Intelligent workstation vehicle for two-person collaborative assembly operation and data acquisition method

Abstract

The application discloses a kind of double-person collaborative assembly operation intelligent workstation and data acquisition method, belong to intelligent manufacturing technical field.Intelligent workstation vehicle includes mobile vehicle body, tool control unit, multiple instance industrial terminal, image acquisition unit and information reading device.The method comprises: creating first job session and second job session of process level isolation on the same multiple instance industrial terminal, respectively binding two operators;In response to tool taking event, torque data, tool taking and returning record and image segment index are bound to corresponding job session;Bidirectional index structure between video segment and torque record is constructed;Remote server generates traceability report according to bidirectional index response traceability request;Electronic terminal can play video or show torque details in response to traceability interface operation.The application solves the data confusion problem in double-person collaborative assembly operation, realizes the accurate separation of double-person data and full-link traceability.

Description

Technical Field

[0001] This invention relates to the field of intelligent manufacturing technology, specifically to an intelligent workstation vehicle for two-person collaborative assembly operations and its supporting data acquisition method, system and storage medium. Background Technology

[0002] In modern discrete manufacturing, especially in the assembly of large and complex products in fields such as aerospace, rail transportation, and construction machinery, a two-person collaborative assembly mode is often adopted due to the large size, heavy weight, and numerous processes of the parts. Two operators perform different assembly tasks at the same workstation and during the same time period. For example, one person is responsible for tightening the bolts on the left side, while the other person is responsible for tightening the bolts on the right side; or one person is responsible for assembling the main structure, while the other person is responsible for connecting the auxiliary wiring harness.

[0003] In this two-person collaborative work scenario, existing intelligent assembly data acquisition systems suffer from the following technical shortcomings: Traditional handheld torque wrenches or torque-controlled tools, while capable of collecting tightening torque data, typically only support single-account login, making it impossible to distinguish data generated by two operators simultaneously on the same terminal. When two operators alternate using the same tool or use different tools at the same time, torque data, tool retrieval records, and video footage are easily confused, leading to difficulties in tracing assembly quality. It becomes impossible to accurately determine which operator performed the tightening of a particular bolt, nor can video clips of specific processes be correlated with corresponding torque records. Therefore, how to accurately bind torque data, tool usage records, and video data to their respective operators and establish a bidirectional traceable index relationship between the data is a pressing technical problem that needs to be solved in this field. Summary of the Invention

[0004] The present invention aims to provide an intelligent workstation vehicle for two-person collaborative assembly operations and a data acquisition method to solve the technical problems of data confusion and the inability to establish bidirectional and accurate traceability between torque data and image fragments in the prior art.

[0005] To achieve the above objectives, the present invention provides the following technical solutions.

[0006] This invention provides a data acquisition method for two-person collaborative assembly operations, applied to a mobile workstation cart integrating an intelligent tool cabinet, a dual-account industrial terminal, and an image acquisition bracket. The method includes the following steps:

[0007] A first job session and a second job session are created simultaneously on the same dual-account industrial terminal. The first job session is bound to a first operator identity identifier, and the second job session is bound to a second operator identity identifier.

[0008] In response to a tool retrieval event, the operator account identifier of the currently retrieval tool is obtained, and the torque data, tool retrieval and return records, and image fragment index associated with the tool retrieval event are all bound to the first job session or the second job session corresponding to the operator account identifier.

[0009] The image segment index points to one or more video segments, and each video segment is associated with the identifier of at least one torque record generated within the time period it covers, forming a bidirectional index between the video segment and the torque record. The bidirectional index means that the metadata of each video segment contains a set of associated torque record identifiers, and the data structure of each torque record also contains the unique identifier of the video segment.

[0010] By running two independent work sessions simultaneously on the same dual-account industrial terminal, binding the identities of two different operators, and automatically assigning torque data, tool retrieval and return records, and video clip indexes to the corresponding sessions based on the account affiliation at the time of tool retrieval, precise separation of data from both operators is achieved, fundamentally solving the data ambiguity problem. Simultaneously, by establishing a bidirectional index between video clips and torque records, it becomes possible to quickly locate the relevant video frame from any torque record, and to trace back to all torque data generated within that time period from any video clip, significantly improving the traceability efficiency of assembly quality.

[0011] Furthermore, the method also includes: configuring each of the multiple independent compartments of the smart tool cabinet with an electronic lock and a tool presence sensor; unlocking at least one of the independent compartments authorized by the operator in response to any operator completing identity authentication through the dual-account industrial terminal; recording a first timestamp of the operator retrieving a tool from the independent compartment and a second timestamp of the operator returning the tool, and associating and storing the first timestamp, the second timestamp, and the operator's identity identifier in a tool retrieval and return log.

[0012] By combining electronic locks and tool presence sensors, automated management and precise recording of tool retrieval and return are achieved. Each operator can only unlock their authorized slots, preventing accidental tool removal or loss. The tool retrieval and return log records the specific time and operator identity of each retrieval and return, providing reliable data support for the entire tool lifecycle management.

[0013] Furthermore, the method for constructing the image segment index includes: controlling a camera mounted on the image acquisition bracket to continuously acquire video streams of the assembly area; dividing the video stream into multiple video segments according to a preset process switching signal or a segmentation command manually triggered by the operator; assigning a unique segment identifier to each video segment and recording the start and end timestamps of the video segment; querying a torque database to extract the identifiers of all torque records between the start and end timestamps, forming a set of associated torque record identifiers for the video segment; writing the set of associated torque record identifiers into the metadata field of the video segment, and simultaneously adding the unique segment identifier of the video segment to the data structure of each torque record.

[0014] The continuous video stream is intelligently segmented into meaningful video segments using process switching signals or manual segmentation commands, avoiding the storage of large amounts of irrelevant video data. Each video segment is automatically associated with all torque record identifiers within its time window, and each torque record is also back-annotated with its corresponding video segment identifier, forming a complete bidirectional index structure. This design allows quality inspectors to jump to the corresponding video screen with one click when viewing torque data for a certain process, intuitively confirming the compliance of the operation; when reviewing monitoring videos, they can also instantly understand which bolts were tightened during that time period and whether the torque values ​​were up to standard.

[0015] Furthermore, the method also includes: when the network connection between the mobile workstation and the remote server is interrupted, storing the torque data to be uploaded, tool retrieval records, and image fragment indexes into a local cache database; assigning a globally unique sequence number to each piece of data to be uploaded; after the network connection is restored, uploading the data to be uploaded one by one in ascending order of the sequence numbers; receiving an upload success confirmation message returned by the remote server, the confirmation message containing the sequence number range of the successfully received data; deleting the confirmed uploaded data from the local cache database according to the sequence number range, and retransmitting the data for which no confirmation message has been received.

[0016] By employing local caching and a serial number-based breakpoint resumption mechanism, data is guaranteed to be neither lost nor duplicated in workshop environments with unstable networks or temporary network outages. A globally unique serial number ensures the sequentiality and integrity of uploaded data, while a confirmation mechanism based on a serial number range enables efficient batch confirmation and precise retransmission control.

