCCS) is sent in real time to each servo drive (SD) of the three-axis steering mechanism (TASM) via a high-speed fieldbus (such as EtherCAT) for execution.

[0136] The parallel steering mechanism is driven to execute a set of control commands, and at the same time, pose correction data stream is generated through binocular visual feedback;

[0137] Upon receiving the Real-Time Steering Control Command Set (RSCCS), the parallel steering mechanism (referring to the three-axis steering mechanism TASM) immediately triggers a response from its servo system (SS). The servo driver (SD) for each motion axis (YA, PA, RA) parses the target position (TP), velocity curve (VC), and acceleration curve (AC) from the command, driving the servo motor (SM) to rotate. This motor rotation is converted into precise angular movement of the steering platform via a high-rigidity, low-backlash transmission mechanism (such as a harmonic reducer HD or precision crossed roller bearings). Position Closed-Loop Control (PCLC) is the core: a high-resolution absolute encoder (AE, resolution e.g., 20 bits / revolution) mounted on the motor or output shaft provides real-time feedback of the actual position (AP) to the driver. The high-speed PID controller (Proportional-Integral-Derivative Controller, PIDC) inside the drive continuously compares AP and TP, generating a torque command (TC) to drive the motor, quickly eliminating position error (PE), and ensuring that the platform moves accurately and smoothly to the angle required by the command (YCA, PCA, RCA). The entire execution process emphasizes fast dynamic response (e.g., millisecond-level response), high positioning accuracy (e.g., repeatability ±0.05°), and smooth, shock-free operation.

[0138] While the steering mechanism performs compensation actions, an independent monitoring system is activated: a binocular vision system (BVS). This system typically consists of two high-frame-rate industrial cameras (ICs, such as global shutter CMOS cameras) mounted above or to the side of the steering platform at a specific distance (baseline distance, BD, e.g., 200mm) and angle (convergence angle, CA), covering the entire bin area. The cameras synchronously trigger to capture image pairs (IPs) of the bin at the start, during (at a preset frequency), and after the mechanism's movement. Based on the stereo vision principle (SVP), the system processes the image pairs through the following steps: 1) Camera Calibration (CC): Precisely acquiring the intrinsic and extrinsic parameters and relative pose of the two cameras. 2) Stereo Matching (SM): Finding the correspondence between identical feature points (such as corner points and texture points) on the bin in the two images. 3) 3D Reconstruction (3DR): Calculates the 3D coordinates of the matching points using triangulation. 4) Pose Estimation (PE): Matches the reconstructed feature point coordinates with the bin model to calculate the bin's current real-time pose (RTTP).

[0139] The calculated real-time pose (RTTP) data is encapsulated into a continuous pose correction data stream (PCDS). PCDS typically includes fields such as timestamp (TS), position (X, Y, Z), attitude (Roll, Pitch, Yaw), and confidence (CF). This data stream is transmitted in real-time at a high frequency (e.g., 50Hz) to the upper-level controller (such as a PLC or industrial PC) of the steering control system. The core functions of PCDS are twofold: 1) Closed-loop feedback (CLF): During the execution of the RSCCS, the actual achieved pose (RTTP) is compared in real-time with the target pose required by the command (implied by the RSCCS). If the deviation exceeds the allowable range (e.g., attitude angle error > 0.5°), the controller can dynamically fine-tune the RSCCS (e.g., issue additional compensation commands) to achieve dynamic closed-loop control and improve the final positioning accuracy. 2) Process Monitoring & Verification (PMV): After the mechanism completes its actions, PCDS provides independent measurements of the final pose to verify whether the steering operation successfully corrected the bin to the allowable tolerance range of the target reference direction (CTB) (e.g., the angle with WLRDV < 0.3°). Binocular vision feedback provides non-contact, high-precision independent measurements, which is an important supplement and verification to the mechanism's own encoder feedback, greatly enhancing the system's reliability and fault tolerance.

[0140] When the angle between the pose correction data stream and the reference direction is less than the preset angle, the pneumatic locking device is triggered and a standard material box with a direction verification code is output.

[0141] The control system continuously monitors the material box attitude angles (primarily the yaw angle, as it is most relevant to the alignment with the workstation line) calculated from the pose correction data stream (PCDS). The system compares the real-time yaw angle (Current_Yaw) in the PCDS with the target yaw angle (Target_Yaw) represented by the workstation line reference direction vector (WLRDV) in real time, calculating the angle between them (Angle Difference, AD): AD = |Current_Yaw - Target_Yaw|. Simultaneously, the system presets a critical threshold: the preset angle (PAT). PAT is the maximum allowable directional deviation angle according to process requirements, and its value is set based on the subsequent workstation docking accuracy requirements (e.g., 0.5 degrees or 1.0 degrees). When the system detects that AD ≤ PAT (i.e., the angle is less than or equal to the preset angle), it considers the material box orientation correction to have met the acceptable standard. This judgment is usually based on several consecutive (e.g., 3) PCDS samples meeting the condition to avoid misjudgments caused by momentary jitter. At this point, the system issues a "Orientation Correction Complete" (OCC) signal.

[0142] Once the "Orientation Correction Complete" (OCC) signal is generated, the system immediately triggers the pneumatic locking device (PLD). The pneumatic locking device is a key mechanism to ensure that the orientation of the hopper is not accidentally disturbed or deviated when it leaves the steering platform and enters the next section of the conveyor line. A typical PLD includes: 1) Lateral Positioning Cylinder (LPC): usually arranged in pairs on both sides of the hopper. When triggered, compressed air (CA, pressure, e.g., 0.6 MPa) drives the cylinder piston rod to extend, pushing a push rod (PR) with a rubber or polyurethane buffer head against the side wall of the hopper, eliminating any possible residual positional offset and rotational freedom of the hopper in the horizontal plane (XY plane). 2) Bottom Clamping Cylinder (BCC): This system raises a lifting block (LB) with friction plates from below, slightly lifting the bottom of the hopper (e.g., 1-2 mm) or applying a certain vertical clamping force to prevent the hopper from sliding during conveyor start / stop. All cylinder actions are controlled by high-speed solenoid valves (SV), with extremely short response times (e.g., <100ms). Sensors (e.g., magnetic sensors, MS) provide real-time feedback on the cylinder's positioning status. The locking action is completed instantaneously and held until the hopper is safely grabbed or pulled by the downstream conveyor mechanism.

[0143] After the pneumatic locking device (PLD) confirms the locking is in place (via feedback from its sensors), the system executes two key actions: 1) Releasing the steering mechanism: The servo motor (SM) of the three-axis steering mechanism (TASM) switches to zero-force or low-force holding mode, or directly releases the position holding, preparing for the material box to be moved out. 2) Outputting a standard material box with an orientation verification code: The control system generates a unique orientation verification code (OVC). This verification code typically includes: material box ID, calibration completion timestamp, final PCDS measurement value (Yaw, Pitch, Roll), measured included angle AD, preset angle PAT, locking status, etc. This code is attached to the material box's electronic tag (such as RFID) or sent to the MES system via a communication network bound to the material box ID. Simultaneously, a control signal is sent to the downstream conveying mechanism of the inbound line (such as a roller conveyor, RC, or a belt conveyor, BC) to release the brakes and smoothly deliver the material box out of the steering station at a standard speed. At this point, the delivered tote box is oriented precisely according to the workstation reference direction (WLRDV), its position is accurately positioned and locked, and it comes with an electronic certificate (OVC) proving its orientation is correct. Therefore, it is called a Standardized Tote with Orientation Verification Code (STOVC). The STOVC is transported to the entrance of the parallel workstation line, ready to enter the designated workstation unit for subsequent operations. The entire steering correction process is now complete.

