Intelligent assembly assistance guidance method and system
By collecting assembly station data in real time and generating dynamic influencing factors based on historical consumption and environmental factors, the problems of lagging material management and scattered data have been solved, enabling accurate prediction and guidance of material demand and improving the continuity and efficiency of assembly operations.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- ZHEJIANG SCI-TECH UNIV
- Filing Date
- 2026-04-20
- Publication Date
- 2026-06-09
Smart Images
Figure CN122175295A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of intelligent manufacturing technology, and in particular to an intelligent assembly-aided guidance method and system. Background Technology
[0002] In industrial assembly production, especially in the assembly line work of complex products such as two-wheeled vehicles and complete vehicles, operators need to complete multiple processes at fixed workstations, frequently handle various materials, and rely on standardized operating procedures to ensure quality. With the advancement of intelligent manufacturing, traditional assembly models are no longer suitable for the demands of efficient and precise production, and existing technologies have some technical problems.
[0003] First, the material management model is passive and outdated. Current material inventory monitoring at assembly stations relies heavily on periodic manual checks, and material replenishment depends on operator judgment or fixed-cycle reporting. There is a lack of forward-looking forecasting based on actual consumption trends, often resulting in assembly line shutdowns due to material shortages or excessive replenishment leading to warehouse space occupation and material waste. Second, data collection during operations is fragmented and isolated. Key data such as material inventory, operating cycle time, and work-in-process queues at assembly stations are mostly obtained through independent equipment or manual recording, lacking a unified data collection and integration platform. Finally, dynamic production factors are not fully considered. Existing guidance methods do not adequately account for the impact of declining operator efficiency, changes in environmental temperature and humidity on material handling, or the effects of operating cycle time deviations and work-in-process backlogs on material consumption. This leads to significant discrepancies between predicted and actual material demand, and a disconnect between guidance instructions and actual production, making it difficult to ensure the continuity and stability of assembly operations.
[0004] Therefore, there is an urgent need for an intelligent assembly-aided guidance method and system to solve the above problems. Summary of the Invention
[0005] The purpose of this invention is to provide an intelligent assembly-assisted guidance method, comprising the following steps:
[0006] Real-time acquisition of assembly station operation process data, including material inventory data, operation cycle data, work-in-process queue data, and current operation process;
[0007] Based on the material inventory data and combined with the corresponding historical material consumption data retrieved from the operation process data, the baseline material consumption trend of the assembly station is predicted.
[0008] By analyzing the deviations between the operation cycle data and the standard cycle, as well as the deviations between the work-in-process queue data and the standard queue, dynamic influencing factors characterizing the current production tension are generated.
[0009] The dynamic influencing factors are used to correct the baseline material consumption trend in real time to obtain corrected material demand forecast information, and material replenishment guidance information is generated based on the material demand forecast information.
[0010] By combining the material replenishment guidance information with the current work process, real-time assembly guidance instructions for the operator are generated and output.
[0011] Furthermore, the step of acquiring real-time operation process data of the assembly station includes:
[0012] Periodically read sensor data associated with the material storage container, the sensor data reflecting the real-time physical quantity of the material, and convert the physical quantity into a specific value representing the current inventory as the material inventory data, and obtain the identification information of the material.
[0013] Acquire video surveillance data of the assembly area, identify the completion cycle of preset standard operation actions, and calculate the operation cycle data;
[0014] By querying real-time data from the production management system, the quantity of products awaiting processing downstream of the current workstation is obtained and used as the work-in-process queue data.
[0015] Furthermore, the step of predicting the baseline material consumption trend of the assembly station includes:
[0016] Retrieve the material consumption records of the assembly station under comparable historical production conditions to form a historical consumption rate time series;
[0017] Obtain the operator proficiency decay factor, which characterizes the current work efficiency status of the operator;
[0018] Obtain the environmental temperature and humidity influence coefficients that characterize the physical environment's impact on the material handling process;
[0019] The historical consumption rate time series is learned and fitted to obtain an initial baseline consumption rate. The initial baseline consumption rate, the operator proficiency decay factor, and the environmental temperature and humidity influence coefficient are fused and calculated to generate the final baseline material consumption rate, which serves as the baseline material consumption trend for the assembly station.
[0020] Furthermore, the step of generating dynamic influencing factors characterizing the current level of production stress includes:
[0021] The beat deviation coefficient is calculated by using the operating beat data and the standard beat, and the beat deviation coefficient is used to quantify the change in actual operating speed.
[0022] A queue pressure coefficient is calculated using work-in-process queue data and a standard queue length. This queue pressure coefficient is used to quantify the pressure caused by work-in-process backlog.
[0023] The rhythm deviation coefficient and the queue pressure coefficient are combined to generate a unified dynamic influence factor.
[0024] Furthermore, the step of generating material replenishment guidance information based on the corrected material demand forecast information includes:
[0025] Based on the baseline material consumption rate and the dynamic influence factor, the real-time predicted consumption rate is obtained, and the real-time predicted consumption rate is used as the corrected material demand forecast information.
[0026] Based on the current material inventory data and the real-time predicted consumption rate, calculate the expected material depletion time.
[0027] Based on the expected depletion time of the materials and the preset logistics response time, determine whether a pre-supply instruction needs to be issued;
[0028] When it is determined that a pre-replenishment instruction needs to be issued, the material replenishment guidance information is generated. The information includes at least material identification information, suggested replenishment quantity, and a time prompt for initiating the replenishment operation.
[0029] Furthermore, the step of generating and outputting real-time assembly guidance instructions for the operator includes:
[0030] The material replenishment guidance information is integrated with the currently executing assembly process information;
[0031] If the material replenishment guidance information indicates that the material is sufficient and the replenishment plan is normal, then a regular guidance instruction containing the current standard operating procedure is generated and output.
[0032] If the material replenishment guidance information indicates a risk of replenishment delay or shortage, a special guidance instruction containing a warning message, a suggestion to wait, or an alternative process will be generated and output.
