Food safety sampling inspection process management method and system

Abstract

The invention discloses a food safety sampling inspection process management method and system, and the method comprises the steps: receiving a sampling inspection task which comprises at least two process links; based on a preset PFMEA knowledge base, determining a potential failure mode corresponding to the process link and a risk value of the potential failure mode; according to the risk value, an execution plan is generated through the key chain, the execution plan comprises buffer duration, and the size of the buffer duration is dynamically calculated and determined based on the risk value; in the process of executing the execution plan, the progress is monitored, and the buffer duration consumption rate is calculated; based on the risk values of all the currently executed process links, a dynamic early warning threshold value is obtained through calculation; and when the buffer duration consumption rate is greater than or equal to an early warning threshold value, triggering early warning and executing corresponding measures corresponding to the potential failure mode. According to the method, unified quantitative management and dynamic cooperative regulation and control of quality risks and progress risks in the food safety sampling inspection process are realized.

Description

Technical Field

[0001] This invention relates to the field of food safety, and in particular to a method and apparatus for managing the food safety sampling and inspection process. Background Technology

[0002] Food safety sampling inspection is a core regulatory tool for ensuring food safety. The process typically involves multiple stages, including task assignment, sampling, logistics, testing, and reporting. It involves various resources such as sampling personnel, vehicles, and testing equipment, and is accompanied by potential risks such as sample failure, data deviation, and delays. Therefore, efficient and reliable management of the sampling inspection process is crucial.

[0003] Existing food safety sampling and management systems and related technologies suffer from the following shortcomings: Passive risk identification: Existing systems primarily focus on process recording and tracking, lacking mechanisms for proactive, structured identification and assessment of potential failure modes. Risk response often relies on post-event remediation rather than pre-event prevention, failing to fundamentally eliminate quality incidents. Inefficient schedule management: Traditional methods (such as the critical path method) do not adequately consider resource constraints, and task duration estimates typically include a large amount of implicit safety time, making them susceptible to human factors such as "student syndrome," leading to significant deviations between project plans and actual execution, and low resource utilization.

[0004] Isolated Risk Management: Existing technologies typically manage quality risks (such as sample failures and data errors) and schedule risks (such as task delays) separately, lacking a unified quantitative model to link the two. This results in a fragmented risk management system, making it difficult to comprehensively optimize and control from a holistic system perspective. Summary of the Invention

[0005] In order to overcome the deficiencies in the prior art, the first objective of this invention is to provide a method for managing the food safety sampling inspection process, and the second objective of this invention is to provide a management system for the food safety sampling inspection process.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0007] Firstly, a method for managing the food safety sampling inspection process, including:

[0008] Receive a random inspection task, which includes at least two process steps;

[0009] Based on a pre-set PFMEA knowledge base, the potential failure modes and their risk values ​​corresponding to the process steps are determined.

[0010] Based on the risk value, an execution plan is generated through the critical chain, wherein the execution plan includes a buffer duration, the size of which is dynamically calculated and determined based on the risk value;

[0011] During the execution of the execution plan, the progress is monitored and the buffer time consumption rate is calculated; based on the risk values ​​of all currently executed process stages, a dynamic early warning threshold is calculated.

[0012] When the buffer duration consumption rate is greater than or equal to the warning threshold, a warning is triggered and the corresponding countermeasures for the potential failure mode are executed.

[0013] Optionally, the steps for generating an execution plan include:

[0014] Identify the logical relationships and resource constraints among the various process stages in the sampling task;

[0015] Based on the aforementioned logical relationships and resource constraints, the critical chain is determined;

[0016] For each process step in the critical chain, determine its pessimistic and aggressive projected project durations;

[0017] Based on the pessimistic duration, aggressive duration, and risk value of all process links in the critical chain, the buffer time is calculated.

[0018] The output contains the critical chain and buffer duration.

[0019] Optionally, the calculation of the buffer duration based on the pessimistic duration, aggressive duration, and risk value of all process stages on the critical chain includes:

[0020] For each process step on the critical chain, calculate the difference between its pessimistic and aggressive durations, and use this as the first safe time for that step.

[0021] For each process step, an adjustment factor is calculated based on its corresponding risk value and a preset risk sensitivity coefficient.

[0022] Multiply the first safety time of each process step by the corresponding adjustment factor to obtain the second safety time of that step;

[0023] The buffer duration is calculated based on the second safety time of all process steps on the critical chain.

[0024] Optionally, the risk sensitivity coefficient is determined through the following steps:

[0025] Based on the type of the sampling task, a preset coefficient value is used as the risk sensitivity coefficient;

[0026] The types of sampling inspection tasks include routine supervision sampling inspection, special sampling inspection and emergency sampling inspection, and the coefficient values ​​of the three are pre-configured to increase sequentially.

[0027] Optionally, the risk sensitivity coefficient is calibrated through the following steps:

[0028] Obtain actual execution data for historical spot check tasks;

[0029] Based on the comparison between the actual execution data and the historical sampling task plan data, the optimized risk sensitivity coefficient is calculated through regression analysis.

[0030] Optionally, the monitoring of progress and calculation of buffer duration consumption rate includes:

[0031] Collect real-time progress data for each stage of the process;

[0032] Based on the real-time progress data, the amount of buffer time consumed is determined, and the buffer time consumption rate is calculated.