[0144] The adaptive steering module uses multi-sensor fusion technology to detect the position and orientation of the material bins in real time. A three-axis steering mechanism dynamically adjusts the bin's direction, ensuring that all materials entering the production line maintain a uniform orientation and eliminating orientation deviations caused by manual loading. Standardized bin orientation is a prerequisite for automated flow. This design significantly reduces the failure rate of robotic arm gripping and shuttle docking, while also reducing the frequency of manual intervention, enabling truly fully automated production line operation.

[0145] S203 uses an infrared dual-mode sensor to detect the real-time position of the material box in the workstation unit, and combines the status change button command triggered by the worker to generate a status mark signal with verification.

[0146] Specifically, the detection area of ​​the dual-mode infrared sensor within the workstation unit can be calibrated based on the coordinate parameters of the three-zone isolation topology map.

[0147] After completing the physical partitioning of the workstation unit, the system first calls the three-zone isolation topology map (which stores the precise spatial coordinate range of the "Unstarted Zone," "In Progress Zone," and "Completed Zone" within each workstation unit). These coordinates define the boundaries of each zone with millimeter-level precision (e.g., the X-axis range of the Unstarted Zone is 0-1500 mm, and the Y-axis range is 0-800 mm). The dual-mode infrared sensor deployed on top of the workstation unit includes two operating modes: Passive IR (PIR) mode for detecting the surface temperature distribution of the material bin, and Active IR (AIR) mode for constructing a point cloud of the material bin contour by emitting and receiving infrared beams. The system converts the coordinate parameters in the three-zone isolation topology map into spatial mapping relationships in the sensor coordinate system. For example, it maps the physical partition boundaries (such as the rectangular vertex coordinates of the "In Progress Zone") to the sensor's field of view (FOV) through a coordinate transformation matrix. The calibration process uses the servo motor built into the sensor to fine-tune the pitch angle (PA) and yaw angle (YA) to ensure that the detection area completely coincides with the boundary of the physical partition, with the error controlled within ±5 mm.

[0148] Dynamic boundary verification is required during calibration. The system controls the sensors to perform a full-area scan of the empty workstation unit: in passive infrared mode (PIR), it detects whether the background thermal radiation distribution is uniform (excluding equipment thermal interference); in active infrared mode (AIR), it projects a light curtain onto the partition wall surface and verifies whether the time difference of flight (ToF) of the returned beam is consistent with the preset physical boundary distance. If a deviation is detected (e.g., the actual position of a zone boundary shifts by 10 mm), the system automatically triggers a compensation algorithm: it generates a compensation vector (CV) based on the offset direction (e.g., positive X-axis) and offset value (OV), and updates the sensor's spatial mapping table (SMT). After calibration, the detection area of ​​each zone is assigned a unique zone code (ZC), such as ZC01 for "not started zone", ZC02 for "in progress zone", and ZC03 for "completed zone", and is bound to the workstation unit's identification identifier (WID).

[0149] To address mechanical deformation or vibration interference during long-term operation, the system incorporates a periodic self-check process. Every 24 hours, or upon detecting an abnormal vibration event (triggered by the accelerometer built into the workstation unit), automatic recalibration is initiated: the sensors emit positioning beams towards calibration targets (with built-in reflective material) at the corners of each zone. New calibration parameters are calculated based on the deviation value (DV) between the feedback target coordinates (TC) and the theoretical coordinates (ThC). The calibration results are written to the non-volatile memory (NVM) of the workstation control unit (WCU) to ensure that parameters are not lost after power failure. Finally, the dual-mode infrared sensors of all workstation units establish a dynamic detection grid (DDG) that is strictly aligned with the physical zones.

[0150] Scan the spatial coordinates of the standard material bin within the workstation unit and output the material bin-zone location mapping table;

[0151] When a standard hopper with a Direction Check Code (DCC) (output from the inbound line) enters the workstation unit, the dual-mode infrared sensor initiates a collaborative scan: the active infrared mode (AIR) emits 32 fan-shaped infrared beams (wavelength 850 nm) to scan the hopper surface, generating approximately 5000 point cloud data (PCD); the passive infrared mode (PIR) simultaneously acquires the hopper surface temperature field (resolution 0.1 degrees Celsius). After filtering and denoising, the point cloud data is fitted with the hopper's six-degree-of-freedom pose (6-DoF Pose) using the Iterative Closest Point (ICP) algorithm, including three-dimensional coordinates (X, Y, Z) and rotation angles (Roll, Pitch, Yaw). For example, the system might detect the hopper's center coordinates as (1200 mm, 450 mm, 200 mm) with a Y-axis rotation angle of 5 degrees.

[0152] Pose data is input into the Zone Matching Engine (ZME). This engine calculates the spatial relationship between the 3D bounding box (BB) of the hopper and the dynamically detected mesh (DDG) defined by the three-zone isolation topology map.

[0153] Calculate the overlap area (OA) between the outer envelope and the boundaries of each partition;

[0154] Determine the zone code (ZC) where the center of gravity (CoG) of the hopper is located;

[0155] If the bin spans multiple zones (e.g., part in the "in progress zone" and part in the "completed zone"), the primary zone is determined based on whether the overlap ratio (OR) exceeds a preset threshold (e.g., 80%).

[0156] The matching results generate a Crate-Zone Mapping Table (CZMT), which includes the Crate ID (CID), the zone code (ZC), the confidence level (CL, ranging from 0-100%), and the timestamp (TS). For example: "CID00573→ZC02 (in progress zone), confidence level 95%, timestamp 2023-08-15 14:30:25".

[0157] To improve mapping accuracy, the system introduces multi-sensor fusion verification:

[0158] If the confidence level (CL) of the dual-mode infrared is less than 90%, the laser distance sensor (LDS) on the side wall of the workstation unit is triggered for secondary verification to measure the distance to the wall (DtW) between the material box and each partition boundary wall.

[0159] When the system detects that the hopper is in motion (determined by the velocity vector of the point cloud data), it starts the Motion Prediction Algorithm (MPA) to estimate its position at the next moment and update the mapping table.

[0160] The final output CZMT is uploaded to the Central Monitoring System (CMS) via the real-time data bus (RDB) of the workstation control unit (WCU).