[0033] The generated real-time assembly guidance instructions are visualized through the display device at the assembly station and supplemented with voice or signal lights.
[0034] Furthermore, this application also discloses an intelligent assembly-assisted guidance system, comprising:
[0035] The acquisition module is used to acquire the operation process data of the assembly station in real time. The operation process data includes material inventory data, operation cycle data, work-in-process queue data, and the current operation process.
[0036] The prediction module is used to predict the baseline material consumption trend of the assembly station based on the material inventory data and the corresponding historical material consumption data retrieved from the operation process data.
[0037] The generation module is used to analyze the deviation between the operation cycle data and the standard cycle and the deviation between the work-in-process queue data and the standard queue, and generate dynamic influencing factors that characterize the current production tension.
[0038] The correction module is used to correct the baseline material consumption trend in real time using the dynamic influencing factors, obtain the corrected material demand forecast information, and generate material replenishment guidance information based on the material demand forecast information.
[0039] The output module is used to integrate the material replenishment guidance information with the current work process to generate and output real-time assembly guidance instructions for the operator.
[0040] Furthermore, the prediction module includes:
[0041] The allocation unit is used to retrieve the material consumption records of the assembly station under historical comparable production conditions to form a historical consumption rate time series.
[0042] The first acquisition unit is used to acquire the operator proficiency decay factor, which represents the current work efficiency status of the operator.
[0043] The second acquisition unit is used to acquire the environmental temperature and humidity influence coefficient, which characterizes the influence of the physical environment on the material handling process.
[0044] The generation unit is used to learn and fit the historical consumption rate time series to obtain an initial baseline consumption rate. The initial baseline consumption rate, the operator proficiency decay factor, and the environmental temperature and humidity influence coefficient are fused and calculated to generate the final baseline material consumption rate, which serves as the baseline material consumption trend for the assembly station.
[0045] This application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the above-described intelligent assembly-assisted guidance method.
[0046] This application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-described intelligent assembly-assisted guidance method.
[0047] The beneficial effects of this application are as follows:
[0048] Firstly, this invention collects material inventory data in real time through multiple channels, combines historical consumption patterns, operator status, and environmental factors to predict baseline consumption trends, and then dynamically adjusts demand based on production urgency to achieve accurate prediction of material needs. Based on the prediction results, it generates guidance information including replenishment timing and quantity, triggering an automated replenishment process, significantly reducing manual intervention, avoiding material shortages and shutdowns or excessive stockpiling, and improving the timeliness and rationality of material supply.
[0049] Secondly, this invention dynamically generates guidance instructions based on material replenishment status and the current process. When materials are sufficient, it outputs standard operating procedures; when there is a replenishment risk, it provides early warnings and alternative process suggestions, ensuring that guidance instructions are precisely matched with material supply and production rhythm. Simultaneously, through multiple channels such as visualization, voice, and signal lights, it ensures that operators receive and execute instructions promptly, reducing misoperations and wasted time, and improving operational standardization and continuity.
[0050] Thirdly, this invention fully considers static influencing factors such as operator skill decline and changes in environmental temperature and humidity, as well as dynamic production states such as operational cycle deviations and work-in-process inventory. By incorporating quantitative factors into the prediction and decision-making process, it makes material demand forecasts more closely aligned with actual consumption patterns and guides instructions to be more adaptable to real-time production scenarios. This effectively reduces the impact of dynamic production changes on assembly efficiency, improves the stability of assembly quality, and enhances overall production efficiency. Attached Figure Description
[0051] Figure 1 This is a schematic diagram of a method flow proposed in an embodiment of this application.
[0052] Figure 2 This is a schematic diagram of the system structure proposed in an embodiment of the present invention.
[0053] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation
[0054] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
[0055] like Figure 1 As shown, this application provides an intelligent assembly-assisted guidance method, including the following steps:
[0056] S1, real-time acquisition of the operation process data of the assembly station, the operation process data including material inventory data, operation cycle data, work-in-process queue data and current operation process;
[0057] S2, Based on the material inventory data and combined with the corresponding historical material consumption data retrieved from the operation process data, predict the baseline material consumption trend of the assembly station.
[0058] S3, Analyze the deviation between the operation cycle data and the standard cycle and the deviation between the work-in-process queue data and the standard queue, and generate a dynamic influencing factor characterizing the current production tension.
[0059] S4. The dynamic influencing factor is used to correct the baseline material consumption trend in real time to obtain the corrected material demand forecast information, and material replenishment guidance information is generated based on the material demand forecast information.
[0060] S5. Based on the material replenishment guidance information and the current work process, generate and output real-time assembly guidance instructions for the operator.
[0061] As described in steps S1-S5 above, in industrial assembly scenarios, material consumption is directly related to production rhythm. Operators need to match production plans while ensuring work quality. Material shortages or untimely replenishment can lead to assembly line interruptions. At the same time, fluctuations in operating rhythm and work-in-process inventory can dynamically change the material consumption rate. If demand is predicted based solely on fixed patterns, it cannot adapt to real-time production changes, thereby affecting overall production efficiency. Therefore, it is necessary to establish a demand prediction and guidance mechanism that takes into account both historical patterns and dynamic working conditions.
[0062] However, existing technologies mostly rely on fixed-cycle replenishment or simply rely on historical consumption data to predict material demand, without considering real-time production factors such as operational cycle deviations and changes in work-in-process queues. This results in low demand forecast accuracy, a disconnect between replenishment instructions and actual working conditions, and assembly guidance that is mostly based on preset processes without integrating material replenishment status. When material shortages or delays occur, the guidance strategy cannot be adjusted in a timely manner, which can easily lead to operator errors or excessively long waiting times.
[0063] This invention collects multi-dimensional operational data from assembly stations in real time, combines historical consumption data with dynamic production status to build a predictive model, and achieves accurate prediction of material demand and real-time assembly guidance. This solves the problems of delayed material replenishment and lack of dynamic adaptability of guidance instructions in traditional assembly processes, thereby improving the efficiency and continuity of assembly operations.