[0033] Optionally, the calculation of a dynamic early warning threshold based on the risk values ​​of all currently executed process stages includes:

[0034] The risk index is calculated based on the set of risk values ​​for all currently executing process stages.

[0035] Based on the preset mapping relationship, the early warning threshold corresponding to the risk index is determined.

[0036] Optionally, the method further includes an optimization step, which includes at least one of the following:

[0037] After the sampling inspection task is completed, the frequency score of the risk value used to calculate the potential failure mode in the PFMEA knowledge base is updated according to the actual failure situation.

[0038] Based on the actual buffer duration consumption rate, the parameters of the buffer duration calculation model are calibrated, including the risk sensitivity coefficient.

[0039] Optionally, updating the frequency score in the PFMEA knowledge base used to calculate the risk value of potential failure modes based on actual failure conditions includes:

[0040] Record the number of times each potential failure mode occurs during the execution of the sampling task;

[0041] The frequency score of the failure mode is adjusted based on the number of occurrences after statistical analysis.

[0042] Optionally, triggering the early warning and executing the corresponding response measures for the potential failure mode includes a resource scheduling step:

[0043] When an alert is triggered, it indicates that multiple process stages are competing for the same resource;

[0044] According to a preset priority rule, the multiple process steps are sorted, wherein process steps belonging to the critical chain have a higher priority than non-critical chain process steps, and under the same conditions, process steps with higher risk values ​​have a higher priority.

[0045] Based on the sorting results, the resources are allocated to the process steps with the highest priority.

[0046] Secondly, a food safety sampling inspection process management system includes:

[0047] The task receiving module is used to receive sampling inspection tasks, which include at least two process steps.

[0048] The risk analysis module is used to determine the potential failure modes and their risk values ​​corresponding to the process steps based on a preset PFMEA knowledge base.

[0049] The plan generation module is used to generate an execution plan through the critical chain based on the risk value, wherein the execution plan includes a buffer duration, the size of which is dynamically calculated and determined based on the risk value;

[0050] The monitoring and early warning module is used to monitor the progress and calculate the buffer time consumption rate during the execution of the execution plan; and to calculate a dynamic early warning threshold based on the risk value of the current execution process stage.

[0051] The response execution module is used to trigger an early warning and execute the corresponding countermeasures for the potential failure mode when the buffer duration consumption rate is greater than or equal to the early warning threshold.

[0052] Thirdly, an electronic device includes:

[0053] One or more processors;

[0054] Memory;

[0055] One or more computer programs;

[0056] The one or more computer programs are stored in the memory and configured to be executed by the one or more processors, and the one or more computer programs are configured to perform the above-described food safety sampling inspection process management method.

[0057] Fourthly, a computer-readable storage medium having a computer program stored thereon, characterized in that the computer program, when executed by a processor, implements the above-mentioned food safety sampling inspection process management method.

[0058] Due to the application of the above-mentioned technical solution, the present invention has the following advantages compared with the prior art:

[0059] 1. By combining the proactive risk identification of process failure mode and impact analysis (PFMEA) with the dynamic process monitoring of critical chain technologies, early warnings can be triggered and pre-set countermeasures can be initiated before potential failure modes have an impact, thereby helping to reduce the probability of quality accidents and improve the reliability of the sampling process.

[0060] 2. By introducing a risk-enhanced buffer duration calculation model, the risk priority coefficient is incorporated into the buffer duration calculation, transforming the setting of the buffer duration from relying on experience to algorithmic decision-making based on risk quantification. The generated execution plan is more in line with the actual risk situation, which helps to improve the executability of the plan, the accuracy of the schedule forecast, and the overall execution efficiency.

[0061] 3. By dynamically calculating the early warning threshold based on the risk value set of all currently executed process stages, the early warning mechanism can adapt to the real-time risk situation of the project, realizing the transformation from static threshold monitoring to dynamic risk perception, and improving the accuracy of early warning and the timeliness of management intervention.

[0062] 4. By updating the frequency scores of potential failure modes in the PFMEA knowledge base based on actual data after execution and calibrating the risk sensitivity coefficient using historical data, the risk model and planning model can continuously optimize themselves, gradually improving the management accuracy of subsequent sampling tasks and realizing the closed-loop integration of quality risk and schedule management.

[0063] 5. By setting resource allocation rules that prioritize critical chains and high-risk items, resources can be automatically allocated to the most critical and risky links for task completion when resource conflicts occur. This helps to solve resource competition problems and improve the utilization efficiency of critical resources.

[0064] 6. By using critical chain, buffer duration consumption status and comprehensive risk index as core monitoring indicators, it provides managers with a global and intuitive project status view, enabling them to focus on anomaly handling and key decisions, and reducing the workload of daily progress tracking and coordination.

[0065] To make the above and other objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0066] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0067] Figure 1 This is a flowchart illustrating the management method for food safety sampling and inspection. Detailed Implementation

[0068] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0069] Example: See Figure 1 As shown, a method for managing the food safety sampling inspection process includes:

[0070] S1. Receive a sampling inspection task, which includes at least two process steps.

[0071] Specifically, the system receives sampling inspection tasks submitted by external systems or users through preset task receiving interfaces, such as web service interfaces, message queues, or manual input interfaces. These sampling inspection tasks typically originate from government regulatory plans, enterprise self-inspection needs, or responses to sudden food safety incidents.

[0072] When creating a sampling inspection task, at least the basic attribute information of the task must be included, such as task number, task source, food category involved, target sampling location, planned sampling quantity, specified testing items, and overall task completion time limit requirements.