[0161] Parse the worker button commands and the bin-zone location mapping table, and generate a status change permission flag when the operation conforms to the zone transfer rules;

[0162] The workstation's control panel features three-color physical buttons: green ("Not Started → In Progress"), yellow ("In Progress → Completed"), and red ("Emergency Revert"). When a worker presses a button, the Button Command Parser (BCP) captures the following information: Button Type (BT), Operator ID (OID), and Press Time (PT). The system immediately retrieves the latest bin-zone location mapping table (CZMT) for the current workstation to obtain the target bin's real-time assigned zone (CZ). For example, when a worker presses the green button, the system detects that the target bin is currently located in ZC01 (Not Started Zone).

[0163] The Zone Transfer Rule Validator (ZTRV) matches a preset finite state machine (FSM) to the button type (BT) and the current partition (CZ). This state machine defines the valid transition paths:

[0164] Only "ZC01→ZC02" (green button is active) is allowed;

[0165] Only "ZC02→ZC03" (yellow button is active) is allowed;

[0166] The red button forces a revert to the previous state (requires supervisor authorization).

[0167] If an illegal operation is detected (such as pressing the yellow button when the hopper is in ZC01), the system will immediately trigger an audible and visual alarm (Alarm Type, AT03) and display an error code (Error Code, EC: "ILLEGAL_TRANSFER") on the operation screen.

[0168] If the operation is valid, the system will further perform a secondary verification of the physical location:

[0169] Re-trigger the dual-mode infrared sensor scan to ensure the hopper has not moved during the operation interval;

[0170] Compare the rechecked zone (RZ) of the second scan with the current partition (CZ) in the mapping table, requiring consistency of over 98%;

[0171] Verify that the bin is completely detached from the original partition boundary (overlap area ratio OR < 5%).

[0172] Once all passes, the Rule Validation Engine (RVE) generates a Status Change Permission Flag (SCPF), which is a binary flag bit (FB=1) along with the operation context (button type BT, bin CID, workstation WID, timestamp TS).

[0173] Retrieve the historical status timestamp of the material bin, check whether the current operation interval meets the process threshold, and output the timing compliance assessment value;

[0174] After generating the Permit Flag (SCPF), the system accesses the Central Timing Database (CTDB) and retrieves the Lifecycle Status Record (LSR) by bin identification ID (CID). The record contains detailed information for each status change: original state (From State, FS), new state (To State, TS), change time (Change Time, CT), and operator ID (OID). For example, the history of bin CID00573 shows that its most recent status change was "from ZC01 to ZC02," with a change time of 2023-08-15 14:25:30.

[0175] The Temporal Compliance Analyzer (TCA) calculates the time interval (TI) between two critical operations:

[0176] If the current operation is "In Progress → Completed" (yellow button), calculate the difference between the current press time (PT) and the change time (CT) of the previous "Not Started → In Progress" (green button);

[0177] Process thresholds (PTH) are dynamically loaded from the bag manufacturing knowledge base (KB), for example, "minimum processing time for sewing = 300 seconds".

[0178] The calculation formula is: Temporal Compliance Evaluation Value (TCEV) = (Actual Interval TI / Process Threshold PTH) × 100%. If TCEV ≥ 100%, it is considered compliant (e.g., TI = 320 seconds, PTH = 300 seconds, then TCEV = 106.7%); if TCEV < 100%, it is considered non-compliant (e.g., TI = 280 seconds, TCEV = 93.3%).

[0179] For violations (TCEV < 100%), the system executes a tiered response:

[0180] If 90% ≤ TCEV < 100%, a yellow warning (Warning Level 2) is issued, requiring workers to confirm whether mandatory implementation is required;

[0181] If TCEV < 90%, a red alarm (Warning Level, WL3) is triggered, and the button is locked until the supervisor enters the Authorization Code (ACO).

[0182] The final output Time-of-Performance Compliance Assessment (TCEV) is tied to the operation request as a percentage, for example, "TCEV = 106.7%" means that the operation interval is reasonable.

[0183] It integrates the status change permission flag with the timing compliance assessment value, and adds a digital signature to generate a verified status marker signal.

[0184] The Signal Fusion Engine (SFE) receives two core inputs: a State Change Permit Flag (SCPF, containing operational legality information) and a Timing Compliance Assessment Value (TCEV, containing timing compliance information). The engine executes the following fusion logic:

[0185] If SCPF = 1 (legal operation) and TCEV ≥ 100% (timing compliant), generate the final permission code (FinalPermission Code, FPC = 0x01);

[0186] If SCPF = 1 but TCEV < 100% (timing violation), generate FPC = 0x02 (violation needs to be recorded);

[0187] If SCPF = 0 (illegal operation), ignore TCEV and generate FPC = 0x00 (change rejected).

[0188] At the same time, the engine associates the operation context: bin CID, workstation WID, button type BT, operator employee number OID, and timestamp TS.

[0189] To ensure data immutability, the system adds a lightweight digital signature (LDS) to the fusion results:

[0190] Concatenate the FPC with the context data into a string, for example:

[0191] FPC:01|CID:00573|WID:A12|BT:YELLOW|OID:9527|TS:20230815143025;

[0192] Use the Hardware Security Module (HSM) of the workstation control unit (WCU) to generate a hash value (HV = SHA-256 (string));

[0193] The HV is asymmetrically encrypted using the production line master key (MK) to generate a signature ciphertext (SC).

[0194] The signing process takes less than 50 milliseconds, ensuring real-time performance.

[0195] The final generated Verified Status Mark Signal (VSMS) contains three core fields:

[0196] Plaintext Packet (PP): Contains operation information such as FPC, CID, and WID;

[0197] Signature Ciphertext (SC): Used to verify data integrity and origin;

[0198] Verification Metadata (VM): Includes hash algorithm type (Hash Type, HT = "SHA-256") and key version number (Key Version, KV = "MKv2").

[0199] The signal is sent to the Manufacturing Execution System (MES) and the blockchain evidence storage module via Industrial Ethernet (IE), triggering the subsequent shuttle scheduling and evidence storage process.

[0200] The system uses an infrared sensor matrix to accurately track the three-dimensional coordinates of the material bin within the workstation, forming a dual verification mechanism with manual operation by the worker. This ensures that status changes conform to both physical location constraints and are confirmed by human intervention, preventing logistical chaos caused by accidental triggering. This human-machine collaborative verification mechanism guarantees automation accuracy while retaining necessary human intervention permissions, preventing the risk of sensor misjudgment and providing flexible operational space for handling special working conditions.

[0201] S204, in response to the verification result of the completed area in the status flag signal, the MES dynamic scheduling engine is triggered to allocate shuttle car task instructions to the central shuttle line and generate a material box transfer path.