[0064] The core principle of this application lies in comprehensively capturing multi-dimensional real-time operational data from assembly stations, combining historical patterns and dynamic production status to accurately predict material demand, and then generating real-time guidance that adapts to material supply and process advancement, solving the problems of delayed material replenishment and lack of flexibility in guidance instructions in traditional assembly. The solution first collects real-time data on material inventory, operating cycle time, work-in-process queue, and current work processes to provide comprehensive and real-time basic data support for subsequent decision-making, ensuring that all analyses are based on actual working conditions. Then, by retrieving material consumption records under comparable historical production conditions and combining them with current material inventory data, it predicts benchmark material consumption trends, establishing a demand forecasting basis based on historical patterns. Subsequently, by analyzing the deviations between operating cycle time and standard cycle time, and between work-in-process queue and standard queue, it generates dynamic influencing factors characterizing the current production tension, quantifying the impact of real-time production rhythm changes on material consumption. Finally, it uses these dynamic influencing factors to analyze the benchmark material consumption trend. Real-time corrections are made to ensure that material demand forecasts accurately match the current production pressure, preventing a disconnect between baseline trends and actual demand. Based on the corrected forecasts, material replenishment guidance information, including material identity, replenishment quantity, and replenishment timing, is generated to achieve proactive material replenishment. Finally, by integrating the material replenishment guidance information with the current work process, regular or special assembly guidance instructions are generated and output based on material sufficiency or the existence of replenishment risks. This ensures that operators' work actions are dynamically matched with the material supply status, guaranteeing the continuity and standardization of assembly operations, maximizing the use of working hours, reducing downtime for waiting for materials, and ultimately improving the overall operational efficiency and quality stability of the assembly station.
[0065] In one embodiment, the step of acquiring real-time work process data at the assembly station includes:
[0066] S11, periodically read the sensor data associated with the material storage container through the data interface. The sensor data reflects the real-time physical quantity of the material, and converts the physical quantity into a specific value representing the current inventory as the material inventory data, and obtains the identification information of the material.
[0067] S12, acquire video monitoring data of the assembly area, identify the completion cycle of the preset standard operation action, and calculate the operation cycle data. Alternatively, the operation cycle data can be obtained by connecting to the control system of the assembly tool and its working cycle signal.
[0068] S13. By querying real-time data from the production management system, the quantity of products awaiting processing downstream of the current workstation is obtained as the work-in-process queue data. Alternatively, the work-in-process queue data can be obtained directly by parsing the sensor network signals deployed between workstations.
[0069] As described in steps S11-S13 above, real-time and accurate acquisition of assembly station operation process data is achieved through multi-channel automated data collection and integration of multiple data types. This provides reliable data support for subsequent prediction of baseline material consumption trends, generation of dynamic influencing factors, and output of real-time assembly guidance instructions. The operation process data covers material inventory, operation cycle time, work-in-process queue, and current operation procedures. This data forms the core foundation connecting the entire intelligent assembly assisted guidance process. The real-time nature of the data directly determines the timeliness of material demand forecasting, the accuracy of the data affects the rationality of dynamic influencing factors, and the comprehensiveness of the data ensures the relevance of assembly guidance instructions. If data collection is delayed, inaccurate, or incomplete, all subsequent decision-making processes will deviate from the actual operating conditions, making effective assembly assisted guidance impossible.
[0070] Existing technologies have several problems in data collection during operations. Material inventory data largely relies on periodic manual checks, which is time-consuming, labor-intensive, and prone to errors due to human negligence. Furthermore, the intervals between checks cause data lag, failing to reflect real-time consumption status. Operational cycle time data is often recorded manually, which is highly subjective and difficult to continuously track changes in the cycle time for each work cycle. Work-in-process queue data largely depends on manual entry and updates from the production management system, resulting in untimely information synchronization and an inability to accurately reflect the actual backlog downstream. To address these issues, this step employs a multi-dimensional data collection solution involving sensor acquisition, system linkage, and automated identification. Through standardized data acquisition processes, it ensures the real-time nature, accuracy, and comprehensiveness of various operational process data.
[0071] Step S11 is used to acquire material inventory data and material identification information. It periodically reads sensor data associated with the material storage container through a preset data interface. The sensor data acquisition method can be selected according to the material type. For small parts, a weight sensor can be used to collect weight data; for large components, an infrared sensor or photoelectric sensor can be used to collect quantity data. The acquisition cycle is calibrated offline to 5 seconds per acquisition to ensure real-time data transmission. The data interface uses an industrial Ethernet interface to establish stable communication with the sensing device. The acquired physical quantities are converted into specific values representing the current inventory through a preset conversion formula. For example, if the weight of a single bolt of a certain type is 0.5 grams, and the weight sensor collects a total material weight of 500 grams, the current inventory is calculated to be 1000 bolts. Simultaneously, the RFID module integrated into the sensing device reads the electronic tag on the material packaging to obtain the material identification information, which includes key attributes such as material model, specifications, and batch number, ensuring that subsequent material replenishment accurately matches current operational needs.
[0072] Step S12 is used to acquire operation cycle data, providing two complementary acquisition methods to adapt to different assembly scenarios. The first method acquires video stream data through high-definition video surveillance equipment deployed in the assembly area. The video surveillance equipment's frame rate is set to 30 frames per second, and a target detection algorithm is used to identify the start and end frames of preset standard operation actions. For example, in the assembly of a two-wheeled vehicle, the action of tightening wheel bolts, from the moment the wrench contacts the bolt until the torque is reached, calculates the time difference between two frames as the completion cycle of a single operation action. The average completion cycle of 10 consecutive operation actions is then used as the current operation cycle data. The second method connects to the control system of the assembly tools, such as the control system of an electric torque wrench or pneumatic press machine, via an RS485 interface. It receives the tool's work cycle signal in real time, which includes timestamps for action start, operation, and stop. The operation cycle data is obtained by calculating the time interval between two adjacent start signals. Both acquisition methods can be flexibly selected according to the tool configuration and environmental conditions of the assembly station, ensuring the continuity and accuracy of operation cycle data acquisition.