[0017] This invention also provides an intelligent workstation vehicle for two-person collaborative assembly operations, comprising:

[0018] Mobile vehicle body;

[0019] An intelligent tool cabinet is fixedly installed on the mobile vehicle body. The intelligent tool cabinet contains multiple independent compartments, and each independent compartment is equipped with an electronic lock and a tool presence sensor.

[0020] A dual-account industrial terminal is fixedly installed on the mobile vehicle and communicates with the intelligent tool cabinet. The dual-account industrial terminal is configured to run a first job session and a second job session simultaneously.

[0021] An image acquisition bracket is fixedly installed on the mobile vehicle body. A camera is installed at the upper end of the image acquisition bracket, and the camera is communicatively connected to the dual-account industrial terminal.

[0022] The barcode scanner is connected to the dual-account industrial terminal.

[0023] By integrating modules such as intelligent tool cabinets, dual-account industrial terminals, image acquisition brackets, and barcode scanners onto a single mobile vehicle, a one-stop collaborative assembly workstation for two workers is created. The mobile vehicle allows the workstation to be flexibly deployed to different assembly stations, eliminating the need for redundant infrastructure construction at each station. The dual-account industrial terminal ensures data separation between the two workers at the hardware level.

[0024] Furthermore, the dual-account industrial terminal includes: a first display area for displaying the user interface of the first job session; a second display area for displaying the user interface of the second job session; a central processing unit configured to execute task scheduling, data acquisition, and data binding logic for the first job session and the second job session; and a memory module for caching the running data of the first job session and the second job session.

[0025] Two operators can simultaneously view their respective interfaces through two physically or logically separated display areas without interfering with each other. The central processing unit uses multi-threading technology to handle task scheduling and data acquisition for both sessions simultaneously, ensuring real-time performance and concurrency. The memory module provides independent cache space for the two sessions to avoid data conflicts.

[0026] Furthermore, the smart tool cabinet also includes: a tool identification tag reader, which is set on the front panel of the smart tool cabinet, for reading radio frequency identification tags or QR code tags affixed to the tools; and a compartment status indicator light, which is set for each independent compartment and is used to indicate the locked status, idle status, or tool not returned after timeout of the independent compartment.

[0027] The tool identification tag reader can automatically identify the type and number of tools placed or removed, eliminating the need for manual input by the operator and improving operational efficiency and accuracy. The slot status indicator lights visually display the current status of each slot through different colors or flashing patterns, allowing operators to easily locate available tools or identify tools that have not been returned within the specified time.

[0028] Furthermore, the image acquisition bracket includes: a vertical pole, the lower end of which is fixed to the mobile vehicle body; a horizontal arm, one end of which is connected to the upper end of the vertical pole via a universal joint; a gimbal, which is fixedly installed at the other end of the horizontal arm, and the camera is mounted on the gimbal; the gimbal has a built-in pitch motor and a horizontal rotation motor for adjusting the shooting angle of the camera.

[0029] The universal joint allows the pan / tilt arm to be adjusted in multiple degrees of freedom, adapting to the shooting needs of operators of different heights and in different assembly positions. The built-in motor of the pan / tilt head enables remote or automated adjustment of the shooting angle, automatically aligning with key assembly areas in conjunction with process switching signals.

[0030] This invention also provides a data acquisition system for dual-person collaborative assembly operations, comprising: a dual-person collaborative assembly intelligent workstation vehicle as described above; a remote server connected to the dual-account industrial terminal via network communication; and a database server connected to the remote server for storing torque data, tool retrieval and return records, image clip indexes, and operator identity information; the dual-account industrial terminal has functions of dual-person identity verification, tool control, assembly torque acquisition, synchronous recording of work images, binding of work data to different accounts, bidirectional indexing of video and torque records, network outage caching, and breakpoint resume transmission.

[0031] By combining the intelligent workstation vehicle with remote servers and database servers to form a complete end-to-end system, a closed-loop process from on-site data collection to cloud storage and analysis has been achieved.

[0032] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method described above.

[0033] By using a computer-readable storage medium, the data acquisition method of the present invention is solidified into executable program code, which facilitates deployment and operation on different hardware platforms and improves the versatility and portability of the technical solution.

[0034] Compared with the prior art, the present invention has the following beneficial effects:

[0035] (1) By running two independent work sessions simultaneously on the same dual-account industrial terminal, binding the identity identifiers of two operators respectively, and automatically allocating torque data, tool pick-up and return records and image fragment indexes to the corresponding sessions according to the account ownership when the tool is picked up, the precise separation of dual-person data is achieved, and the probability of data confusion between dual-person operations is reduced by more than 80%.

[0036] (2) By establishing a bidirectional index between video clips and torque records, each video clip is associated with all torque record identifiers within its time window, and each torque record also points back to its respective video clip, realizing one-click bidirectional tracing from torque value to video frame and from video frame to torque data, improving the efficiency of quality problem tracing by more than 80%.

[0037] (3) Through the combination of electronic lock and tool in-situ sensor, the automated management and accurate recording of tool retrieval and return are realized, and the completeness rate of tool retrieval and return records is increased to over 99%.

[0038] (4) By using local caching and serial number breakpoint resume mechanism, data is not lost or duplicated in unstable network environment, and the data preservation integrity rate is increased to more than 99%. Attached Figure Description

[0039] Figure 1 This is a schematic diagram of the overall system architecture of the intelligent workstation vehicle for two-person collaborative assembly operations according to the present invention.

[0040] Figure 2 This is a flowchart of the data acquisition method for two-person collaborative assembly operations according to the present invention.

[0041] Figure 3 This is a timing diagram of the tool's retrieval and return control process according to the present invention. Detailed Implementation

[0042] The technical solution of the present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. It should be understood that the preferred embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the invention.

[0043] Example 1: System Overall Architecture

[0044] like Figure 1 As shown in the figure, the present invention provides an intelligent workstation vehicle and data acquisition system for dual-person collaborative assembly operations, including a mobile vehicle body, a tool control unit (i.e., an intelligent tool cabinet), a multi-instance industrial terminal (i.e., a dual-account industrial terminal), an image acquisition unit (i.e., an image acquisition bracket and a camera), an information reading device (i.e., a barcode scanner), and a remote server and a database server.

[0045] The mobile cart is constructed from welded stainless steel square tubing and features four swivel casters at the bottom, two of which are equipped with brakes to ensure stable parking during operation. The work surface is covered with anti-static material to prevent damage to electronic components from static electricity. Storage drawers are located beneath the work surface for storing spare parts and auxiliary tools.

[0046] The tool control unit is fixedly mounted on the platform of the mobile vehicle. The casing of the tool control unit is made of cold-rolled steel sheet with a rust-proof paint coating. The front panel of the tool control unit has multiple independent compartments, each equipped with an electronic lock and a tool presence sensor. The electronic lock uses an electromagnetic lock cylinder, and its locking and unlocking are controlled by a multi-instance industrial terminal. The tool presence sensor uses a microswitch or photoelectric sensor; it triggers a closing signal when a tool is placed in the compartment and a closing signal when the tool is removed.