[0202] Specifically, the completed area bin ID, workstation coordinates, and verification timestamp can be extracted from the verification status flag signal to construct a bin transfer trigger event;

[0203] When the system receives a Verified Status Tag Signal (VSTS) from a workstation, the signal has been digitally signed to ensure its integrity and tamper-proof nature. The Event Parsing Module (EPM) of the central dispatch system first decrypts and verifies the VSTS. After confirming the legitimate source of the signal, it extracts three core fields: Material Box Identification (MBID, unique code such as "MB20240515007"), Workstation Coordinates (WC, format "Production Line A-Workstation 3-X:1200mm, Y:450mm"), and Verification Timestamp (VTS, accurate to milliseconds such as "20240515143025.358"). The extraction process uses field positioning technology to accurately segment the data stream according to a preset signal structure template (including start identifier, length definition, and checksum). For example, the MBID field occupies bytes 6-21 of the signal stream and is parsed into a readable string using ASCII code; the WC field is converted into a three-dimensional value (X, Y, Z) in the factory's unified coordinate system, where the Z-axis is usually the default height value; the VTS field needs to be synchronized to the system's high-precision network clock (HNC) to ensure consistent time references across the entire plant. These fields together constitute the original documentation for the material bins entering the transfer process.

[0204] The extracted field data enters the Event Construction Engine (ECE). This engine first associates the work order information in the MES database, binding the MBID with the corresponding Production Order Number (PON) and Material Type (MT, such as "outer canvas component of a bag"). Next, it locates the specific Workstation Unit ID (WUID) and its corresponding Parallel Cluster Zone (PCZ) based on the WC. A crucial step is adding event attributes: the Trigger Type (TT) is marked as "Release in Completed Area," the Urgency Level (UL) is automatically calculated based on the remaining delivery time of the work order (e.g., 2 hours remaining is set to the highest level 5, 24 hours remaining is set to the medium level 3), and the VTS is attached as the event generation time. The final output is a structured Material Box Transfer Trigger Event (MBTTE), whose data packet format includes: {Event ID, MBID, PON, MT, WUID, PCZ, WC, VTS, TT, UL}. For example: "EVT20240515143025_001,MB20240515007,PO-1142,Canvas Assembly,WS3-08,ZONE-B,(1200,450,0),20240515143025.358,Completed Zone Release,UL3".

[0205] To ensure the validity of the event, the system performs double verification:

[0206] Spatiotemporal consistency verification: Compare the VTS and the local log records of the workstation to confirm that the material bin is indeed located in the completed area at that time point (to prevent signal replay due to network latency);

[0207] State Transition Compliance Verification: Query the historical state sequence of this MBID (e.g., "Not Started Area → In Progress Area → Completed Area") to verify that the current trigger conforms to the preset Finite State Machine (FSM) rules for state transitions. If the verification fails, an alarm event is generated and manual intervention is notified; if successful, the MBTTE is marked as "Verified" and pushed to the input queue of the MES Dynamic Scheduling Engine (MDSE). At this point, a bin transfer request with complete context information is formally established, providing accurate input for subsequent resource scheduling.

[0208] Based on the material bin transfer trigger event, collect the shuttle car location, aisle occupancy status, and workstation demand priority to generate a dynamic resource status matrix;

[0209] Upon receiving the MBTTE, the MES Dynamic Scheduling Engine (MDSE) immediately activates the Real-time Data Collector (RDC). The RDC acquires the latest factory status through three types of sensor networks:

[0210] Shuttle Vehicle Position (SVP): Each shuttle is equipped with an Ultra-Wideband Tag (UWBT) that communicates with UWB Anchors (UWBA, accuracy ±10cm) deployed on the roof of the factory building. It calculates and reports the vehicle's three-dimensional coordinates on the central shuttle line in real time (e.g., "SV02:X=3500mm,Y=0,Z=200mm").

[0211] Channel Occupation Status (COS): A Lidar Scanner (LS) is installed at a key node of the central shuttle. The scanning results are converted into channel segment status (e.g., "Segment L203: 85% occupancy") using the Occupancy Grid Algorithm (OGA). This is then cross-validated using the binary duty cycle signal (0 / 1) from the Infrared Beam Sensor (IBS).

[0212] Workstation Demand Priority (WDP): Extract the Task Queue Depth (TQD), Process Urgency Factor (PUF), and Equipment Readiness Status (ERS) of all workstation units from the MES work order pool, and calculate the dynamic priority using a Weighted Scoring Model (WSM) (e.g., Workstation 5 demand priority = 0.6 × TQD + 0.3 × PUF + 0.1 × ERS, score range 1-100).

[0213] The collected raw data is spatiotemporally aligned using a Situation Fusion Processor (SFP):

[0214] Map the SVP coordinates onto a virtual grid map (VGM, grid size 200mm×200mm) of the central shuttle line to mark vehicle positions;

[0215] Convert the COS data into a segment occupation flag (SOF) for the channel, such as "Segment ID: L201, Status: Idle (0) / Occupied (1) / Partially Occupied (0.6)";

[0216] Sort the WDPs by workstation unit number and generate a priority vector (PV, e.g., [WS3-08:78,WS5-12:92,...]).

[0217] The fused data is organized into a Dynamic Resource Situation Matrix (DRSM) with a fixed structure. This matrix is ​​a three-dimensional tensor.

[0218] Dimension 1: Shuttle vehicle resource list (vehicle ID, location, speed, load);

[0219] Dimension 2: Channel topology status (segment ID, length, occupancy rate, and direction of passage);

[0220] Dimension 3: Workstation demand distribution (workstation ID, priority, number of material boxes to be transferred).

[0221] For example, a matrix slice: "SV02@(3500,0,200)|L203:0.6|WS3-08:78".

[0222] To improve the real-time nature of decision-making, DRSM employs an incremental update mechanism (IUM):

[0223] The basic matrix is ​​fully refreshed every 5 seconds;

[0224] When an MBTTE event is triggered or a sensor detects a sudden change in state (such as a shuttle abruptly stopping or a sudden lane closure), a partial update is initiated. Version control (VC) is introduced during the update process; each modification generates a new matrix version number (e.g., DRSM_Ver2.15), ensuring the scheduling engine always makes decisions based on the latest situation. The final output DRSM carries a time validity label (e.g., "Valid period: 20240515143030-20240515143035"), providing a reliable instantaneous factory snapshot for path planning.

[0225] Using a dynamic resource situation matrix as input, a dynamic window algorithm is used to solve for the optimal path and output the initial path point sequence.

[0226] The path planning module inputs the DRSM and MBTTE (including the target workstation coordinates WC) into the Dynamic Window Approach (DWA). DWA first constructs the Motion Parameter Space (MPS):

[0227] Based on the shuttle's dynamics model (maximum speed V_max = 1.5 m / s, maximum acceleration a_max = 0.8 m / s²),... 2 Given a maximum angular velocity ω_max = 45° / s, calculate the feasible velocity window (VW) and angular velocity window (AVW) under the current state.

[0228] Uniform sampling (e.g., 100 sampling points) is performed in the two-dimensional space formed by VW×AVW to generate candidate velocity pairs {(v,ω)} (e.g., v=0.7m / s, ω=15° / s).

[0229] For each (v,ω), a trajectory prediction (TP) is simulated for the next 3 seconds, and discrete path points are calculated based on the vehicle's kinematic equations (time step Δt = 0.1s, 30 points in total).

[0230] Each predicted trajectory needs to be filtered through a triple evaluation function:

[0231] Obstacle Proximity (OP): Calculates the Euclidean distance (ED) between the trajectory point and obstacles (other vehicles, road-blocking equipment) in the DRSM. If it is less than the safe radius R_safe = 0.5m, the trajectory is eliminated.