[0073] Step S13 is used to obtain work-in-process (WIP) queue data, and two collection methods are provided. The first method queries real-time data from the production management system via an HTTP interface. The production management system records the product flow status of each workstation in real time. By filtering the pending product records of downstream workstations of the current workstation, the WIP queue data is obtained. For example, if the current assembly workstation is a two-wheeled vehicle frame assembly workstation, and its downstream is the wheel assembly workstation, querying the production management system shows that the number of pending frames at the wheel assembly workstation is 8. This number is the WIP queue data for the current workstation. The second method analyzes the sensor network signals deployed on the conveyor lines between workstations. The sensor network consists of infrared sensors, deployed at 0.5-meter intervals. Sensor trigger signals indicate that a product has passed. By counting the number of sensor triggers that do not enter the downstream workstation but are located between the current workstation and the downstream workstation, the WIP queue data is directly obtained. Both methods rely on system data and on-site sensing, respectively, to ensure that the WIP queue data can reflect the actual backlog in real time, providing an accurate basis for the subsequent generation of dynamic influencing factors.
[0074] In one embodiment, the step of predicting the baseline material consumption trend of the assembly station includes:
[0075] S21, retrieve the material consumption records of the assembly station under historical comparable production conditions to form a historical consumption rate time series;
[0076] S22, obtain the operator proficiency decay factor that characterizes the current operator's work efficiency status. The operator proficiency decay factor is obtained by querying the personnel work hour management system to obtain the continuous working time of the current operator at the assembly station, and by combining the statistical model of the operator's historical average work efficiency changing with continuous working time.
[0077] S23, obtain the environmental temperature and humidity influence coefficient characterizing the influence of the physical environment on the material handling or processing process, wherein the environmental temperature and humidity influence coefficient is obtained by reading the real-time data collected by the temperature and humidity sensor deployed at the assembly station and comparing and matching it with the ideal storage and working environment parameters corresponding to the material in the preset material characteristic database.
[0078] S24, a time series analysis model is used to learn and fit the historical consumption rate time series to obtain an initial baseline consumption rate. The initial baseline consumption rate, the operator proficiency decay factor, and the environmental temperature and humidity influence coefficient are fused and calculated to generate the final baseline material consumption rate, which serves as the baseline material consumption trend for the assembly station.
[0079] As described in steps S21-S24 above, by integrating historical consumption patterns, operator status, and environmental influencing factors, and using time series analysis and multi-parameter fusion calculation, an accurate baseline material consumption trend is generated.
[0080] Material consumption at assembly stations does not follow a fixed pattern. Operator efficiency decreases with continuous working hours, and changes in the physical environment, such as temperature and humidity, affect the ease of material retrieval and processing efficiency. These factors directly alter the material consumption rate. Predicting consumption trends based solely on historical data fails to account for the dynamic variables in actual production, leading to significant discrepancies between predictions and actual consumption. This, in turn, impacts the timeliness and accuracy of subsequent material replenishment guidance. Therefore, it is necessary to construct a baseline consumption trend prediction mechanism that considers multiple influencing factors.
[0081] Existing technologies for predicting material consumption trends often rely solely on historical average consumption data, neglecting the impact of individual operator conditions and environmental changes on consumption rates. This results in insufficient adaptability of the prediction models. For example, a decrease in efficiency after four hours of continuous work by the same operator, or environmental humidity exceeding the ideal processing range for materials, will cause the actual consumption rate to deviate from the historical average. Traditional methods cannot capture these dynamic changes. To address this issue, this paper constructs a baseline sequence by retrieving comparable historical data, introduces an operator proficiency decay factor and environmental temperature and humidity influence coefficients, and combines this with a time series analysis model to achieve multi-dimensional parameter fusion. This comprehensively covers key variables affecting material consumption, improving the prediction accuracy of baseline consumption trends.
[0082] Step S21 is used to construct a historical consumption rate time series. Historical material consumption records for the current assembly station are retrieved from the production management system's database. Comparable data with the same product model, production batch, and work process as the current production conditions are selected to ensure that the historical data is relevant to the current operating conditions. The selected historical data is arranged in chronological order, and the material consumption rate is calculated hourly to form a historical consumption rate time series. For example, in five consecutive production batches, the hourly consumption rates of a certain material at this station are 10 units, 11 units, 9 units, 10 units, and 11 units, respectively, forming the corresponding time series data.
[0083] Step S22 is used to obtain the operator's proficiency decay factor. The continuous working time of the current operator is queried through the personnel and work hour management system, and historical work data of the operator is retrieved to establish a linear statistical model of the historical average work efficiency changing with continuous working time. The parameters of this model are calibrated through offline experiments. For example, it is set that there is no efficiency decay within 1 hour of continuous work, and efficiency decays by 2% for every additional hour of continuous work. When working continuously for 4 hours, efficiency decays by 8%, and the corresponding operator proficiency decay factor is 0.92. This model dynamically calculates the decay factor under the current continuous working time, accurately reflecting the impact of the operator's current work efficiency status on the material consumption rate.
[0084] Step S23 is used to obtain the environmental temperature and humidity influence coefficient. Temperature and humidity data are collected in real time by temperature and humidity sensors deployed at the assembly station. The sensor collection cycle is 5 minutes / time to ensure data real-time performance. The collected real-time temperature and humidity data is compared with a pre-set material characteristic database. The material characteristic database stores the ideal storage and operating environment parameters for various materials. For example, the ideal operating temperature for a certain type of plastic part is 20℃-25℃, and the ideal operating humidity is 40%-60%. If the real-time temperature is 28℃ and the humidity is 65%, which exceeds the ideal range, the environmental temperature and humidity influence coefficient is calculated to be 0.95 through a preset mapping rule. This environmental temperature and humidity influence coefficient quantifies the inhibitory effect of environmental factors on the efficiency of material handling and processing.