[0047] The multi-instance industrial terminal is fixedly mounted on the platform of the mobile vehicle, located to one side of the tool control unit. The multi-instance industrial terminal uses an industrial-grade tablet PC or embedded industrial computer, equipped with a touch screen. The multi-instance industrial terminal communicates with the tool control unit via an RS485 bus or Ethernet to transmit control commands for the electronic lock and status signals from the tool's on-premises sensor. Internally, the multi-instance industrial terminal runs a multi-instance operating system or application, and through containerization technology or cgroups and namespaces process isolation technology, it can simultaneously create and maintain two independent job sessions.

[0048] The image acquisition unit is fixedly mounted on the platform of the mobile vehicle, located behind the tool control unit. A camera is installed on the top of the image acquisition unit, which communicates with the multi-instance industrial terminal via a USB or Ethernet interface to acquire video images of the assembly area in real time.

[0049] The information reading device communicates with multiple industrial terminals via USB interface or Bluetooth wireless connection, and is used to scan product barcodes, operator badge QR codes, and tool QR codes.

[0050] The remote server communicates with multiple industrial terminals via the workshop LAN, and the database server communicates with the remote server to store torque data, tool retrieval and return records, image clip indexes, and operator identification information.

[0051] Example 2: Detailed Structure of the Tool Control Unit

[0052] like Figure 1As shown, the internal structure of the tool control unit includes multiple independent compartments, the size of which is customized according to the specifications of the tools stored. For commonly used electric and pneumatic wrenches, the internal dimensions of the independent compartment are set to 400mm × 300mm × 200mm; for smaller torque wrenches and sockets, the internal dimensions of the independent compartment are set to 250mm × 200mm × 150mm.

[0053] Each individual compartment is equipped with an electronic lock at its front opening. The electronic lock consists of a latch and an electromagnet. When the multi-instance industrial terminal outputs an unlock signal, the electromagnet is energized and engages, causing the latch to retract, allowing the operator to open the compartment door to retrieve tools. When a lock signal is output, the electromagnet is de-energized and releases, the latch extends, and the compartment door is locked. The electronic lock's status feedback signal is returned to the multi-instance industrial terminal via a digital input interface to confirm the lock's actual status.

[0054] The tool presence sensor is installed at the bottom of an independent compartment. When a tool is placed in the compartment, its weight presses against the sensor contacts, causing the sensor to output a high-level signal. When the tool is removed, the sensor contacts reset, outputting a low-level signal. The multi-instance industrial terminal periodically polls the status of all tool presence sensors, recording the time of change and the corresponding compartment number when a status change is detected.

[0055] The front panel of the tool control unit is also equipped with a tool identification tag reader. This reader uses a high-frequency RFID read / write module or a QR code scanning module to read RFID tags or QR code tags affixed to the tool handle. When the operator brings a tool close to the reader, the reader automatically identifies the tool's unique code and sends this code information to a multi-instance industrial terminal to confirm the type and serial number of the tool being handled.

[0056] Each individual compartment has a corresponding compartment status indicator light on its outer side. The compartment status indicator light uses RGB tri-color LEDs, and different colors and flashing patterns indicate different states: solid green indicates that the compartment is idle and available for use; solid red indicates that the compartment is locked and contains tools; flashing orange indicates that the tools have not been returned within the time limit; and off indicates that the compartment is disabled.

[0057] Example 3: Detailed Structure of a Multi-Instance Industrial Terminal

[0058] like Figure 1 As shown, the front view of the multi-instance industrial terminal displays its display and interaction areas. The touch screen of the multi-instance industrial terminal is divided into a first display area and a second display area. The two display areas can be of equal size or adjusted according to actual needs. For example, the first display area occupies 50% of the left side of the screen, and the second display area occupies 50% of the right side of the screen.

[0059] The first display area is used to present the user interface of the first work session, including current operator information, current process information, real-time torque data display, tool status information, and alarm information. The second display area is used to present the user interface of the second work session, and its content and structure are symmetrical to the first display area.

[0060] The multi-instance industrial terminal integrates a central processing unit (CPU). The CPU uses an industrial-grade ARM or x86 architecture processor with a clock speed of at least 1.8GHz and multi-core parallel processing capabilities. The CPU runs a real-time operating system or a Linux operating system, and uses multi-threading technology to simultaneously handle task scheduling, data acquisition, and data binding logic for both the first and second job sessions. The CPU executes the task scheduling, data acquisition, and data binding logic for both job sessions by calling the operating system's underlying task scheduling interface, data acquisition driver interface, and data binding application interface.

[0061] The multi-instance industrial terminal also integrates a memory module, which uses DDR4 or LPDDR4 industrial-grade memory chips with a capacity of no less than 8GB. The memory module is divided into two independent cache areas, serving the first job session and the second job session respectively, ensuring that the running data of the two sessions do not interfere with each other.

[0062] The multi-instance industrial terminal features multiple interfaces on its side, including an Ethernet interface, a USB interface, an RS485 interface, an HDMI interface, and an audio interface. The Ethernet interface connects to the workshop local area network (LAN) for data communication with a remote server. The USB interface connects to barcode scanners, cameras, and other USB peripherals. The RS485 interface connects to the electronic lock and tool presence sensor of the tool control unit.

[0063] Example 4: Detailed Structure of the Image Acquisition Unit

[0064] like Figure 1 As shown, the image acquisition unit includes a pole, a cross arm, a universal joint, and a pan-tilt unit.

[0065] The upright is made of aluminum alloy round tubing with an outer diameter of 40mm and a wall thickness of 3mm. Its height is adjustable from 800mm to 1200mm. The lower end of the upright is fixed to the platform of the mobile vehicle via a flange, which is secured to the platform with four bolts. The upper end of the upright has threaded holes for connecting a universal joint.

[0066] The crossarm is also made of aluminum alloy round tube, and its length is adjustable from 500mm to 800mm. One end of the crossarm is connected to the upper end of the pole via a universal joint, and the other end is fixed to the pan-tilt head with screws.

[0067] The universal joint uses a ball joint structure, allowing the crossarm to be adjusted to any angle in the horizontal and vertical directions. The locking handle of the universal joint is used to fix the angle after adjustment and prevent loosening.

[0068] The gimbal is a two-axis motorized gimbal with a built-in pitch motor and a horizontal rotation motor. The pitch motor has a rotation range of -90 degrees to +90 degrees, and the horizontal rotation motor has a rotation range of -180 degrees to +180 degrees. The bottom of the gimbal is fixed to the horizontal arm via a standard 1 / 4-inch screw hole, and the top is equipped with a camera via a quick-release plate. The gimbal connects to a multi-instance industrial terminal via an RS485 bus or PWM signal cable, receiving remote control commands to achieve automatic adjustment of the camera's shooting angle.

[0069] The camera is an industrial-grade network camera with a resolution of at least 1920×1080 pixels and a frame rate of at least 30fps, supporting H.264 or H.265 video encoding formats. The camera has a built-in infrared LED, ensuring clear imaging even in low-light conditions. The camera connects to multiple industrial terminals via Ethernet cables, transmitting video streams in real time.

[0070] Example 5: Detailed Process of Data Acquisition Method

[0071] like Figure 2 As shown in the figure, the data acquisition method for two-person collaborative assembly operations provided by this embodiment of the invention includes the following steps.