[0232] Target Alignment (TA): Evaluates the azimuth deviation between the trajectory endpoint and the target workstation WC. The smaller the deviation, the higher the score (e.g., the cos(θ_diff) function).

[0233] Velocity Priority (VP): Encourages speeds close to the maximum permissible speed (e.g., the v / V_max ratio).

[0234] The overall score S = α × OP + β × TA + γ × VP (weights α = 0.5, β = 0.3, γ = 0.2), and the (v, ω) with the highest S is selected as the optimal control command. For example, the optimal solution is: v_opt = 1.2 m / s, ω_opt = 10° / s.

[0235] Generate the Initial Path Point Sequence (IPPS) based on the optimal (v,ω):

[0236] Starting from the current position of the shuttle (P0), recursively calculate subsequent points according to the equation of motion:

[0237] P_{k+1}.x=P_k.x+v_opt×cos(θ_k)×Δt;

[0238] P_{k+1}.y=P_k.y+v_opt×sin(θ_k)×Δt;

[0239] θ_{k+1}=θ_k+ω_opt×Δt.

[0240] The sequence length is determined by the prediction duration (e.g., 3 seconds corresponds to 30 points), and the point spacing is uniform (approximately 0.12m).

[0241] Example output: IPPS = [(3500,0),(3620,30),(3748,45),...,(target point)], with each point accompanied by a timestamp (e.g., t = 0.0s, 0.1s,...3.0s). This sequence represents the theoretically optimal path under ideal, undisturbed conditions.

[0242] The system detects spatiotemporal conflict points in the path point sequence and adjusts them through a virtual track allocation mechanism to generate the final material box transfer path.

[0243] The Conflict Detector (CD) performs spatiotemporal overlay analysis of IPPS and DRSM:

[0244] Spatial Conflict: Map IPPS waypoints to the DRSM's Virtual Grid Map (VGM) and check the occupancy status of the grid containing each point. If a grid is marked "occupied" (e.g., due to other vehicles or obstacles), mark the point as a Spatial Conflict Point (SCP).

[0245] Time Conflict: The absolute time T_abs = VTS + t_k of the vehicle's arrival at each waypoint is calculated based on the IPPS timestamp and compared with the predicted trajectories of other vehicles in DRSM. If two vehicles enter the same channel segment (spatial overlap area) in the same time period (e.g., T_abs ± 0.5s), it is marked as a Temporal Conflict Point (TCP).

[0246] The conflict detection results generate a conflict heatmap (CHM), for example, "Path point P15(4020,120)@t=2.4s: Conflict with SV03 in segment L215".

[0247] In response to the detected conflict, the Virtual Track Allocator (VTA) initiates an adjustment mechanism:

[0248] Time slot negotiation: Allocating exclusive time windows (ETW) to conflicting road segments. For example, if SV03 is required to release segment L215 at t=2.3-2.7s, the usage time of this vehicle SV02 is adjusted to t=2.8-3.1s.

[0249] Path replanning: Inserting avoidance path segments (APS) near conflict points. Common strategies include:

[0250] Speed ​​adjustment: Reduce speed before the conflict point (e.g., reduce v from 1.2m / s to 0.6m / s) to prolong the passage time;

[0251] Path offset: Generate detour points (e.g., the original path point (4020, 120) is offset to (4050, 100));

[0252] Waiting point insertion: Add a stationary point to the safe zone (e.g., stay at (3980,100) for t=2.3-2.7s).

[0253] The adjustment process employs the Iterative Relaxation Method (IRM) until the conflict is resolved.

[0254] The final output Material Box Transfer Path (MBTP) includes:

[0255] Spatial path sequence: Optimized path point coordinates (including avoidance points);

[0256] Time schedule: Target arrival time for each waypoint (accurate to 0.1 seconds);

[0257] Speed ​​control curve: segmented speed commands (e.g., "P0-P5: v = 1.2 m / s, P6-P10: v = 0.6 m / s");

[0258] Conflict resolution: Recorded track allocation protocol (e.g., "L215 segment usage right: SV02@t=2.8-3.1s").

[0259] For example, the complete path description is: "Starting from (3500,0), stopping at (3620,30)@t+1.0s → detouring at (4050,100)@t+2.8s → arriving at the target (4500,450)@t+4.2s". This path is sent to the shuttle actuators via the central control bus and simultaneously synchronized to the MES system as the basis for scheduling.

[0260] Based on real-time logistics data and equipment status, the MES engine uses a dynamic window algorithm for multi-objective path planning, comprehensively considering factors such as transportation efficiency, equipment load balancing, and process priority to generate the optimal material transfer plan. Dynamic scheduling breaks through the limitations of traditional fixed paths, enabling the production line to intelligently respond to abnormal situations such as equipment failures and urgent orders, thereby improving overall logistics efficiency.

[0261] S205, according to the material box transfer path, control the shuttle car to send the completed material box back to the cutting area via the exit line, simultaneously release the completed area space in the workstation unit, and trigger the blockchain storage module to record the entire process status data.

[0262] Specifically, the shuttle car operation can be controlled based on the final material box transfer path, and trajectory deviation alarm events can be generated in real time through the UWB positioning module;

[0263] Once the central scheduling system generates the Final Material Transfer Path (FMTP), this path contains a series of ordered path points (PPs), each with three-dimensional coordinates (X-axis, Y-axis, Z-axis height) and an expected arrival timestamp (TS). The control system sends the FMTP to the embedded controller (EC) of the shuttle vehicle (SV). After parsing the path instructions, the EC drives the shuttle's motion actuators, including the drive wheel servo motor (DWSM), steering servo (SS), and lifting hydraulic cylinder (LHC). As the shuttle travels along the preset path, its core positioning device is the Ultra-Wideband Positioning Module (UWB-PM). The module consists of at least four UWB Anchors (UAs) installed at the four corners of the workshop ceiling and a UWB Tag (UT) on the top of the shuttle. The UAs continuously emit nanosecond-level pulse signals, and the UTs receive the signals and calculate the distance (DIST) to each UA. The Time Difference of Arrival (TDOA) algorithm is used to calculate the shuttle's real-time three-dimensional position (RP), achieving a positioning accuracy of centimeter-level (e.g., ±3 cm).

[0264] The Trajectory Monitoring System (TMS) compares the shuttle's real-time position (RP) with the corresponding expected position (EP) in the FMTP multiple times per second (e.g., 10 times / second). It calculates the Euclidean Distance Deviation (EDD) and Time Progress Deviation (TPD) in the XY plane. If the EDD exceeds a preset Spatial Tolerance Threshold (STT, e.g., 15 cm) or the TPD exceeds a Temporal Tolerance Threshold (TTT, e.g., ±5 seconds), a Trajectory Deviation Alert Event (TDAE) is generated. The TDAE includes the alert level (AL, categorized as mild, moderate, and severe), deviation direction (DD), current coordinates, expected coordinates, and a timestamp. For example, when it is detected that the shuttle has deviated 20 cm eastward from the expected path due to slipping on oil stains on the ground (EDD=20cm>STT=15cm), the system immediately generates a medium alarm event (AL=Medium) and transmits it to the central control console and the shuttle EC via the workshop wireless LAN (WLAN).