[0073] The sampling inspection task includes at least two process stages. Each process stage refers to a sub-task or phase that constitutes a complete sampling inspection workflow, with a clear logical sequence and dependencies. For example, food safety sampling inspection process stages may include, but are not limited to: task assignment, planning, on-site sampling, sample transportation, sample handover, sample monitoring, data review, and report generation. Each process stage typically has its own executing entity (e.g., sampling personnel, testing personnel), required resources (e.g., sampling tools, testing equipment), estimated timeframe, and logical relationship with other stages (e.g., sequential or parallel).

[0074] In one optional implementation, when the system receives a sampling inspection task, it can automatically parse the process steps included in the task based on a predefined sampling inspection process template. In another optional implementation, the process steps can be manually defined and assembled by the user through configuration.

[0075] S2. Based on the preset PFMEA knowledge base, determine the potential failure modes and their risk values ​​corresponding to the process steps.

[0076] After receiving the sampling inspection task and analyzing the process steps, the system assigns a quantified risk value to each process step. Specifically, the system accesses and queries the built-in PFMEA knowledge base (Impact Analysis Knowledge Base). The PFMEA knowledge base is a pre-built structured database that stores historical risk data and analysis results related to the food safety sampling inspection process. Specifically, the core data fields of the PFMEA knowledge base include, but are not limited to: process step identifier, potential failure mode, potential failure cause, potential failure impact, severity score, frequency score, detectability score, risk priority coefficient, and corresponding countermeasures. Among them, the initial values ​​of severity score, frequency score, and detectability score are pre-set based on a comprehensive preset of historical sampling inspection case data, industry regulatory standards, and expert review results.

[0077] The potential failure mode refers to a non-conformity or defect that may occur in a process stage, causing the output of that stage to fail to meet the expected requirements. For example, a potential failure mode in the "sample transportation" stage could be "the sample is not stored at the required temperature during transportation", and a potential failure mode in the "sample testing" stage could be "expired test reagents are used".

[0078] In an optional implementation, the process of determining the potential failure modes and their risk values ​​corresponding to process stages is as follows: Based on each process stage determined in step S1, the system automatically matches it with its process stage identifier or type as query conditions in the PFMEA knowledge base to obtain one or more potential failure mode records associated with that process stage. For each matched potential failure mode, the system extracts its current frequency score, severity score, and detectability score, and calculates its risk value. The risk value described in this application is a risk priority coefficient, calculated as follows: Risk Priority Coefficient = Frequency Score × Severity Score × Detectability Score. The frequency score measures the likelihood or probability of the potential failure mode occurring; the severity score measures the severity of the consequences of the potential failure mode on the final sampling results if it occurs; and the detectability score measures the ease with which the potential failure mode can be successfully identified before it has an impact using existing control measures. For example, for the potential failure mode of "not storing samples at the required temperature during transportation", if its frequency score is 5, severity score is 8, and detectability score is 4, then its risk priority coefficient is 5×8×4=160.

[0079] The PFMEA knowledge base supports dynamic maintenance for self-learning. In one optional implementation, after a sampling inspection task is completed, if the system detects a significant deviation between the actual occurrence of a potential failure mode and a preset frequency score (e.g., the actual occurrence frequency deviates from the theoretical frequency corresponding to the preset frequency by more than a preset threshold, such as 20%), it can automatically trigger a score review process. After review and confirmation, the frequency score of the failure mode is updated. In another optional implementation, if a new process step or a potential failure mode not yet recorded in the PFMEA knowledge base is encountered, the system supports users manually supplementing relevant information and score data. After the necessary review process, this information is written into the PFMEA knowledge base to continuously expand its risk coverage.

[0080] S3. Based on the risk value, generate an execution plan through the critical chain, wherein the execution plan includes a buffer duration, the size of which is dynamically calculated and determined based on the risk value.

[0081] After obtaining the process steps and their corresponding risk values ​​in step S2, the system enters the execution plan generation stage. The purpose of this step is to create an execution plan that incorporates risk information and can effectively address uncertainties during the execution process.

[0082] The generation of the execution plan relies on critical chain technology. The critical chain refers to a sequence of activities that determines the shortest completion time of the entire sampling task, identified after comprehensively considering the logical relationships (such as sequence and dependencies) and resource constraints (such as the availability of personnel and equipment) between various process stages. The task sequence is an ordered set of process stages constituting the critical chain; that is, a series of process stages arranged according to their logical execution order and resource dependencies. A delay in any stage of this sequence will directly lead to a delay in the final completion time of the entire project.

[0083] Specifically, the process of generating an execution plan includes:

[0084] S31. Identify the logical relationships and resource constraints between each process step in the sampling task.

[0085] Specifically, the system first performs a systematic analysis of the process steps defined in step S1. This analysis includes two dimensions: the logical relationship between the process steps and the resource constraint dimension.

[0086] In terms of the logical relationships between process stages, the system parses the inherent temporal dependencies between each process stage. These logical relationships define the execution order between process stages. For example, the "sample testing" process stage can only begin after the "sample receiving" process stage is completed, constituting a serial relationship. For process stages that may be parallel, such as multiple different testing items, they may form a parallel relationship.