[0072] Step 501: Start the multi-instance industrial terminal and load the operating system and applications. After the application initializes, create a first display area and a second display area on the touch screen, ready to receive login requests from the first operator and the second operator, respectively.

[0073] Step 502: The first operator scans the QR code on their personal work badge using a barcode scanner, or enters their username and password in the first display area to complete identity authentication. After verifying the operator's identity, the multi-instance industrial terminal creates a first job session in the first display area and binds the first operator's first identity identifier to the first job session. Similarly, the second operator completes identity authentication in the second display area, creates a second job session, and binds a second identity identifier.

[0074] Step 503: The multi-instance industrial terminal determines its own list of authorized independent lock slots based on the permission configurations of the first and second operators. For example, the first operator is authorized to use five lock slots numbered 001 to 005, and the second operator is authorized to use five lock slots numbered 006 to 010. The multi-instance industrial terminal sends an authorization configuration command to the tool control unit, and the tool control unit updates the authorization status of each electronic lock according to the command.

[0075] Step 504: When the first operator needs to use a tool, they click the "Get Tool" button on the first display area, or directly pull open the door panel corresponding to the authorized compartment. After detecting the first operator's tool request, the multi-instance industrial terminal sends an unlocking command to the tool control unit, unlocking one or more independent compartments authorized to the first operator. At the same time, the multi-instance industrial terminal records the current time as the first timestamp for tool retrieval, and records the first operator's primary identification and the compartment number retrieved.

[0076] Step 505: The first operator removes the tool from the unlocked independent slot. The tool in-place sensor detects that the tool has been removed, and the status signal changes from high to low. After the multi-instance industrial terminal polls for this status change, it confirms again that the tool has been removed and records the precise time of tool removal.

[0077] Step 506: The first operator uses tools to perform assembly operations, such as tightening bolts. The tool's built-in torque sensor or an externally connected torque acquisition module collects tightening torque data in real time, including torque value, angle value, and tightening time. The torque data is transmitted to the multi-instance industrial terminal via wired or wireless means. Based on the currently active job session (the first job session in this case), the multi-instance industrial terminal automatically binds the received torque data to the first operator's primary identity identifier and stores it in a local database or directly uploads it to a remote server.

[0078] Step 507: During the assembly process, the camera continuously captures video streams from the assembly area. The multi-instance industrial terminal divides the video stream into multiple video segments based on preset process switching signals or segmentation commands manually triggered by the operator. Process switching signals can originate from process change commands issued by the MES system or be manually triggered by the operator via the "Next Process" button on the touchscreen. Each video segment is assigned a unique segment identifier, such as a UUID format string, and its start and end timestamps are recorded.

[0079] Step 508: The multi-instance industrial terminal queries the torque database to extract the identifiers of all torque records between the start and end timestamps, forming a set of associated torque record identifiers for the video segment. This set of associated torque record identifiers is written into the metadata field of the video segment, and a unique segment identifier for the video segment is added to the data structure of each torque record, forming a bidirectional index relationship. Specifically, the set of associated torque record identifiers for video segment V1 is {R1, R2, R3}, with the inverted index of torque record R1 pointing to V1, the inverted index of torque record R2 pointing to V1, and the inverted index of torque record R3 pointing to V1.

[0080] Step 509: After the first operator finishes using the tool, they return it to its original independent slot. The tool presence sensor detects the tool being placed in, and the status signal changes from low to high. The multi-instance industrial terminal records the current time as the second timestamp for tool return and records the tool return event. If the tool is not returned within the timeout period (exceeding a preset timeout threshold, such as 30 minutes), the multi-instance industrial terminal controls the slot status indicator light to flash orange to indicate this.

[0081] Step 510: The multi-instance industrial terminal packages all data generated by the first operator during this operation, including torque data, tool retrieval and return records, image segment indexes, and operator identification, and uploads it to the remote server. After the upload is complete, the data in the local cache is cleared based on the server's confirmation message.

[0082] Example 6: Detailed Implementation of the Breakpoint Resume Mechanism

[0083] The breakpoint resume mechanism provided in this embodiment of the invention includes the following steps.

[0084] Step 601: Multiple industrial terminals periodically check the network connection status with the remote server. The detection methods include sending heartbeat packets or checking the TCP connection status. If no response is received after three consecutive heartbeat packets, the network connection is considered interrupted.

[0085] Step 602: After the network connection is interrupted, the multi-instance industrial terminal stores the newly generated torque data, tool retrieval records, and image segment indexes into a local cache database. The local cache database uses a lightweight SQLite database and is stored on the solid-state drive of the multi-instance industrial terminal.

[0086] Step 603: Assign a globally unique sequence number to each piece of data to be uploaded. The sequence number is a 64-bit integer, generated as follows: the high 32 bits are the device number, and the low 32 bits are the locally incrementing sequence number. For example, for device number 0001, the sequence number of the first piece of data is 0x0001000000000001, the sequence number of the second piece of data is 0x0001000000000002, and so on.

[0087] Step 604: After the network connection is restored, the multiple industrial terminals upload the data to be uploaded one by one in ascending order of serial number. The upload protocol uses HTTPS POST request, and the data body is encapsulated in JSON format, containing the serial number and the specific data content.

[0088] Step 605: After receiving the uploaded data, the remote server performs data verification and deduplication. If the data verification passes and the sequence number has not been processed, the data is stored in the database server, and a successful upload confirmation message is returned to the multi-instance industrial terminals. The confirmation message contains the sequence number range of the successfully received data, for example, {"confirmedRange":"0x0001000000000001-0x0001000000000020"}.

[0089] Step 606: After receiving the confirmation message, the multi-instance industrial terminal deletes the confirmed uploaded data from the local cache database according to the serial number range. For data for which no confirmation message has been received (serial number not within the confirmation range), the multi-instance industrial terminal adds it to the retransmission queue and re-initiates the upload request.

[0090] Step 607: Repeat steps 604 to 606 until all data to be uploaded in the local cache database has been successfully confirmed for upload.

[0091] Example 7: Detailed Implementation of Bidirectional Indexing of Image Segments and Torque Records

[0092] During the assembly process, cameras continuously capture video streams, which are then received and buffered in real time by multiple industrial terminals. When a process change signal is detected, the multiple industrial terminals perform the following operations.

[0093] The first step is to end the recording of the current video segment, record the end timestamp, and assign a unique segment identifier to the current video segment. The segment identifier is in the format of "VIDEO_" + device number + "" + date and timestamp + "" + random number, for example, "VIDEO_0001_20260610_143025_8372".

[0094] The second step is to save the current video clip as an MP4 file, with the filename being the clip identifier plus the ".mp4" suffix.

[0095] The third step is to query the torque database and extract the identifiers of all torque records between the start and end timestamps. An example of the SQL query is as follows:

[0096] SELECT torque_record_id FROM torque_data

[0097] WHERE tightening_time>='2026-06-10 14:25:00'

[0098] AND tightening_time<='2026-06-10 14:30:25'

[0099] AND device_id='0001';

[0100] The fourth step is to convert the obtained torque record identifier set into a JSON array format, for example, ["TR_0001_20260610_142530_001", "TR_0001_20260610_142532_002", "TR_0001_20260610_142535_003"].