[0265] The system performs a tiered response for different types of TDAEs:

[0266] Mild warning (AL = Low, EDD 5~15cm): EC automatically triggers the Micro-Correction Algorithm (MCA), which gradually corrects the path segment by finely adjusting the angle of the steering servo SS (e.g. ±2°);

[0267] Medium alarm (AL = Medium, EDD 15-30cm): The central dispatch system intervenes, freezes the current path instruction, and starts the Dynamic Replanning Module (DRM) to generate a local detour path within 2 meters based on the real-time environmental map (including the location of other devices);

[0268] High Alarm (AL = High, EDD > 30cm or collision risk): Immediately cut off power and trigger the Sound-Light Alarm (SLA), then wait for manual intervention.

[0269] All alarm events are recorded in the Operation Log (OL) for post-event analysis and algorithm optimization.

[0270] When the trajectory deviation alarm event is empty, the outbound steering module is triggered to rotate to the outbound direction and outputs the outbound calibration execution command;

[0271] After the shuttle arrives at the designated loading point in the Completed Area (CA) of the workstation unit, the robotic arm (RA) automatically transfers the Material Tote (MT) to the shuttle rack. At this time, the Track Monitoring System (TMS) continuously monitors the TDAE queue status. The system determines that the safe transfer conditions are met only when no new TDAEs are generated along the entire path from the loading point to the exit (i.e., the alarm event queue is empty) and all historical alarms have been cleared. At this point, the central controller sends a Preparation Command (PC) to the Exit Direction Module (EDM). The EDM is an electromechanical device located at the intersection of the central shuttle line and the exit line. Its core component is the Rotary Platform (RP), which carries the track and is driven by a High-Torque Servo Motor (HTSM).

[0272] The steering calibration process consists of three steps:

[0273] Orientation pre-alignment: The EDM's built-in magnetic encoder (ME) reads the current platform angle (CA) and compares it with the exit line baseline angle (ELBA) to calculate the required rotation angle difference (AD). The HTSM drives the platform to rotate at a low speed (e.g., 0.5 radians / second), switching to jog mode (JM) when approaching the target angle.

[0274] Precise Positioning: When the platform approaches the ELBA (e.g., within ±1° range), the Laser Alignment Sensor (LAS) is activated. The LAS emits a cross-line laser (CLL) onto the exit line track, detecting its overlap with a preset baseline marker (BM). If the deviation exceeds 0.1°, a micro-stepping motor (MSM) is triggered for sub-degree compensation (e.g., 0.05° / step).

[0275] Mechanical locking: After alignment (deviation ≤0.1°), the pneumatic locking pin (PLP) is inserted into the positioning hole (PH) of the platform base to achieve physical locking. At the same time, the limit switch (LS) sends an in-position signal (IPS).

[0276] Once the lock is established, EDM generates an Exit Calibration Execution Command (ECEC). ECEC contains the following key parameters:

[0277] Platform Status Code (PSC): For example, "PSC=200" indicates successful calibration;

[0278] Calibration Timestamp (CTS);

[0279] Laser Alignment Precision (LAP, e.g., 0.05°).

[0280] The instruction is sent synchronously to the shuttle EC and Exit Line Controller (ELC) via Industrial Ethernet (IE), paving the way for the shuttle to enter the exit line.

[0281] Based on the completion signal of the outgoing calibration execution command, the locking of the workstation unit partition wall is released, and a workstation space release signal is generated;

[0282] When the Workstation Control System (WCS) receives the ECEC command (and PSC=200), it confirms that the outbound line is ready. At this time, the WCS initiates the Space Release Sequence (SRS). First, the WCS checks the status of the completed area CA within the current workstation: it confirms that the hopper has been removed via the Infrared Sensor (IRS) (Empty Signal, ES), and verifies that there is no residue on the platform (load value ≤0.5kg) via the Pressure Sensor (PS).

[0283] The core operation for releasing space is unlocking the Liftable Partition Wall (LPW). The LPW consists of three sections:

[0284] The wall separating the unstarted area from the ongoing area (Wall-1, W1);

[0285] The partition walls between the ongoing and completed zones (Wall-2, W2);

[0286] The partition wall between the area and the external passageway has been completed (Wall-3, W3).

[0287] During the material bin transfer process, W3 is in the Locked-Down State (LDS), physically isolating CA from the passageway. Upon release, WCS sends an UnlockCommand (UC) to the Electric Actuator (EA) of LPW. EA first releases the Electromagnetic Lock (EML), then drives the actuator to raise W3 to a preset height (e.g., 1.2 meters) at a constant speed (e.g., 10 mm / s), forming the transfer passageway. An ultrasonic rangefinder (URF) embedded in the wall monitors the lifting height in real time, triggering a StopSignal (SS) when the target value is reached.

[0288] After W3 is fully raised, WCS performs two key operations:

[0289] State Reset: Switch the workstation cell state machine (SM) from "Completed" to "Ready", allowing workers to move the hopper from the In-Progress Area (IPA) into the CA;

[0290] Signal Generation: Based on the "lift confirmation" signal from the W3 position sensor (POS) and the "empty area confirmation" signal from the IRS, a Workstation Space Release Signal (WSRS) is synthesized. The WSRS includes:

[0291] Workstation ID (WID);

[0292] Release Timestamp (RTS);

[0293] Wall Status Code (WSC, e.g., "WSC=OPEN").

[0294] The signal is broadcast to the MES (Manufacturing Execution System) and the central shuttle scheduler, indicating that the workstation is ready to receive new tasks.

[0295] The system integrates trajectory deviation alarm events, workstation space release signals, and quality inspection data, and writes them into the blockchain for evidence storage through the BFT consensus mechanism.

[0296] The Blockchain Notarization Module (BNM) continuously monitors three types of data sources:

[0297] Track deviation alarm event TDAE (from TMS);

[0298] Workstation space release signal WSRS (from WCS);

[0299] Quality data (QD, from the online vision inspection system) includes bin ID, defect type (DT), and qualified flag (QF).

[0300] The Data Fusion Unit (DFU) associates these three types of data by Material ToteID (MTID) and encapsulates them into a Notarization Data Packet (NDP). Each NDP contains:

[0301] Metadata Header (MH): Version number, data source IP, time window;

[0302] Business Data Body (BDB): Key-value pairs for TDAE / WSRS / QD;

[0303] Integrity Check Code (ICC): A hash value (HashValue, HV) based on the SHA-256 algorithm.

[0304] The evidence storage process employs the Byzantine Fault Tolerance Consensus Mechanism (BFT-CM). BNM is connected to a private blockchain network consisting of four nodes (Node, N).