[0087] In terms of resource constraints, the system identifies and confirms the types and quantities of resources necessary to execute each process step. These resource constraints refer to resources with limited availability that may become bottlenecks in the execution of the sampling task, such as sampling personnel, testing equipment, and transport vehicles. By accessing and querying integrated resource calendars or real-time resource pool status information, the system can identify whether multiple process steps are competing for the same limited resource within a specific time period of the sampling task timeline, i.e., resource conflict.

[0088] S32. Based on the logical relationships and resource constraints, determine the critical chain.

[0089] In one optional implementation, the system employs a scheduling algorithm to perform calculations and resource allocation while simultaneously satisfying the logical relationships and resource availability constraints between all process stages. The critical chain is defined as: under resource constraints, the sequence of process stages with the longest total duration from the start to the end of the sampling task. The actual duration of this sequence determines the shortest time required to complete the entire sampling task.

[0090] In one optional implementation, the process of determining the critical chain may include: First, based on the logical relationship network between process stages, a theoretical critical path is initially calculated. Then, resource constraints are introduced; when the system detects resource contention, resource arbitration and scheduling are performed according to preset rules, adjusting the planned time of the affected process stages. Finally, in the generated schedule, the sequence of process stages with the minimum total float (e.g., zero or close to zero) is determined as the critical chain. The total float refers to the maximum amount of time that the start or end time of a process stage can be delayed without affecting the final total duration of the sampling task. Since any delay in any process stage on the critical chain will directly lead to a delay in the total duration of the sampling task, this critical chain is the focus of subsequent schedule monitoring and buffer time management.

[0091] S33. For each process step on the critical chain, determine its pessimistic duration, aggressive duration, and corresponding risk value.

[0092] After the critical chain is determined, the system enters the data preparation phase before calculating the buffer duration. The system first determines the pessimistic and aggressive project durations for each process step on the critical chain, as well as the risk value (i.e., risk priority coefficient) determined in step S2.

[0093] The project duration refers to the estimated continuous working time required to complete a process stage (such as "on-site sampling" or "sample testing"). It is the most basic timing parameter in sampling tasks, used to schedule the start and end times of each process stage. In traditional project management methods, the project duration estimated separately for each process stage usually includes safety time reserved to cope with potential delays in that process stage. Safety time refers to the additional time margin added to cope with various uncertainties that may occur during the execution of each process stage (such as personnel operation delays, temporary equipment failures, resource coordination waits, etc.). Currently, the decentralized safety time management method is prone to inefficiency due to factors such as "student syndrome." Student syndrome is a common behavioral phenomenon in which task performers, even when sufficient time is allocated for the task, tend to delay starting work until the deadline is approaching. This results in wasted early time, and the decentralized safety time cannot effectively cope with actual risks.

[0094] To address this issue, this application employs critical chain project management. Specifically, the system estimates an aggressive duration for each process stage on the critical chain. The aggressive duration refers to the estimated time required to complete the process stage under ideal conditions, excluding most confounding factors; it typically corresponds to a moderate confidence level of completion (e.g., 50%). In contrast, the pessimistic duration typically corresponds to a higher confidence level of completion (e.g., 90%), representing the longest possible time after considering various adverse factors, and is therefore significantly longer than the aggressive duration. The specific value of the confidence level of completion can be predefined by the system based on the organization's historical data statistical analysis results, or configured by the user.

[0095] As an optional implementation, the system can perform an intermediate calculation to provide an optimized reference benchmark for subsequent risk-enhanced buffer duration calculation models. Specifically, the system identifies and defines the difference between the pessimistic and aggressive durations of each process stage on the critical chain as the first safety time for that process stage. The first safety time represents a time margin reserved specifically for that stage to cope with its own uncertainties.

[0096] Next, the system aggregates the first safety times of all process steps on the critical chain to form an initial buffer duration. One common aggregation method is to calculate the square root of the sum of the squares of all first safety times to obtain the initial buffer duration. The formula can be expressed as:

[0097] ;

[0098] in, This represents the calculated initial buffer duration;

[0099] This indicates the total number of process steps on the critical chain;

[0100] Indicates the first [chain number] on the key chain. The pessimistic project timeline for each stage of the process;

[0101] Indicates the first [chain number] on the key chain. The project timeline for each stage of the process.

[0102] It is important to emphasize that this initial buffer duration... It does not directly participate in the calculation of the risk-enhanced buffer duration in subsequent step S34. Its main purpose is to serve as a benchmark reference value for calculating the expected buffer duration consumption rate during the model optimization and calibration phase after the entire sampling task is completed (see subsequent step S62).

[0103] S34. Based on the pessimistic duration, aggressive duration, and risk value of all process links in the critical chain, the buffer duration is calculated.

[0104] After completing the data preparation in step S33, the system invokes the risk-enhanced buffer duration calculation model to generate the buffer duration that will be directly incorporated into the execution plan. This model directly and independently integrates the risk value representing quality risk into the safety time estimation of each process stage, and obtains the final protective time buffer through aggregation calculations, thereby achieving comprehensive control over both schedule and quality risks.

[0105] The specific calculation process includes the following steps:

[0106] S341. Calculate the first safe time.

[0107] For each process stage in the critical chain, calculate the difference between its pessimistic and aggressive durations, and use this difference as the first safe time for that stage. That is, the... The first safety time for each process step is The first safety time reflects the basic time margin reserved for the inherent uncertainties of the process, without considering the impact of risk values.

[0108] S342. Calculate the adjustment factor.