[0101] The fifth step is to write the JSON array into the metadata field of the video clip. The metadata can be stored in the user data box of the MP4 file, or it can be stored in a separate XML or JSON file along with the video file.

[0102] Step 6: Iterate through each torque record obtained from the query and add a unique segment identifier for the video clip to its data structure. An example of the update SQL statement is as follows:

[0103] UPDATE torque_data SET related_video_segment_id='VIDEO_0001_20260610_143025_8372'

[0104] WHERE torque_record_id IN ('TR_0001_20260610_142530_001', 'TR_0001_20260610_142532_002', 'TR_0001_20260610_142535_003');

[0105] Step 7: Begin recording a new video clip and record a new start timestamp.

[0106] Through the above steps, a bidirectional index relationship was established between video clips and torque records. In subsequent quality traceability processes, quality inspectors can query the data using the following two methods:

[0107] Method 1: Reverse lookup of video clips from torque records. Quality inspectors select a torque record in the system. Based on the "related_video_segment_id" field in the torque record, the system quickly locates the corresponding video clip file and automatically jumps to the vicinity of the tightening time corresponding to the torque record in that video clip for playback.

[0108] Method 2: Query torque records from video clips. Quality inspectors select a video clip in the system. The system parses the metadata of the video clip, obtains a set of associated torque record identifiers, and then queries the torque database for the details of the torque records corresponding to these identifiers, displaying them in a list or chart format.

[0109] Example 8: Detailed Implementation of Multi-Instance Terminal Session Management

[0110] Multi-session management of multi-instance industrial terminals is achieved through operating system-level process isolation or containerization technology. Specifically, the multi-instance industrial terminal uses the Linux operating system and creates two independent container environments through cgroups and namespaces technologies. Each container environment runs a complete application instance, corresponding to the first job session and the second job session, respectively.

[0111] Each job session's application instance includes the following functional modules:

[0112] Identity authentication module: Responsible for operator login authentication and permission verification. Supports multiple authentication methods, including QR code scanning, RFID card reading, and username / password input.

[0113] Task Management Module: This module receives assembly tasks from the MES system, including process lists, process parameters, and quality standards. It breaks down the tasks into specific operational steps and guides the operator through the process step-by-step via the user interface.

[0114] Data acquisition module: responsible for communicating with hardware devices such as torque tools, tool in-situ sensors, cameras and barcode scanners to collect torque data, tool status, video streams and barcode information in real time.

[0115] Data binding module: responsible for binding the collected data according to the currently active job session and operator identity to ensure the correctness of data ownership.

[0116] Data storage module: responsible for writing the bound data to the local cache database or uploading it directly to the remote server.

[0117] User interface module: Responsible for presenting the operation interface in the corresponding display area, including task progress, torque data curve, tool status indication and alarm information.

[0118] Data between the two work sessions is completely isolated and invisible to each other. However, in certain scenarios, such as when it is necessary to coordinate the work rhythm of two operators, multi-instance industrial terminals can transmit limited information between the two sessions through inter-process communication mechanisms, such as the number of operation steps completed by the other party, but will not transmit specific torque data and tool retrieval records.

[0119] Example 9: Detailed Implementation of the Tool Retrieval and Return Process

[0120] like Figure 3 As shown, the specific procedure is as follows when an operator needs to collect tools.

[0121] The first step involves the operator standing in front of the tool control unit and initiating a tool request through the corresponding display area of ​​the multi-instance industrial terminal. The multi-instance industrial terminal first verifies the operator's identity and permissions, confirming that the operator is authorized to use the target tool.

[0122] The second step involves the multi-instance industrial terminal sending an unlocking command to the tool control unit, specifying the number of the independent lock slot to be unlocked. Upon receiving the command, the controller of the tool control unit energizes the electromagnet of the corresponding electronic lock, causing the bolt to retract and the lock slot door to pop open.

[0123] Third, the operator opens the compartment door and removes the tool. The tool-in-place sensor detects that the tool has been removed, and the status signal changes. The controller of the tool control unit reports the status change event to the multi-instance industrial terminal.

[0124] Fourth, the multi-instance industrial terminal records the first timestamp of tool retrieval, and records the operator's identification, grid number, and tool number (read via a tool identification tag reader). The multi-instance industrial terminal writes this tool retrieval event into the tool retrieval and return record table.

[0125] Fifth, multiple industrial terminals begin listening to the torque data acquisition channel associated with the tool. When torque data arrives, it is bound to the corresponding operator identity based on the currently active job session.

[0126] The specific procedure for returning tools to an operator is as follows.

[0127] The first step is for the operator to return the tool to its original slot. The tool-in-place sensor detects that the tool has been placed, and the status signal changes.

[0128] The second step is for the controller of the tool management unit to report the status change event to the multi-instance industrial terminal.

[0129] The third step involves the multi-instance industrial terminal recording a second timestamp of the tool's return and calculating the tool's usage time (second timestamp minus first timestamp). If the usage time exceeds a preset timeout threshold, the multi-instance industrial terminal sets a timeout flag.

[0130] Fourth, the multi-instance industrial terminal sends a locking command to the tool control unit, the electromagnet of the electronic lock is de-energized, the bolt pops out, and the compartment door panel is locked.

[0131] Fifth, the multi-instance industrial terminal stops listening to the torque data acquisition channel associated with the tool and writes the tool return event to the tool retrieval record table.

[0132] Example 10: Complete Architecture of a Two-Person Collaborative Assembly Operation Data Acquisition System

[0133] like Figure 1 As shown, the dual-person collaborative assembly operation data acquisition system provided in this embodiment of the invention includes one or more dual-person collaborative assembly operation intelligent workstation vehicles, a remote server, and a database server.

[0134] Each intelligent assembly station vehicle, operated by two people, communicates with a remote server via the workshop's local area network (LAN). The workshop LAN uses a gigabit Ethernet or Wi-Fi 6 wireless network to ensure real-time and reliable data transmission.

[0135] The remote server utilizes a high-performance server cluster, deploying data receiving, data processing, and data distribution services. The data receiving service receives uploaded data from each intelligent workstation, performing preliminary verification and format conversion. The data processing service performs in-depth processing on the uploaded data, including data cleaning, anomaly detection, and statistical analysis. The data distribution service distributes the processed data to downstream application systems, such as MES systems, quality traceability systems, and reporting systems.

[0136] The database server uses a relational database management system, such as MySQL or PostgreSQL, to store torque data, tool retrieval records, video clip indexes, and operator identification information. The database server also has a file storage system deployed to store video clip files.

[0137] The database tables are designed as follows:

[0138] The torque data table includes the following fields: torque record identifier, operator identifier, torque value, angle value, tightening time, tool number, workpiece number, bolt position, and associated video clip identifier.

[0139] The tool retrieval and return record table includes the following fields: record identifier, operator identifier, tool number, slot number, retrieval time, return time, and timeout flag.

[0140] The image segment index table contains the following fields: segment identifier, start timestamp, end timestamp, associated torque record identifier set, file path, and operator identifier.