[0085] Step S24 generates the final baseline material consumption rate. An ARIMA time series analysis model is used to learn and fit the historical consumption rate time series. The model order is determined to be ARIMA(1,1,1) using the AIC criterion. The initial baseline consumption rate is obtained through model training. The initial baseline consumption rate, operator skill attenuation factor, and environmental temperature and humidity influence coefficients are linearly fused and calculated using the following formula:
[0086] ;
[0087] Among them, the Indicating the baseline material consumption rate, the Indicates the initial baseline consumption rate, the The operator's proficiency decay factor is represented by the following. The environmental temperature and humidity influence coefficient is used to represent the influence coefficient. For example, if the initial baseline consumption rate is 10 pieces / hour, the operator's skill attenuation factor is 0.92, and the environmental temperature and humidity influence coefficient is 0.95, the calculated final baseline material consumption rate is 8.74 pieces / hour. This baseline material consumption rate is the baseline material consumption trend of the assembly station. It comprehensively considers historical patterns, operator status, and environmental factors, ensuring the accuracy and adaptability of the prediction results.
[0088] In one embodiment, the step of generating dynamic influencing factors characterizing the current level of production stress includes:
[0089] S31: Calculate the beat deviation coefficient, which is obtained by dividing the operation beat data by the standard beat, and is used to quantify the change in actual operation speed.
[0090] S32: Calculate the queue pressure coefficient, which is obtained by dividing the work-in-process queue data by the standard queue length, and is used to quantify the pressure caused by work-in-process backlog.
[0091] S33: The beat deviation coefficient and the queue pressure coefficient are fused to generate a unified dynamic influence factor. The fusion method is to multiply the two coefficients or to perform a weighted calculation on the two coefficients according to a preset rule.
[0092] As described in steps S31-S33 above, by quantifying the operational cycle deviation and the work-in-process queue deviation, a dynamic influencing factor is generated using a coefficient fusion method. This accurately represents the current production tension level and provides a scientific quantitative basis for real-time correction of the baseline material consumption trend. The production tension level is directly related to the actual material consumption rate. A faster operational cycle means increased material consumption per unit time. Accelerated work-in-process queues will prompt operators to increase their work speed, thereby increasing material demand. Only by accurately capturing this dynamic change can material demand forecasts be aligned with real-time production conditions, avoiding a disconnect between the baseline consumption trend and actual demand.
[0093] Current technologies fail to systematically quantify production stress levels, either ignoring their impact on material consumption or considering only one factor—cycle time or queue—leading to an inability to comprehensively reflect production dynamics. For example, focusing solely on faster cycle time while neglecting work-in-process inventory will underestimate material consumption rates; focusing solely on queue backlogs while ignoring slower cycle time will overestimate material demand. To address this issue, this paper calculates the deviation coefficients of two key dimensions separately, then generates a unified factor through standardized fusion, achieving a comprehensive and accurate quantification of production stress levels.
[0094] Step S31 is used to calculate the cycle time deviation coefficient. The standard cycle time is retrieved from the production process document, which specifies the preset standard operating cycle for each assembly process. For example, the standard cycle time for a two-wheeled vehicle bolt assembly process is 10 seconds / piece. The operation cycle time data is obtained through video monitoring recognition or the assembly tool control system as described in step S12. If the actual collected operation cycle time data is 8 seconds / piece, the operation cycle time data is divided by the standard cycle time to obtain a cycle time deviation coefficient of 0.8. This cycle time deviation coefficient quantifies the degree of deviation between the actual operating speed and the standard speed. A coefficient less than 1 indicates that the actual operating speed is faster than the standard and the production intensity is high. A coefficient greater than 1 indicates that the operating speed is slower than the standard and the intensity is low. Its value change directly reflects the impact of the operation rhythm on material consumption.
[0095] Step S32 is used to calculate the queue pressure coefficient. The standard queue length is retrieved from the process parameters of the production management system. This parameter is preset based on the production line capacity and turnover efficiency. For example, the standard queue length for a certain assembly station is 5 pieces. The work-in-process queue data is obtained through querying the production management system as described in S13 or through sensor network parsing. If the actual work-in-process queue data is 8 pieces, the work-in-process queue data is divided by the standard queue length to obtain a queue pressure coefficient of 1.6. This queue pressure coefficient quantifies the production pressure caused by the backlog of work-in-process. A coefficient greater than 1 indicates the existence of backlog. The larger the value, the more serious the backlog and the higher the production tension. It directly reflects the forcing effect of downstream demand on the material consumption of the current workstation.
[0096] Step S33 is used to fuse the two coefficients to generate a dynamic impact factor, providing two standardized fusion methods. The first is a weighted calculation, with a preset weight of 0.4 for the cycle time deviation coefficient and 0.6 for the queue pressure coefficient. These weights are calibrated through offline experiments to ensure a reasonable weighting of the two dimensions' influence on production intensity. For example, with a cycle time deviation coefficient of 0.8 and a queue pressure coefficient of 1.6, the weighted calculation yields a dynamic impact factor of 0.8 × 0.4 + 1.6 × 0.6 = 1.28. The second method is direct multiplication. Again, using the above coefficients as an example, multiplying yields a dynamic impact factor of 0.8 × 1.6 = 1.28. The two fusion methods can be flexibly selected according to different assembly scenarios. The final generated dynamic impact factor quantifies the deviation of the two independent dimensions into a unified indicator, accurately representing the current production intensity and ensuring that the corrected material demand forecast can adapt to changes in production rhythm in real time.