[0109] For each process step, based on its corresponding risk value (risk priority coefficient) ) and preset risk sensitivity coefficient An adjustment factor is calculated. The risk sensitivity coefficient... It is a configurable system parameter used to globally control the overall impact of risk values ​​on the final size of the buffer duration.

[0110] In an optional implementation, for the first Adjustment factors for each process step The calculation method is as follows:

[0111] ;

[0112] in:

[0113] This represents the risk sensitivity coefficient;

[0114] Indicates the first [chain number] on the key chain. The risk priority coefficient corresponding to each process step;

[0115] A preset normalization constant (e.g., a value of 1000) is used to balance the dimensions of the risk value so that it is on a comparable order of magnitude to the value 1.

[0116] This formula ensures that when the risk value is zero or extremely low, the adjustment factor approaches 1, and the first safety time is not amplified. As the risk value increases, the adjustment factor increases accordingly, thus providing more time reserve for the safety time of this process step.

[0117] S343, Aggregate to obtain buffer duration.

[0118] Based on the second safety time of all process steps in the critical chain, the buffer time required for the entire sampling task is calculated. To reasonably assess the overall uncertainty arising from the cumulative risks across multiple stages, the system employs an aggregation algorithm based on the square root of the sum of squares. The buffer duration... The calculation formula is as follows:

[0119] ;

[0120] in, This represents the total number of process steps in the critical chain. Statistically, this formula treats the adjusted second safety time for each process step as an independent risk variable, and its aggregated result (buffer duration) provides a unified time reserve for addressing the comprehensive risks of the entire critical chain. The buffer duration calculated in this way deeply integrates the schedule uncertainty information and quality risk information of each process step, achieving unified management and dynamic balance of these two types of risks at the quantitative level.

[0121] To further illustrate the calculation process and effects of this application, an exemplary calculation is provided below:

[0122] Suppose that the critical chain of a certain sampling inspection task consists of three process stages (A, B, C), and its parameters and risk values ​​are as follows:

[0123] Process step Pessimistic duration (days) Aggressive duration (days) Risk priority factor A 10 6 200 B 8 5 100 C 12 8 300

[0124] 1. Calculate the first safe time for each process step. :

[0125] A: 10 − 6 = 4 days;

[0126] B: 8 - 5 = 3 days;

[0127] C: 12 − 8 = 4 days;

[0128] 2. Calculate the risk adjustment factor for each activity stage. :

[0129] A: 1 + 0.15 × (200 / 1000) = 1.03;

[0130] B: 1 + 0.15 × (100 / 1000) = 1.015;

[0131] C: 1 + 0.15 × (300 / 1000) = 1.045;

[0132] 3. Calculate the second safety time for each activity (first safety time × adjustment factor):

[0133] A: 4 × 1.03 = 4.12 days;

[0134] B: 3 × 1.015 = 3.045 days;

[0135] C: 4 × 1.045 = 4.18 days;

[0136] 4. Calculate the buffer duration:

[0137] ;

[0138] In contrast, if PFMEA risk (i.e., the traditional critical chain approach) is not considered, the buffer time is:

[0139] ;

[0140] As can be seen, the risk-enhanced buffer duration calculation model provided in this application, by taking into account the quality risks of each stage, increases the buffer from 6.40 days to 6.62 days, providing more targeted time reserves for high-risk stages, making the execution plan more resilient, and achieving unified management of quality risks and schedule risks at the quantitative level.

[0141] The risk sensitivity coefficient The value of directly affects the magnitude of risk adjustment, and the determination and optimization of its specific value is the key to the system's self-learning capability.

[0142] In an optional implementation, the system uses a preset coefficient value as a risk sensitivity coefficient based on the type of the sampling task. The types of sampling inspection tasks include routine supervision sampling inspection, special sampling inspection and emergency sampling inspection, and the coefficient values ​​of the three are pre-configured to increase sequentially. The value range is typically 0.1-0.3. For example:

[0143] For routine supervision and spot checks, the processes are standardized and the risks are relatively controllable, so a lower risk sensitivity coefficient is used, for example... Take 0.12;

[0144] For special sampling inspection projects, which are highly targeted and complex, a medium risk sensitivity coefficient is adopted, for example... Take 0.18;

[0145] For emergency sampling inspection projects, which have high timeliness requirements and significant uncertainties, a higher risk sensitivity coefficient is adopted, for example... Take 0.25.

[0146] In this way, different types of sampling tasks can obtain buffer protection that matches their risk characteristics.

[0147] In another alternative implementation, the system call risk sensitivity coefficient is calibrated using historical data. The calibration process is a continuous optimization step, designed to ensure that... The value of is more likely to reflect the risk consumption pattern in the actual operation of the organization. For specific calibration methods, please refer to the following step S62.

[0148] S35. Output the execution plan containing the critical chain and buffer duration.

[0149] After calculating the buffer duration, the system enters the final synthesis and output stage of the execution plan. The purpose of this step is to integrate the results of all the aforementioned planning steps into a complete, executable execution plan that includes embedded risk information.

[0150] Specifically, the system synthesizes the critical chain determined in step S32 (i.e., the task sequence that determines the shortest project duration), the logical relationships and resource allocation schemes between various process stages analyzed in step S31, and the buffer duration calculated in step S34. The execution plan refers to a comprehensive scheme that includes the task sequence, time arrangement, resource allocation, and buffer duration.