[0305] Production execution node (N1: deployed on the MES server);

[0306] Quality control node (N2: deployed in the quality inspection system);

[0307] Device IoT node (N3: deployed at the workshop gateway);

[0308] Audit evidence storage node (N4: deployed in the cloud).

[0309] When an NDP is sent to the blockchain network:

[0310] Proposal phase: N1, as the Proposer Node (PN), broadcasts the NDP to N2 / N3 / N4;

[0311] Voting phase: Each node verifies the validity of the NDP's ICC and the rationality of its business logic (such as whether the TDAE and WSRS timestamps conflict), and casts a vote of approval (YES) or rejection (NO).

[0312] Consensus reached: If ≥3 nodes (including N1) vote YES within the timeout (Timeout, TO, e.g., 2 seconds), consensus is reached; otherwise, a retry (Retry, RT) is initiated.

[0313] Once consensus is reached, the NDP is packaged into a block (B), and the block structure includes:

[0314] Block Header (BH): Parent Block Hash (PBH), timestamp, Merkle Root (MR);

[0315] Block Body (BB): A set of NDPs;

[0316] Consensus Proof (CP): Digital Signature (DS) of each node.

[0317] New blocks are appended to the ledger using Chained Encryption (CE): the hash value of the previous block (PH) is used as input to generate an immutable link. Finally, the system outputs a Blockchain Notarization Record (BNR), whose unique access address (UAA) is written back to the MES database for subsequent quality traceability or dispute arbitration.

[0318] When the shuttle performs transfer tasks, the system synchronously updates the workstation status and generates tamper-proof production traceability records. UWB precise positioning ensures traceability of the logistics trajectory, forming a complete digital twin data chain from raw materials to finished products. This design achieves deep integration of physical and information flows. Blockchain-based evidence storage provides a reliable basis for quality traceability, while the real-time space release mechanism significantly improves workstation utilization. The entire system enhances production capacity.

[0319] As can be seen, a four-segment physical layout channel is constructed, including the inbound line, the central shuttle line, the parallel workstation line, and the outbound line. Through the adaptive steering module of the inbound line, the material boxes input from the external cutting area are received and the material box direction is corrected to generate standard material boxes. Based on the real-time position detection of the material boxes in the workstation unit by the infrared dual-mode sensor, a status mark signal with verification is generated. In response to the verification result of the completed area in the status mark signal, the shuttle car task instruction is assigned to the central shuttle line to generate the material box transfer path. According to the material box transfer path, the shuttle car is controlled to send the completed material boxes back to the cutting area via the outbound line, and the completed area space in the workstation unit is released simultaneously. This enables precise orientation of material box flow and efficient reuse of workstation space, ensuring reliable traceability and real-time collaboration of production status.

[0320] Another embodiment of the present invention provides a layout system for a fully automated material handling production line for bag manufacturing, see [link to relevant documentation]. Figure 3 The system may include:

[0321] The construction module 301 is used to construct a four-segment physical layout channel including an inbound line, a central shuttle line, parallel workstation lines and an outbound line. The workstation lines are composed of multiple independent workstation units connected in parallel. Each workstation unit is divided into a physically isolated unstarted area, an in-process area and a completed area.

[0322] The receiving module 302 is used to receive the material box input from the external cutting area through the adaptive steering module of the inbound line and perform material box orientation correction to generate a standard material box with uniform orientation and aligned with the workstation line.

[0323] The detection module 303 is used to detect the real-time position of the material box in the workstation unit based on the infrared dual-mode sensor, and generate a verified status mark signal in combination with the status change button command triggered by the worker.

[0324] The allocation module 304 is used to respond to the verification result of the completed area in the status flag signal, trigger the MES dynamic scheduling engine, allocate shuttle car task instructions to the central shuttle line, and generate a material box transfer path.

[0325] The control module 305 is used to control the shuttle car to send the completed material box back to the cutting area via the exit line according to the material box transfer path, simultaneously release the completed area space in the workstation unit, and trigger the blockchain storage module to record the status data of the entire process.

[0326] This invention also provides a storage medium storing a computer program, wherein the computer program is configured to execute the steps in any of the above method embodiments when running.

[0327] Specifically, in this embodiment, the storage medium can be configured to store a computer program for performing the following steps:

[0328] S201, construct a four-segment physical layout channel including an inbound line, a central shuttle line, parallel workstation lines and an outbound line. The workstation lines are composed of multiple independent workstation units connected in parallel. Each workstation unit is divided into a physically isolated unstarted area, an in-process area and a completed area.

[0329] S202, through the adaptive steering module of the inbound line, receives the material box input from the external cutting area and performs material box orientation correction to generate a standard material box with uniform orientation and aligned with the workstation line.

[0330] S203 uses an infrared dual-mode sensor to detect the real-time position of the material box in the workstation unit, and combines the status change button command triggered by the worker to generate a status mark signal with verification.

[0331] S204, in response to the verification result of the completed area in the status flag signal, the MES dynamic scheduling engine is triggered to allocate shuttle car task instructions to the central shuttle line and generate a material box transfer path.

[0332] S205, according to the material box transfer path, control the shuttle car to send the completed material box back to the cutting area via the exit line, simultaneously release the completed area space in the workstation unit, and trigger the blockchain storage module to record the entire process status data.

[0333] This invention also provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to perform the steps in any of the above method embodiments.

[0334] Specifically, the aforementioned electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the aforementioned processor, and the input / output device is connected to the aforementioned processor.

[0335] Specifically, in this embodiment, the processor can be configured to perform the following steps via a computer program:

[0336] S201, construct a four-segment physical layout channel including an inbound line, a central shuttle line, parallel workstation lines and an outbound line. The workstation lines are composed of multiple independent workstation units connected in parallel. Each workstation unit is divided into a physically isolated unstarted area, an in-process area and a completed area.

[0337] S202, through the adaptive steering module of the inbound line, receives the material box input from the external cutting area and performs material box orientation correction to generate a standard material box with uniform orientation and aligned with the workstation line.

[0338] S203 uses an infrared dual-mode sensor to detect the real-time position of the material box in the workstation unit, and combines the status change button command triggered by the worker to generate a status mark signal with verification.

[0339] S204, in response to the verification result of the completed area in the status flag signal, the MES dynamic scheduling engine is triggered to allocate shuttle car task instructions to the central shuttle line and generate a material box transfer path.

[0340] S205, according to the material box transfer path, control the shuttle car to send the completed material box back to the cutting area via the exit line, simultaneously release the completed area space in the workstation unit, and trigger the blockchain storage module to record the entire process status data.

[0341] The above description, based on the embodiments shown in the figures, details the structure, features, and effects of the present invention. The above description is only a preferred embodiment of the present invention, but the present invention is not limited to the scope of implementation shown in the figures. Any changes made in accordance with the concept of the present invention, or equivalent embodiments modified to have equivalent changes, that do not exceed the spirit covered by the specification and figures, should be within the protection scope of the present invention.