[0097] In one embodiment, the step of generating material replenishment guidance information based on the revised material demand forecast information includes:
[0098] S41: Multiply the baseline material consumption rate by the dynamic influence factor to obtain the real-time predicted consumption rate, and use the real-time predicted consumption rate as the corrected material demand forecast information;
[0099] S42: Calculate the expected material depletion time based on the current material inventory data and the real-time predicted consumption rate;
[0100] S43: Based on the expected depletion time of the material and the preset logistics response time, determine whether a pre-supply instruction needs to be issued;
[0101] S44: When it is determined that a pre-supply instruction needs to be issued, the material supply guidance information is generated. The information includes at least the material identification information obtained in step S11, the suggested supply amount, and the suggested time prompt for initiating the supply operation.
[0102] As described in steps S41-S44 above, by dynamically correcting the material consumption rate, accurately calculating the material depletion time, and combining the logistics response time to determine the replenishment timing, material replenishment guidance information containing core replenishment information is generated, thereby achieving the foresight and accuracy of material replenishment and ensuring continuous material supply to the assembly station without excessive inventory backlog.
[0103] While the baseline material consumption trend takes into account historical patterns, operator status, and environmental factors, it does not incorporate changes in consumption rate caused by real-time production pressure. Increased production pressure directly accelerates material consumption. If replenishment is still planned according to the baseline trend, it will lead to delayed replenishment or insufficient replenishment. At the same time, replenishment operations require a certain logistics response time. If replenishment is not predicted and initiated in advance, there will be a situation where production stops after materials are exhausted and waiting for replenishment. Therefore, it is necessary to determine the replenishment strategy and generate clear guidance information based on the real-time corrected consumption trend and logistics timeliness.
[0104] Existing technologies for guiding material replenishment sometimes only trigger replenishment when material inventory falls below a fixed threshold, failing to consider real-time consumption rate changes. If the consumption rate suddenly increases, materials may be depleted before the threshold is reached. Other methods, while incorporating demand forecasting, do not consider logistics response time, leading to late replenishment initiation and untimely delivery. Furthermore, traditional replenishment information often only includes material type and quantity, lacking replenishment timing prompts, which can result in premature replenishment occupying storage space or late replenishment impacting production. To address these issues, this paper uses dynamic influencing factors to correct the baseline consumption trend and obtain a real-time predicted consumption rate. This is then combined with inventory and logistics time to determine replenishment timing, ultimately generating complete guidance information including identification, suggested replenishment quantity, and time prompts. This achieves dynamic adaptation of replenishment decisions and comprehensive guidance information.
[0105] Step S41 is used to obtain the corrected material demand forecast information. The baseline material consumption rate is calculated by step S24, for example, the previously calculated 8.74 units / hour. The dynamic influence factor is generated by step S33, for example, 1.28. Multiplying the baseline material consumption rate by the dynamic influence factor directly yields the real-time predicted consumption rate of 8.74 × 1.28 ≈ 11.19 units / hour. This rate accurately reflects the actual material consumption under the current production pressure and serves as the corrected material demand forecast information, providing an accurate basis for subsequent exhaustion time calculations.
[0106] Step S42 is used to calculate the expected material depletion time. The current material inventory data is obtained from step S11, for example, 1000 units, and the real-time predicted consumption rate is 11.19 units / hour. By dividing the current material inventory data by the real-time predicted consumption rate, the expected material depletion time is obtained as 1000 ÷ 11.19 ≈ 90 hours. That is, from the current moment, the material will be depleted in 90 hours. This time point provides a core reference for judging the timing of subsequent replenishment.
[0107] Step S43 determines whether a pre-replenishment instruction needs to be issued. A preset logistics response time is retrieved from the logistics management system; this time is the average time from when a material replenishment request is initiated until the material arrives at the assembly station, for example, 24 hours. The estimated material depletion time is compared with the preset logistics response time. If the estimated material depletion time is less than or equal to the preset logistics response time, it indicates that initiating replenishment now can ensure the material arrives before depletion. If the estimated material depletion time is greater than the preset logistics response time, replenishment is not initiated, and monitoring continues. For example, if the current estimated depletion time is 90 hours, which is greater than 24 hours, a pre-replenishment instruction is not issued. The estimated depletion time is recalculated hourly until it decreases to 24 hours or less.
[0108] Step S44 generates material replenishment guidance information. When it is determined that a pre-replenishment instruction needs to be issued, key replenishment information is integrated. Material identification information is obtained from step S11 and includes material model, specifications, batch number, etc., for example, "Model A-123 bolt, Specification M8×20, Batch 20240501". The recommended replenishment quantity is set based on the current material inventory data and subsequent production plan to meet the needs of the next production cycle. For example, if the current inventory is 1000 units and the subsequent production plan requires 1000 units, the recommended replenishment quantity is set to 1000 units. The suggested time for initiating the replenishment operation is determined based on the expected material depletion time and the preset logistics response time. For example, if the expected depletion time is 24 hours, the suggested time for initiating the replenishment operation is the current moment, ensuring material delivery within the logistics response time. The generated material replenishment guidance information integrates the above three core contents, providing clear replenishment execution guidelines for the logistics department and on-site management personnel, ensuring accurate and timely execution of replenishment operations.
[0109] In one embodiment, the step of generating and outputting operator-oriented real-time assembly guidance instructions includes:
[0110] S51: Integrate the material replenishment guidance information with the currently executing assembly process information;
[0111] S52: If the material replenishment guidance information indicates that the material is sufficient and the replenishment plan is normal, then generate and output a regular guidance instruction containing the current standard operating procedure;
[0112] S53: If the material replenishment guidance information indicates a risk of replenishment delay or shortage, a special guidance instruction containing a warning message, a suggestion to wait, or to execute an alternative process is generated and output.
[0113] S54: The generated real-time assembly guidance instructions are visualized through the display device of the assembly station, and auxiliary prompts are provided through voice or signal lights.