[0151] During the synthesis process, the system places the buffer duration at the end of the critical chain. This buffer is a unified time reserve set up to protect the overall duration of the sampling task from uncertainties on the critical chain. Simultaneously, for nodes on non-critical chain paths that merge into the critical chain, the system can selectively calculate and set the merge buffer duration. This merge buffer duration is a local time buffer set up to prevent delays in non-critical chain processes from propagating to the critical chain.

[0152] The final execution plan clearly includes the following elements: the start and end times of each process step; the resources (personnel, equipment, vehicles) required to execute each process step and their allocation time periods; the critical chain path of the sampling task; the location and size of the sampling task buffer time, and the location and size of the optional inflow buffer time.

[0153] After the execution plan is generated, the system transforms it into an executable state through its publishing mechanism. In one optional implementation, the system publishes the execution plan to the task management service and automatically distributes it to the terminals (such as web management terminals or mobile apps) of relevant responsible personnel (e.g., samplers, testers), guiding them to begin execution. In another optional implementation, the system persistently stores the final execution plan data in a database, providing a data foundation for the visualization monitoring and decision support module for subsequent real-time progress monitoring and dynamic early warning.

[0154] S4. During the execution of the execution plan, monitor the progress and calculate the buffer time consumption rate. Based on the risk values ​​of all currently executed process stages, calculate a dynamic early warning threshold.

[0155] Once the execution plan is released and initiated, the system enters the dynamic execution and monitoring phase. The purpose of this step is to shift from static plan management to dynamic perception and response to risk situations. The risk situation refers to the immediate risk status of the sampling task, determined by all currently executing process stages and their risk levels.

[0156] Specifically, this step includes the following steps:

[0157] S41. Collect real-time progress data for each process stage.

[0158] The system automatically collects real-time data from the execution site through multiple channels. Specifically, task performers such as sampling and testing personnel can report the start, completion, and pause status of each step via a mobile app; for automated equipment with interfaces, the system can directly obtain its operating status and testing progress; simultaneously, the system also supports manual input of the completion status of key nodes via a web interface. This real-time progress data is continuously transmitted and stored in the system's time-series database, providing a data foundation for subsequent calculations.

[0159] S42. Based on the real-time progress data, determine the amount of buffer time consumed and calculate the buffer time consumption rate.

[0160] The system dynamically calculates the actual execution progress of the sampling task based on the real-time progress data of each process step collected in step S41. Specifically, the system compares the current time with the current planned completion time of the critical chain in the execution plan to determine the buffer time consumption. The buffer time consumption refers to the difference between the current planned completion time of the critical chain and the current time, which represents the buffer time reserved to cope with uncertainties (i.e., the buffer time calculated in step S34). The portion that has already been used. If the current time is later than the current planned completion time of the critical chain, it means that the buffer time has been consumed.

[0161] Subsequently, the system calculates the buffer duration consumption rate. The buffer duration consumption rate is the ratio of buffer duration consumed to the total buffer duration (…). The ratio of the size of the buffer duration is used to quantify the degree of buffer duration usage. The calculation formula is: Buffer duration consumption rate = (Buffer duration consumption / Buffer duration) × 100%.

[0162] S43. Based on the risk values ​​of all currently executed process stages, a dynamic early warning threshold is calculated.

[0163] Traditional critical chain management uses fixed buffer time consumption rate thresholds (e.g., a 33% consumption rate triggers a yellow alert, and a 66% consumption rate triggers a red alert) to trigger different levels of alerts. This fixed threshold method has significant drawbacks: it fails to consider the risk differences between different sampling tasks and at different execution stages of the same task, resulting in a rigid alert mechanism. For high-risk tasks or high-risk stages, a fixed high threshold may cause alerts to be issued too late, missing the opportunity for early intervention; while for low-risk situations, a fixed low threshold may trigger unnecessary alerts, resulting in a waste of management resources.

[0164] This application adjusts the early warning threshold based on the set of risk values ​​for all currently executed process stages, representing a risk-adaptive dynamic early warning threshold method. This method ensures that the early warning threshold is no longer a fixed value, but rather dynamically fluctuates according to the risk profile of the current sampling task. The risk profile refers to the immediate risk status of the sampling task, jointly determined by all currently executed process stages and their risk priority coefficients. Specifically, the system collects the risk priority coefficients of all current process stages and calculates a comprehensive risk index based on the set of these risk priority coefficients, then dynamically adjusts the early warning threshold according to this comprehensive risk index.

[0165] Specifically, a dynamic early warning threshold is calculated based on the risk values ​​of all currently executed process stages, including:

[0166] S431. Calculate the risk index based on the risk values ​​of all currently executing process steps.

[0167] Specifically, the system first identifies all currently executing process steps. It should be noted that there may be one or multiple currently executing process steps, as some process steps in a sampling task are logically allowed to be executed in parallel, or may be carried out simultaneously due to sufficient resources. For example, in the same sampling task, "on-site sampling" personnel may simultaneously perform sampling operations at multiple different locations; these parallel sampling activities constitute multiple simultaneously executing "on-site sampling" process steps. Similarly, in the laboratory testing phase, different testers may be using multiple devices to perform parallel "sample testing" on different items or different batches of the same sample.

[0168] For each currently executing process step, the system retrieves its risk priority coefficient from its corresponding process failure mode and impact analysis knowledge base. Subsequently, the system calculates a comprehensive risk index based on the set of these risk priority coefficients to characterize the overall risk level of the current sampling task.