[0141] The operator information table contains the following fields: operator ID, name, employee number, role, permission configuration, and list of authorized positions.

[0142] The traceability report can be viewed on the multi-instance industrial terminals, or accessed through any electronic terminal with display function (including but not limited to tablet computers, mobile phones, and personal computers) to execute the playback method as described in claim 9.

[0143] Example 11

[0144] As a further preferred embodiment of the above embodiments, this embodiment also provides a variant of the intelligent workstation vehicle equipped with a head-mounted imaging device, a UPS emergency power supply and a dedicated charging compartment, which aims to solve the technical problems of limited field of view of fixed cameras when personnel move around and interruption of work recording due to sudden power outages.

[0145] Specifically, the mobile vehicle body is also fixedly equipped with a charging compartment for a head-mounted imaging device. The charging compartment contains multiple charging slots, each equipped with charging contacts and a status indicator light, providing charging and storage functions for the head-mounted smart glasses or head-mounted camera device. The head-mounted imaging device connects to the multi-instance industrial terminal via Wi-Fi or Bluetooth wireless communication links, continuously capturing assembly operation video streams from the operator's first-person perspective when the operator moves or their body obstructs the fixed camera's view. When the head-mounted imaging device is placed in the charging compartment, the multi-instance industrial terminal automatically identifies the device number and records the storage time; when the device is removed, the multi-instance industrial terminal automatically activates the image acquisition channel and binds the acquired video stream to the currently active work session.

[0146] The mobile vehicle is also equipped with a UPS (Uninterruptible Power Supply). The input of the UPS is connected to an external AC power interface, and its output is electrically connected to the control circuits of the multi-instance industrial terminal, the camera, the head-mounted imaging device charging compartment, and the tool control unit. The UPS has a built-in battery pack and a power management controller. The power management controller is configured to monitor the external AC power supply status in real time. When an external AC power interruption is detected or the voltage drops below a preset threshold (e.g., below 187V), it automatically switches to battery power within milliseconds (switching time no more than 10ms) to maintain the continuous operation of the multi-instance industrial terminal and the camera. Based on the battery pack capacity configuration, the continuous power supply time is no less than 15 minutes, preferably 30 minutes, ensuring that the torque data acquisition and image recording of the currently ongoing assembly work are not interrupted due to power failure. When the external AC power is restored, the UPS automatically switches back to AC power and manages the charging of the battery pack.

[0147] Example 12

[0148] As a further improvement to the above embodiment, this embodiment includes an audible and visual alarm device fixedly installed at a prominent position on the top or side of the mobile vehicle body. The audible and visual alarm device is communicatively connected to the multi-instance industrial terminal via an RS485 bus or I / O control line, and includes at least a red alarm light, a yellow warning light, and a buzzer.

[0149] The multi-instance industrial terminal is configured to monitor the following abnormal events and trigger corresponding audible and visual alarm actions:

[0150] (1) When the torque data is detected to exceed the upper limit of the preset range Tmax, it is determined to be a torque over-limit event. The multi-instance industrial terminal sends a first control signal to the sound and light alarm device to control the red alarm light to flash at a frequency of 1Hz, and at the same time the buzzer emits a continuous long alarm sound until the operator confirms the abnormality or completes the over-limit processing operation on the terminal interface.

[0151] (2) When it is detected that the tool has not been returned after being taken out of the independent slot for more than a preset timeout threshold (such as 30 minutes), it is determined as a tool timeout failure event. The multi-instance industrial terminal sends a second control signal to the sound and light alarm device to control the yellow warning light to flash at a frequency of 2Hz, and at the same time the buzzer emits an intermittent short alarm sound (0.5 seconds on, 0.5 seconds off).

[0152] (3) When the network connection interruption is detected to last for more than a preset time (e.g., 60 seconds), it is determined to be a network abnormal event. The multi-instance industrial terminal sends a third control signal to the sound and light alarm device to control the yellow warning light to flash slowly at a frequency of 0.5Hz without triggering the buzzer, so as to avoid noise interference with the operation.

[0153] (4) When the operator attempts to collect tools before the inspection is completed, it is determined to be an inspection abnormality event. The sound and light alarm device emits a yellow light and the buzzer emits two short beeps.

[0154] When the multi-instance industrial terminal triggers any audible or visual alarm, it generates a corresponding abnormal event record and stores it in a local cache database. The abnormal event record includes the event type, trigger time, associated operator ID, associated tool number, and associated work order number. After the network is restored, the abnormal event record is uploaded to a remote server along with other work data for subsequent quality traceability and anomaly statistical analysis.

[0155] Example 13

[0156] As a further refinement of the above embodiments, this embodiment details the specific implementation of the pre-operation inspection process.

[0157] An inspection configuration table is pre-stored in the memory module of the multi-instance industrial terminal. This table contains at least one inspection item required for the current workstation. Inspection items include, but are not limited to: whether all tools in each compartment of the tool control unit are complete; whether the intelligent torque tool has sufficient power; whether the camera image is clear; whether the network connection is normal; and whether the UPS power supply is in normal condition. Each inspection item is configured with a corresponding detection method, including manual verification and automatic detection.

[0158] After any operator completes authentication and successfully creates a job session through the multi-instance industrial terminal, the multi-instance industrial terminal performs the following inspection control steps:

[0159] Step S701: Read the inspection configuration table corresponding to the current workstation, generate an inspection task list and display it in the corresponding display area;

[0160] Step S702: For manually confirmed inspection items, a confirmation button is displayed on the display area, and the operator clicks to confirm each item and records the confirmation timestamp; for automatically detected inspection items, the multi-instance industrial terminal automatically calls the corresponding hardware detection interface (including but not limited to camera status query interface, network Ping test interface, UPS status reading interface) to obtain the detection results.

[0161] Step S703: Determine whether all inspection items have passed. If all have passed, unlock the tool retrieval function, allowing the operator to perform subsequent scanning, tool retrieval, and assembly operations. If any item fails, lock all independent slots of the tool control unit, prohibit tool retrieval operations, and display the failed inspection item and processing prompts on the corresponding display area.

[0162] Step S704: Write the inspection results (including inspection item name, test result, test time, and operator identification) into the inspection record table and store it in the local cache database. When the network is available, upload it to the remote server synchronously.

[0163] Even if the operator triggers a tool retrieval event while the inspection is incomplete, the multi-instance industrial terminal will not send an unlocking command to the tool control unit and will not trigger the inspection abnormality audible and visual alarm described in Embodiment Twelve.

[0164] Example 14

[0165] As a further optimization of the above embodiments, this embodiment provides a detailed supplementary explanation of the image acquisition and uploading mechanism.

[0166] Image acquisition rules:

[0167] The multi-instance industrial terminal executes the following control logic when starting image acquisition:

[0168] (1) After scanning the product QR code or selecting the motor number on the display area, the operator clicks the "Start Operation" button to trigger the video recording start command;

[0169] (2) After video recording is started, the multi-instance industrial terminal is prohibited from switching cameras during the recording process (including switching from a fixed camera to a head-mounted device or vice versa) to ensure that the source of the image for a single video clip is unique;

[0170] (3) The minimum duration of each video recording is 10 seconds. Video segments with a recording duration of less than 10 seconds are considered invalid segments, and no image index record is generated or uploaded, in order to reduce storage and transmission overhead.