[0114] As described in steps S51-S54 above, by integrating material replenishment guidance information with current work process information, real-time assembly guidance instructions adapted to the material status are generated according to different scenarios. A multi-channel collaborative output method is adopted to ensure that operators receive the guidance content in a timely and accurate manner, achieving dynamic adaptation between assembly operations and material supply, and ensuring operational continuity and efficiency. The smooth progress of assembly operations depends on the precise matching of process execution and material supply. When materials are sufficient, efficient operation must be carried out according to standard procedures. When there is a shortage of materials or a delay in replenishment, the operation strategy must be adjusted in a timely manner. If the guidance instructions are not combined with the material status and are only mechanically output according to preset procedures, it will cause operators to blindly proceed even when materials are insufficient, leading to work stoppages, waiting, or misoperation, affecting production rhythm and product quality. Therefore, it is necessary to dynamically generate targeted guidance instructions based on the real-time material replenishment status.
[0115] Traditional assembly guidance instructions are often fixed in production process documents or simply displayed on equipment screens, containing only fixed work steps and not linked to real-time material replenishment status. When materials are scarce or replenishment is delayed, they cannot provide effective adjustment suggestions, causing operators to continue working until materials are exhausted, resulting in wasted time. Furthermore, the guidance instructions are output in a single way, relying solely on visual presentation, making them easily overlooked in noisy workshop environments or during busy operations, resulting in poor guidance effectiveness. To address these issues, this paper integrates material replenishment information and process information, generating two types of guidance instructions—regular and special—based on different scenarios. These instructions are then output through multiple channels, including visualization, voice, and signal lights, achieving dynamic adaptability and effective reception.
[0116] Step S51 integrates core information. Material replenishment guidance information is generated by step S44, including material identification information, suggested replenishment quantity, suggested replenishment operation time prompt, and material sufficiency status or shortage risk warning. Current work process information is retrieved from the work order data of the production management system, clarifying the assembly steps to be executed, required material specifications, process parameters, and sequence. A data association algorithm integrates these two types of information to establish a correspondence between material status and process execution. For example, if the current process is "installing two-wheeled vehicle frame fixing bolts," the integration process clarifies whether the replenishment status of the required A-123 bolts supports the continuation of the process, laying the foundation for subsequent scenario-based instruction generation.
[0117] Step S52 is used to generate routine guidance instructions. When the material replenishment guidance information indicates sufficient materials and the replenishment plan is normal (meaning the current material inventory can meet the consumption of the current process and the next two hours of operation), and the replenishment request has been initiated as planned and is expected to arrive one hour before the materials are exhausted), a routine guidance instruction is generated. The instruction is strictly formulated according to production process standards, including the specific steps of the current operation, required tools, and process parameters. For example, "Step 1: Take 3 bolts of model A-123 from the intelligent parts box; Step 2: Tighten the bolts with an electric torque wrench to a torque of 5 N·m; Step 3: Confirm the bolt tightening status via the touchscreen and record the data," ensuring that operators follow standard procedures and guaranteeing assembly quality and efficiency.
[0118] Step S53 generates a special guidance instruction. When the material replenishment guidance information indicates a risk of replenishment delay or shortage—that is, the estimated material depletion time is 1.5 hours, while the preset logistics response time is 2 hours, and replenishment cannot arrive on time—a special guidance instruction is generated. The instruction includes a clear warning, specifying the exact time and reason for the material shortage, and provides actionable adjustment suggestions. If there are alternative processes that do not require the shortage material, such as "decorative part pasting," it is recommended to prioritize these alternative processes. If no alternative process exists, it is recommended to pause the current operation and wait for replenishment, and the instruction suggests using the waiting time for tool inspection and cleaning to avoid unnecessary waiting and maximize the use of working hours.
[0119] Step S54 is used to output guidance instructions through multiple channels. The display device at the assembly station uses a high-definition touch screen to visualize the guidance instructions in the form of text and flowcharts. Voice prompts are achieved through directional speakers deployed at the top of the station. Signal lights are deployed in front of the station. When regular guidance instructions are output, the green light is always on. When special guidance instructions are output, the yellow light flashes and a voice broadcast is triggered simultaneously. If there is an emergency shortage risk, the red light flashes. Through visual and auditory multi-sensory collaborative prompts, it is ensured that operators can capture guidance information in a timely manner and accurately execute corresponding operations during busy operations.
[0120] like Figure 2 As shown, the present invention also discloses an intelligent assembly-assisted guidance system, comprising:
[0121] The acquisition module 1 is used to acquire the operation process data of the assembly station in real time. The operation process data includes material inventory data, operation cycle data, work-in-process queue data, and the current operation process.
[0122] Prediction module 2 is used to predict the baseline material consumption trend of the assembly station based on the material inventory data and the corresponding historical material consumption data retrieved from the operation process data.
[0123] The generation module 3 is used to analyze the deviation between the operation cycle data and the standard cycle and the deviation between the work-in-process queue data and the standard queue, and generate dynamic influencing factors that characterize the current production tension.
[0124] The correction module 4 is used to correct the baseline material consumption trend in real time using the dynamic influencing factor to obtain the corrected material demand forecast information, and generate material replenishment guidance information based on the material demand forecast information.
[0125] Output module 5 is used to integrate the material replenishment guidance information with the current work process to generate and output real-time assembly guidance instructions for the operator.
[0126] In one embodiment, the prediction module includes:
[0127] The allocation unit is used to retrieve the material consumption records of the assembly station under historical comparable production conditions to form a historical consumption rate time series.
[0128] The first acquisition unit is used to acquire the operator proficiency decay factor, which represents the current work efficiency status of the operator.
[0129] The second acquisition unit is used to acquire the environmental temperature and humidity influence coefficient, which characterizes the influence of the physical environment on the material handling process.
[0130] The generation unit is used to learn and fit the historical consumption rate time series to obtain an initial baseline consumption rate. The initial baseline consumption rate, the operator proficiency decay factor, and the environmental temperature and humidity influence coefficient are fused and calculated to generate the final baseline material consumption rate, which serves as the baseline material consumption trend for the assembly station.
[0131] This application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the above-described intelligent assembly-assisted guidance method.
[0132] This application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-described intelligent assembly-assisted guidance method.
[0133] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in this application and in the embodiments can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual-speed SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).