[0169] In an optional implementation, the risk index is defined as a comprehensive risk index, which is the average of the risk priority coefficients of all currently executing process stages, and its calculation formula is as follows:

[0170] ;

[0171] in, This indicates the total number of process steps currently being executed. Indicates the first The risk priority coefficients for each process stage are used as the average risk index. This effectively characterizes the average risk load of a unit process stage, making risk assessments comparable across different numbers of activities, thus logically aligning with the average risk value of the critical chain during the planning phase.

[0172] In another alternative implementation, the risk index is defined as the sum of the risk priority coefficients of all currently executing process steps, i.e.:

[0173] ;

[0174] in, This indicates the total number of process steps currently being executed. Indicates the first Risk priority coefficient for each process stage.

[0175] This approach focuses on the total risk load of all concurrent processes and is more sensitive to the number of concurrent high-risk activities.

[0176] S432. Determine the early warning threshold corresponding to the comprehensive risk index according to the preset mapping relationship.

[0177] After obtaining the risk index, the system determines a dynamic early warning threshold through a preset mapping relationship. The preset mapping relationship defines the conversion rules from the risk index to the specific early warning threshold.

[0178] In an optional implementation, the mapping relationship is calculated by using a risk adjustment factor. This is used to scale the baseline warning threshold (e.g., 33%, 66%) based on the real-time risk level.

[0179] If the comprehensive risk index (average value) is used as the risk index, then the risk adjustment coefficient... The calculation formula is:

[0180] ;

[0181] Right now ;

[0182] If the total risk load (sum) is used as the risk index, then the risk adjustment coefficient... The calculation formula is:

[0183] ;

[0184] Right now ;

[0185] in, This indicates the total number of process steps currently being executed. Indicates the first The risk priority coefficient for each process step is 500, which is a preset normalization constant.

[0186] Subsequently, the system adjusts the risk coefficient. Set dynamic warning thresholds: The green warning threshold is... The yellow alert threshold is When the buffer time consumption rate is less than or equal to the green warning threshold, the system is in the green zone; when the buffer time consumption rate is greater than the green warning threshold but less than or equal to the yellow warning threshold, the system is in the yellow zone; and when the buffer time consumption rate is greater than the yellow warning threshold, the system is in the red zone. This dynamic adjustment mechanism based on a comprehensive risk index allows the warning threshold to decrease as the overall risk level increases, thereby achieving a more sensitive risk response.

[0187] S5. When the buffer duration consumption rate is greater than or equal to the warning threshold, a warning is triggered and the corresponding countermeasures for the potential failure mode are executed.

[0188] When the buffer time consumption rate calculated by the system in step S42 reaches or exceeds the dynamic early warning threshold determined in step S432, the system automatically triggers an early warning at the corresponding level. The operation of triggering the early warning includes generating and distributing early warning information. The early warning information is a structured notification message automatically generated by the system, and its content typically includes the relevant sampling task number, the triggered early warning level, the current buffer time consumption rate, and the identifier of the currently high-risk process link and its corresponding risk priority coefficient. The system distributes the early warning information to preset responsible persons through an integrated messaging service, which may include SMS, email, or an interface with internal instant messaging tools to ensure timely delivery of early warnings.

[0189] In one optional implementation, the system executes corresponding countermeasures for the relevant potential failure modes simultaneously with or after triggering an early warning. The system automatically retrieves and associates pre-set corrective and preventative measures from the process failure mode and related potential failure mode from the process failure mode and related knowledge base, and pushes this list of measures as action suggestions to the relevant responsible persons. This achieves automatic association between risk warnings and targeted countermeasures, providing decision support for managers.

[0190] In an optional implementation, the system also performs intelligent resource scheduling. When an alert is triggered, the system detects resource conflicts in real time. A resource conflict refers to a situation where multiple process stages simultaneously compete for the same critical resource. The critical resource refers to a limited number of resources that are essential to task execution, such as specific testing equipment or specialized testing personnel. If a resource conflict is detected, the system automatically arbitrates and schedules resources according to preset priority rules. These priority rules stipulate that process stages belonging to the critical chain have higher priority than non-critical chain process stages; and when priority categories are the same, process stages with higher risk priority coefficients have higher priority. The system sorts competing process stages according to these rules and dynamically allocates the critical resource to the highest-priority process stage to ensure that resources are prioritized for activities most critical to the success and highest risk of the sampling inspection task.

[0191] Through the above process, this application realizes automated and intelligent closed-loop management from risk monitoring and early warning triggering to measure execution and resource scheduling, which helps to improve the reliability of food safety sampling and inspection process and the efficiency of responding to emergencies.

[0192] S6. Optimization steps.

[0193] After the sampling inspection task is completed, the system enters the closed-loop optimization phase. The purpose of this step is to use the actual data generated during the execution of this task to update and calibrate the parameters in the system's PFMEA knowledge base and risk-enhanced buffer duration calculation model, enabling the system to have self-learning and continuous optimization capabilities, thereby gradually improving the management accuracy and risk response capabilities of subsequent sampling inspection tasks.

[0194] In an optional implementation, the optimization step includes:

[0195] S61. Update the frequency score of the risk value used to calculate the potential failure mode in the PFMEA knowledge base according to the actual failure situation.