[0171] (4) When the duration of a single video recording exceeds 30 minutes, the multi-instance industrial terminal automatically stops the current recording and saves the video segment, and at the same time automatically starts a new video segment to continue recording, ensuring that the size of a single video file is controllable and convenient for storage and transmission;

[0172] (5) After each video segment is recorded, the multi-instance industrial terminal calculates the check value of the video file (using MD5 or SHA-256 algorithm), associates the check value with the segment identifier and stores it for integrity verification during subsequent upload.

[0173] Detailed correlation between video clips and process steps:

[0174] During the construction of the image segment index, each video segment is also associated with the currently executing process step number (processStepId). Specifically, when a multi-instance industrial terminal detects a process switching signal, in addition to performing the segment segmentation operation described in Embodiment 7, it also writes the current process step number into the processStepId field of the video segment metadata, so that each video segment is simultaneously associated with the process step number and the set of torque record identifiers generated within the time window of that process. This supports tracing image records from the "process dimension," allowing quality inspectors to replay all video segments corresponding to a process in the order of process steps.

[0175] Enhanced large file chunk upload and resume functionality:

[0176] Based on the breakpoint resume mechanism, for large video files (e.g., exceeding 10MB), a segmented upload strategy is adopted. The specific steps are as follows:

[0177] Step S801: Before uploading the video file, the multi-instance industrial terminal obtains the file size. If the file size exceeds the preset segmentation threshold (e.g., 5MB), the video file is divided into multiple data segments. Each segment is assigned a segmentation number (incrementing from 0), and the total number of segments is recorded.

[0178] Step S802: Each data segment carries the following parameters when uploaded: video segment identifier (videoId), segment sequence number, total number of segments, segment data MD5 checksum, and business deduplication identifier (generated by concatenating work order number + product number + operator identity identifier + start timestamp).

[0179] Step S803: After receiving the fragment, the remote server checks whether the video segment fragment has been received based on the business deduplication identifier. If all fragments have been received, it returns an "existing" confirmation to the workstation cart, and the workstation cart skips the upload of the video file. If not, it caches the current fragment and records the set of received fragment sequence numbers.

[0180] Step S804: After the remote server receives all the segments of a video clip, it merges the data of each segment and calculates the overall MD5 checksum of the merged file. It then compares the MD5 checksum with the checksum uploaded by the multi-instance industrial terminal in step S801. If they match, it returns a successful upload confirmation message. If they do not match, it returns a failure message and requests that all or missing segments be re-uploaded.

[0181] Example 15

[0182] As a system-level extension of the above embodiments, this embodiment describes the configuration function of the background management system and its collaborative working method with the intelligent workstation vehicle.

[0183] A backend management system is deployed on the remote server, and the backend management system provides the following configuration management functions:

[0184] (1) Employee information management: Configure employee ID, name, login account, role, qualification list (including operable workstation types and usable tool types) and authorized workstation list;

[0185] (2) Tool information management: Configure tool number, tool name, tool type, grid number, calibration cycle and torque calibration parameters (including Pset value, lower limit Tmin, upper limit Tmax);

[0186] (3) Workstation and workstation cart management: Configure workstation number, workstation name, corresponding workstation cart number, and workstation cart authorization code (used for legality verification during the first registration of the terminal).

[0187] (4) Process management: Configure the process route corresponding to the product model. Each process route contains at least one process step. Each process step is configured with process number, process name, list of required tool numbers, torque target value and upper and lower limits, and visual guidance pictures or video links.

[0188] (5) Inspection template management: Configure the list of inspection items and item types (manual confirmation / automatic detection) for each workstation;

[0189] (6) Work order issuance: Issue the work order number, product number, process route number, planned quantity, planned start time and planned end time to the designated workstation.

[0190] The back-end management system communicates with the multi-instance industrial terminals via a web service interface. The interface uses a RESTful API or WebSocket protocol, supporting bidirectional real-time data interaction. When the back-end management system updates the process configuration or issues a new work order, the multi-instance industrial terminals receive the update in real time and cache it locally, eliminating the need for manual synchronization by operators.

[0191] The background management system is also configured to receive operation records, torque data, image indexes, tool retrieval logs, inspection records, and abnormal event records uploaded by the multi-instance industrial terminals, and store the above data in the database server. A unified assembly operation data wide table is established in the database server, using the work order number and product number as primary keys, and associated with operator identification, process step number, torque record identifier set, video clip identifier set, tool retrieval record identifier set, and abnormal event identifier set, realizing one-click data traceability across the entire chain from "personnel-tools-work orders-processes-torque-images-abnormalities".

[0192] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Those skilled in the art can make various modifications and variations to the present invention without departing from its spirit and principles. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for data acquisition in a two-person collaborative assembly operation, characterized in that, The method, applied to a mobile workstation vehicle integrating a tool control unit, a multi-instance industrial terminal, and an image acquisition device, includes: On the same multi-instance industrial terminal, a process-level isolated first job session and a second job session are created, the first job session is bound to a first operator identity, and the second job session is bound to a second operator identity. In response to a tool retrieval event, the operator account identifier of the currently retrieval tool is obtained, and the torque data, tool retrieval and return records, and image fragment index associated with the tool retrieval event are all bound to the first job session or the second job session corresponding to the operator account identifier. The video stream acquired by the image acquisition device is divided into multiple video segments according to the process switching signal or manual segmentation instruction, and a unique segment identifier is assigned to each video segment. Construct a bidirectional index structure between the video segment and the torque records: write the set of torque record identifiers generated within the time period covered by the video segment into the metadata of the video segment, and update the data structure of each torque record to include the unique segment identifier of the video segment.

2. The method according to claim 1, characterized in that, It also includes at least one of the following: Inspection control: After creating the first job session and the second job session, read the inspection configuration table to generate an inspection task list and obtain the confirmation results of each inspection item; If all inspection items pass, the tool retrieval function is unlocked; if any item fails, all independent slots of the tool control unit are locked, the tool retrieval operation is prohibited, and the audible and visual alarm device is triggered to issue an inspection abnormality prompt; the inspection results are written into the inspection record table and stored in the local cache database. Tool Retrieval and Return Management: Each of the multiple independent slots in the tool management unit is equipped with an electronic lock and a tool presence sensor; in response to successful operator authentication, at least one independent slot authorized to the operator is unlocked; the tool presence sensor detects tool retrieval events, records the first retrieval timestamp, and writes the tool number and the retrieval operator's identification into the tool retrieval and return log; the tool return event is detected and the second return timestamp is recorded; if the difference between the return time and the retrieval time exceeds a preset timeout threshold, the audible and visual alarm device is controlled to issue a tool timeout warning, and the timeout flag in the tool retrieval and return log is updated; Index refinement: The currently executing process step number is written into the metadata field of the video segment, so that the video segment is associated with both the process step number and the torque record identifier set; after video recording starts, switching cameras during recording is prohibited; video segments with a single recording duration of less than 10 seconds are considered invalid segments and no image index record is generated; when the single recording duration exceeds 30 minutes, the current recording is automatically stopped and the video segment is saved, and a new video segment is automatically started to continue recording; after each video segment is saved, a file checksum is calculated and associated with the segment identifier for storage.