[0134] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, apparatus, article, or method that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, apparatus, article, or method. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, apparatus, article, or method that includes that element.
[0135] The above description is merely a preferred embodiment of the present invention and does not limit the scope of this application. Any equivalent results or equivalent process transformations made based on the content of the present invention's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the scope of protection of this application.
Claims
1. A method for intelligent assembly-assisted guidance, characterized in that, Includes the following steps: Real-time acquisition of assembly station operation process data, including material inventory data, operation cycle data, work-in-process queue data, and current operation process; Based on the material inventory data and combined with the corresponding historical material consumption data retrieved from the operation process data, the baseline material consumption trend of the assembly station is predicted. By analyzing the deviations between the operation cycle data and the standard cycle, as well as the deviations between the work-in-process queue data and the standard queue, dynamic influencing factors characterizing the current production tension are generated. The dynamic influencing factors are used to correct the baseline material consumption trend in real time to obtain corrected material demand forecast information, and material replenishment guidance information is generated based on the material demand forecast information. By combining the material replenishment guidance information with the current work process, real-time assembly guidance instructions for the operator are generated and output.
2. The intelligent assembly-assisted guidance method according to claim 1, characterized in that, The steps for acquiring real-time assembly station operation process data include: Periodically read sensor data associated with the material storage container, the sensor data reflecting the real-time physical quantity of the material, and convert the physical quantity into a specific value representing the current inventory as the material inventory data, and obtain the identification information of the material. Acquire video surveillance data of the assembly area, identify the completion cycle of preset standard operation actions, and calculate the operation cycle data; By querying real-time data from the production management system, the quantity of products awaiting processing downstream of the current workstation is obtained and used as the work-in-process queue data.
3. The intelligent assembly-assisted guidance method according to claim 1, characterized in that, The step of predicting the baseline material consumption trend of the assembly station includes: Retrieve the material consumption records of the assembly station under comparable historical production conditions to form a historical consumption rate time series; Obtain the operator proficiency decay factor, which characterizes the current work efficiency status of the operator; Obtain the environmental temperature and humidity influence coefficients that characterize the physical environment's impact on the material handling process; The historical consumption rate time series is learned and fitted to obtain an initial baseline consumption rate. The initial baseline consumption rate, the operator proficiency decay factor, and the environmental temperature and humidity influence coefficient are fused and calculated to generate the final baseline material consumption rate, which serves as the baseline material consumption trend for the assembly station.
4. The intelligent assembly-assisted guidance method according to claim 1, characterized in that, The step of generating dynamic influencing factors characterizing the current level of production stress includes: The beat deviation coefficient is calculated by using the operating beat data and the standard beat, and the beat deviation coefficient is used to quantify the change in actual operating speed. A queue pressure coefficient is calculated using work-in-process queue data and a standard queue length. This queue pressure coefficient is used to quantify the pressure caused by work-in-process backlog. The rhythm deviation coefficient and the queue pressure coefficient are combined to generate a unified dynamic influence factor.
5. The intelligent assembly-assisted guidance method according to claim 1, characterized in that, The step of generating material replenishment guidance information based on the corrected material demand forecast information includes: Based on the baseline material consumption rate and the dynamic influence factor, the real-time predicted consumption rate is obtained, and the real-time predicted consumption rate is used as the corrected material demand forecast information. Based on the current material inventory data and the real-time predicted consumption rate, calculate the expected material depletion time. Based on the expected depletion time of the materials and the preset logistics response time, determine whether a pre-supply instruction needs to be issued; When it is determined that a pre-replenishment instruction needs to be issued, the material replenishment guidance information is generated. The information includes at least material identification information, suggested replenishment quantity, and a time prompt for initiating the replenishment operation.
6. The intelligent assembly-assisted guidance method according to claim 1, characterized in that, The step of generating and outputting real-time assembly guidance instructions for the operator includes: The material replenishment guidance information is integrated with the currently executing assembly process information; If the material replenishment guidance information indicates that the material is sufficient and the replenishment plan is normal, then a regular guidance instruction containing the current standard operating procedure is generated and output. If the material replenishment guidance information indicates a risk of replenishment delay or shortage, a special guidance instruction containing a warning message, a suggestion to wait, or an alternative process will be generated and output. The generated real-time assembly guidance instructions are visualized through the display device at the assembly station and supplemented with voice or signal lights.
7. An intelligent assembly auxiliary guidance system, characterized in that, include: The acquisition module is used to acquire the operation process data of the assembly station in real time. The operation process data includes material inventory data, operation cycle data, work-in-process queue data, and the current operation process. The prediction module is used to predict the baseline material consumption trend of the assembly station based on the material inventory data and the corresponding historical material consumption data retrieved from the operation process data. The generation module is used to analyze the deviation between the operation cycle data and the standard cycle and the deviation between the work-in-process queue data and the standard queue, and generate dynamic influencing factors that characterize the current production tension. The correction module is used to correct the baseline material consumption trend in real time using the dynamic influencing factors, obtain the corrected material demand forecast information, and generate material replenishment guidance information based on the material demand forecast information. The output module is used to integrate the material replenishment guidance information with the current work process to generate and output real-time assembly guidance instructions for the operator.
8. The intelligent assembly auxiliary guidance system according to claim 7, characterized in that, The prediction module includes: The allocation unit is used to retrieve the material consumption records of the assembly station under historical comparable production conditions to form a historical consumption rate time series. The first acquisition unit is used to acquire the operator proficiency decay factor, which represents the current work efficiency status of the operator. The second acquisition unit is used to acquire the environmental temperature and humidity influence coefficient, which characterizes the influence of the physical environment on the material handling process. The generation unit is used to learn and fit the historical consumption rate time series to obtain an initial baseline consumption rate. The initial baseline consumption rate, the operator proficiency decay factor, and the environmental temperature and humidity influence coefficient are fused and calculated to generate the final baseline material consumption rate, which serves as the baseline material consumption trend for the assembly station.
9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.