Claims

1. A layout method of a full-automatic material flow production line for luggage production, characterized in that, The method comprises: constructing a four-section physical layout channel comprising an inbound line, a central shuttle line, a parallel station line, and an outbound line, wherein the station line is composed of multiple independent station units in parallel, and each station unit is divided into physically isolated non-starting area, in-process area, and completed area; receiving the material box input by the external cutting area through the adaptive steering module of the inbound line, performing direction correction of the material box, and generating a standard material box with unified direction and aligned with the station line; based on the real-time position detection of the material box in the station unit by the infrared dual-mode sensor, combining the state change button instruction triggered by the worker, and generating a state marking signal with verification; in response to the verification result of the completed area in the state marking signal, triggering the MES dynamic scheduling engine, assigning the shuttle car task instruction to the central shuttle line, and generating a material box transfer path; according to the material box transfer path, controlling the shuttle car to send the completed material box back to the cutting area through the outbound line, synchronously releasing the completed area space in the station unit, and triggering the blockchain storage module to record the full-process state data; the construction of the four-section physical layout channel comprising an inbound line, a central shuttle line, a parallel station line, and an outbound line, wherein the station line is composed of multiple independent station units in parallel, and each station unit is divided into physically isolated non-starting area, in-process area, and completed area, comprising: analyzing the bag process flowchart, extracting the material input / output relationship including sewing and assembly process, and generating a process dependency topology graph; according to the process dependency topology graph, dividing the production line into four sections of inbound buffer area, central shuttle channel, parallel station cluster, and outbound recovery area, and outputting a four-section spatial coordinate framework; in the station cluster of the four-section spatial coordinate framework, deploying a liftable partition wall and loading the physical coordinate parameters of the non-starting area / in-process area / completed area, and generating a three-area isolation topology graph; inputting the three-area isolation topology graph into the logistics simulation system, verifying that the material box transfer path does not conflict with the partition wall action, and outputting a four-section layout channel with isolation parameters.

2. The method of claim 1, wherein, the adaptive steering module of the inbound line receives the material box input by the external cutting area and performs direction correction of the material box, and generates a standard material box with unified direction and aligned with the station line, comprising: extracting the station line reference direction vector from the four-section layout channel with isolation parameters as the correction target reference; by 3D line laser scanning the surface of the inbound material box, extracting the corner point coordinates and the center of gravity offset, and generating a material box pose deviation matrix; according to the pose deviation matrix and the station reference direction, calculating the pitch / yaw compensation angle of the three-axis steering mechanism, and outputting a real-time steering control instruction set; driving the parallel steering mechanism to execute the control instruction set, and simultaneously generating a pose correction data stream through binocular vision feedback; when the included angle between the pose correction data stream and the reference direction is less than a preset angle, triggering the pneumatic locking device and outputting a standard material box with direction verification code.

3. The method of claim 2, wherein, the real-time position detection of the material box in the station unit based on the infrared dual-mode sensor, combining the state change button instruction triggered by the worker, and generating a state marking signal with verification, comprising: based on the coordinate parameters of the three-area isolation topology graph, calibrate the detection area of the dual-mode infrared sensor in the station unit; Scanning the spatial coordinates of the standard bin in the workstation unit, outputting a bin-partition position mapping table; Analyzing the worker button instruction and the bin-partition position mapping table, and verifying that the operation conforms to the partition transfer rule, to generate a state change permission flag; Calling the bin historical state timestamp, detecting whether the current operation interval conforms to the process threshold, and outputting a timing compliance evaluation value; Fusing the state change permission flag and the timing compliance evaluation value, and adding a digital signature to generate a state marker signal with verification.

4. The method of claim 3, wherein, In response to the verification result of the completed area in the state marker signal, triggering the MES dynamic scheduling engine, distributing a shuttle vehicle task instruction to the central shuttle line, and generating a bin transfer path, including: Extracting the completed area bin ID, workstation coordinates, and verification timestamp from the verified state marker signal to construct a bin transfer trigger event; According to the bin transfer trigger event, collecting the shuttle vehicle position, channel occupation state, and workstation demand priority to generate a dynamic resource situation matrix; Taking the dynamic resource situation matrix as input, using a dynamic window algorithm to solve the optimal path, and outputting an initial path point sequence; Detecting the space-time conflict points in the path point sequence, adjusting through a virtual track allocation mechanism to generate a final bin transfer path.

5. The method of claim 4, wherein, According to the bin transfer path, controlling the shuttle vehicle to send the completed bin back to the cutting area through the outbound line, synchronously releasing the completed area space in the workstation unit, and triggering the blockchain storage module to record the whole-process state data, including: Based on the final bin transfer path, controlling the shuttle vehicle to run, and generating a track offset alarm event in real time through the UWB positioning module; When the track offset alarm event is empty, triggering the outbound turning module to rotate to the outbound direction, and outputting an outbound calibration execution instruction; According to the completion signal of the outbound calibration execution instruction, releasing the lock of the workstation unit partition wall, and generating a workstation space release signal; Fusing the track offset alarm event, the workstation space release signal, and the quality inspection data, and writing into the blockchain storage record through the BFT consensus mechanism.

6. A layout system of a full-automatic material flow production line for luggage production, characterized in that, The system comprises: A construction module for constructing a four-segment physical layout channel comprising an inbound line, a central shuttle line, a parallel workstation line, and an outbound line, wherein the workstation line is composed of multiple independent workstation units in parallel, and each workstation unit is divided into physically isolated non-starting area, in-process area, and completed area; A receiving module for receiving the bins input by the external cutting area through the adaptive turning module of the inbound line and performing direction correction of the bins to generate standard bins with unified direction and aligned with the workstation line; A detection module for detecting the real-time position of the bins in the workstation unit based on the infrared dual-mode sensor, and combining the state change button instruction triggered by the worker to generate a state marker signal with verification; A distribution module for triggering the MES dynamic scheduling engine in response to the verification result of the completed area in the state marker signal, distributing a shuttle vehicle task instruction to the central shuttle line, and generating a bin transfer path; A control module for controlling the shuttle vehicle to send the completed bin back to the cutting area through the outbound line according to the bin transfer path, synchronously releasing the completed area space in the workstation unit, and triggering the blockchain storage module to record the whole-process state data; The construction module is specifically configured to: Analyzing the process flow chart of the luggage, extracting the material input / output relationship including sewing and assembly process, and generating a process dependence topology diagram; According to the process dependence topology diagram, the production line is divided into four sections: an incoming buffer zone, a central shuttle channel, a parallel station cluster, and an outgoing recovery zone, and a four-section spatial coordinate framework is output; In the station cluster of the four-section spatial coordinate framework, a liftable partition wall is deployed and loaded with physical coordinate parameters of the not-started area / ongoing area / completed area, and a three-area isolation topology diagram is generated; The three-area isolation topology diagram is input into a logistics simulation system to verify that the material box transfer path does not conflict with the partition wall action, and a four-section layout channel with isolation parameters is output.

7. A storage medium, characterized by The storage medium stores a computer program, wherein the computer program is configured to execute the method of any one of claims 1-5 when running.

8. An electronic device comprising a memory and a processor, characterized in that, The memory stores a computer program, and the processor is configured to execute the computer program to execute the method of any one of claims 1-5.

Images