3. The method according to claim 1, characterized in that, It also includes at least one of the following: Resumable upload: When a network connection interruption with the remote server is detected, the torque data, tool retrieval records, and image segment indexes to be uploaded are stored in the local cache database, and a globally unique sequence number is assigned to each piece of data to be uploaded. The high-order bits of the globally unique sequence number are the device number, and the low-order bits are the locally incrementing sequence number. After the network is restored, the data to be uploaded is uploaded one by one in ascending order of the sequence numbers. A successful upload confirmation message is received from the remote server. The confirmation message contains the sequence number range of the successfully received data. The confirmed uploaded data is deleted from the local cache database according to the sequence number range, and the data for which no confirmation message has been received is added to the retransmission queue. UPS power supply protection: The mobile workstation vehicle is also equipped with a UPS emergency power supply; when an external mains power interruption is detected, the UPS emergency power supply is controlled to switch to battery power supply, and the switching time is no more than 10 milliseconds, so as to maintain the operation of the multi-instance industrial terminal and the image acquisition device. Segmented Upload: When uploading the video file, if the file size exceeds a preset segmentation threshold, segmented upload is performed: the video file is divided into multiple data segments, and each data segment is assigned a segmentation number; each data segment carries a video segment identifier, segmentation number, total number of segments, segmentation check value, and business deduplication identifier during upload. The business deduplication identifier is generated by concatenating the work order number, product number, operator identity identifier, and video start timestamp; after receiving the data, the remote server merges the segments and performs overall verification. If the verification passes, an upload success confirmation message is returned; if the verification fails, the missing segments are required to be re-uploaded.

4. A smart workstation vehicle for two-person collaborative assembly operations, characterized in that, include: Moving vehicle body; The tool control unit is fixedly installed on the mobile vehicle body and includes multiple independent compartments. Each independent compartment is equipped with an electronic lock and a tool presence detector. An industrial computing terminal is fixedly installed on the mobile vehicle body and communicates with the tool control unit. The industrial computing terminal is configured to run a multi-instance operating system to create a first job session and a second job session with process-level isolation. An image acquisition unit is fixedly installed on the mobile vehicle body and is communicatively connected to the industrial computing terminal to acquire image data of the assembly area. An information reading device is communicatively connected to the industrial computing terminal and is used to read product identification and / or operator identification. The industrial computing terminal is configured to perform the method as described in any one of claims 1 to 3.

5. The intelligent workstation cart according to claim 4, characterized in that, It also includes at least one of the following: Terminal and Image Acquisition Unit Structure: The industrial computing terminal includes a first display area and a second display area, respectively displaying the user interfaces of the first and second job sessions; a central processing unit, which creates independent operating environments for the two sessions through containerization technology or cgroups and namespaces process isolation technology; a memory module, divided into two independent cache areas to serve the first and second job sessions respectively; the image acquisition unit includes a fixed image acquisition module and a head-mounted imaging device, the fixed image acquisition module including a pole, a crossarm, a universal joint, and a pan-tilt unit, with a first camera mounted on the pan-tilt unit, and the head-mounted imaging device connected to the industrial computing terminal via a wireless communication link; the front panel of the tool control unit is also equipped with a tool identification tag reader and grid status indicator lights corresponding to each independent grid position; Charging compartment, UPS, and audible and visual alarm: This also includes a charging compartment for the head-mounted imaging device, fixedly installed on the mobile vehicle body. It contains multiple charging slots, each equipped with charging contacts and a status indicator light. When the head-mounted imaging device is placed inside, the industrial computing terminal records the storage time; when it is removed, the corresponding image acquisition channel for the work session is automatically activated. A UPS emergency power supply is fixedly installed inside the mobile vehicle body. Its input is connected to external AC power, and its output is connected to the control circuits of the industrial computing terminal, the image acquisition unit, and the tool control unit, respectively. It is configured to switch to battery power within 10 milliseconds when external AC power is interrupted. An audible and visual alarm device is fixedly installed on the top or side of the mobile vehicle body and communicatively connected to the industrial computing terminal. It is configured to trigger corresponding audible and visual alarm actions in response to torque over-limit events, tool overdue return events, network anomalies, or inspection anomalies.

6. A data acquisition system for two-person collaborative assembly operations, characterized in that, include: A remote server that communicates with the intelligent workstation vehicle; A database server, which is communicatively connected to the remote server; The intelligent workstation vehicle for dual-person collaborative assembly operations as described in claim 4; The remote server is configured to receive and parse the job data uploaded by the first job session and the second job session, and construct a bidirectional index mapping relationship between the video segment identifier and the torque record identifier set; The database server is configured to store torque data, tool retrieval and return records, image segment indexes, inspection records, abnormal event records, and operator identity information.

7. A data processing method based on two-person collaborative assembly operations, characterized in that, Applied to remote servers, including: Receive first operator operation data uploaded from the first operation session of the intelligent workstation vehicle and second operator operation data uploaded from the second operation session. The operation data includes torque data, tool retrieval and return records, and image clip indexes. Parse the image segment index to obtain the video segment identifier and its associated set of torque record identifiers, as well as the video segment identifier that each torque record points to in reverse, and establish a bidirectional index mapping relationship; In response to a quality traceability request, if the traceability request contains a target torque record identifier, the corresponding video segment file is called according to the unique identifier of the video segment in the torque record, and the process jumps to the timestamp corresponding to the torque record to generate a playable traceability report. If the traceability request contains a target video segment identifier, then the associated torque record identifier set in the video segment metadata is parsed, the corresponding torque record details are queried, and a visual traceability report is generated.

8. The data processing method according to claim 7, characterized in that, It also includes the fragmented reception step: The system receives video file segments uploaded by the intelligent workstation vehicle, each segment carrying a video segment identifier, segment number, total number of segments, and segment verification value. Based on the video segment identifier and the service deduplication identifier, determine whether the video segment has been completely received. If it has been completely received, return an existence confirmation message. If the data is not fully received, the current fragment is cached and the set of received fragment sequence numbers is recorded. After all fragments are received, the fragments are merged and the overall checksum is calculated. This checksum is then compared with the checksum provided by the uploader. If they match, a successful upload confirmation message is returned. If they do not match, a failure message is returned and the missing fragments are requested to be re-uploaded.

9. A method for replaying data from a two-person collaborative assembly operation, characterized in that, Applied to electronic terminals, including: In response to user actions on the traceability interface, retrieve the selected target torque record or target video clip; If the selected torque record is the target torque record, read the unique identifier field of the video segment in the torque record, locate the corresponding video segment file, and control the player to start playing the video segment from the position corresponding to the timestamp of the torque record. If the selected video segment is a target video segment, the associated torque record identifier set in the metadata of the video segment is read, the corresponding torque record details are queried and the query results are displayed on the traceability interface, while the video segment is played.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the data acquisition method for two-person collaborative assembly operations as described in any one of claims 1 to 3, or the data processing method based on two-person collaborative assembly operations as described in any one of claims 7 to 8.

Images