[0196] During task execution, the system automatically records and statistically analyzes the occurrence of each potential failure mode during the actual execution process. This occurrence includes whether a specific failure mode occurred and the number of times it occurred. After the sampling inspection task is completed, the system dynamically updates the frequency score of the corresponding failure mode in the PFMEA knowledge base based on these statistical data.

[0197] In an optional implementation, the frequency score update process includes: the system comparing the actual number of occurrences of each potential failure mode in the current task with its expected frequency value, where the expected frequency value is the theoretical occurrence frequency calculated based on the original frequency score of the failure mode. If there is a significant difference between the actual number of occurrences and the expected frequency value, for example, if the deviation between the actual frequency and the theoretical frequency exceeds a preset threshold (e.g., 20%), the system can automatically trigger a frequency score review process for that failure mode, or generate a data anomaly report for management personnel to review. After review and confirmation, the system will adjust and update the frequency score of the failure mode in the knowledge base according to predefined rules or algorithms. Through the above operations, the risk data in the PFMEA knowledge base can continuously approach the real situation, and the accuracy of its risk profile continues to evolve as the task is executed.

[0198] S62. Based on actual buffer duration consumption data, the risk sensitivity coefficient in the risk-enhanced buffer duration calculation model is... Perform calibration.

[0199] To continuously improve the predictive accuracy of the risk-enhanced buffer duration calculation model, the system utilizes accumulated historical sampling task execution data to analyze the risk sensitivity coefficient, a key parameter in the model. Perform periodic or triggered calibration optimization.

[0200] First, the system archives key data after each task is completed. For the completed task, the system records its actual buffer time consumption rate (based on the buffer time calculated in step S34). The consumption rate is calculated, and its expected buffer duration consumption rate is also calculated. The expected consumption rate is a theoretical value used as a calibration benchmark, and it is calculated as follows: assuming that this task adopts traditional critical chain management without introducing risk adjustment, that is, using the initial buffer duration calculated in step S33. As a buffer, the theoretical consumption level is derived from actual progress data. This data (actual consumption rate, expected consumption rate, average risk value of the critical chain, etc.) will be added to the system's historical task dataset.

[0201] Subsequently, when calibration conditions are met (such as historical data volume reaching a threshold or periodic triggering), the system performs calibration calculations. Based on this historical dataset, the system uses mathematical methods such as regression analysis to solve for an optimized risk sensitivity coefficient that minimizes the overall prediction bias of the model. Specifically, calibration aims to minimize the risk enhancement model (using...). To reduce the discrepancy between the predicted and actual consumption, or to narrow the difference between the predicted and actual consumption and the traditional benchmark model (based on...) The difference between predictions. The calibration formula is as follows:

[0202] ;

[0203] in: This indicates the number of historical sampling tasks used for calibration.

[0204] Indicates the first The weight of each historical sampling task. The weight can be set according to the size, complexity, or importance of the task.

[0205] Indicates the first The actual consumption rate of buffer time consumed by each historical sampling task.

[0206] Indicates the first The expected buffer time consumption rate for each historical sampling task. This expected consumption rate is estimated based on a traditional critical chain buffer time management model without incorporating risk adjustment (i.e., using the initial buffer time). The theoretically expected consumption level during management.

[0207] The system calculates the calibrated risk sensitivity coefficient by performing the above calculations. This calibration process can be set to run automatically on a regular schedule, or triggered after the system has accumulated a certain amount of new sampling task execution data. Updated This will be applied to the generation of execution plans for subsequent new sampling tasks, thereby enabling the risk-enhanced buffer duration calculation model to have self-learning and continuous optimization capabilities, and continuously improve the accuracy of planning.

[0208] Through the closed-loop learning and optimization in step S6, this application transforms the food safety sampling inspection process management system from a static tool into an intelligent decision support system that can grow alongside the organization and become increasingly accurate, achieving a leap from single-task management to continuous capability improvement.

[0209] This embodiment also discloses a food safety sampling inspection process management system, including:

[0210] The task receiving module is used to receive sampling inspection tasks, which include at least two process steps.

[0211] The risk analysis module is used to determine the potential failure modes and their risk values ​​corresponding to the process steps based on a preset PFMEA knowledge base.

[0212] The plan generation module is used to generate an execution plan through the critical chain based on the risk value, wherein the execution plan includes a buffer duration, the size of which is dynamically calculated and determined based on the risk value;

[0213] The monitoring and early warning module is used to monitor the progress and calculate the buffer time consumption rate during the execution of the execution plan; and to calculate a dynamic early warning threshold based on the risk value of the current execution process stage.

[0214] The response execution module is used to trigger an early warning and execute the corresponding countermeasures for the potential failure mode when the buffer duration consumption rate is greater than or equal to the early warning threshold.

[0215] This embodiment also discloses an electronic device, including:

[0216] One or more processors;

[0217] Memory;

[0218] One or more computer programs;

[0219] The one or more computer programs are stored in the memory and configured to be executed by the one or more processors, and the one or more computer programs are configured to perform the above-described food safety sampling inspection process management method.

[0220] This embodiment also discloses a computer-readable storage medium storing a computer program thereon, characterized in that the computer program, when executed by a processor, implements the above-mentioned food safety sampling inspection process management method.

[0221] Specific embodiments have been used to illustrate the principles and implementation methods of this invention. The descriptions of the embodiments above are only for the purpose of helping to understand the method and core ideas of this invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this invention. Therefore, the content of this specification should not be construed as a limitation of this invention.

Claims

1. A method for managing the food safety sampling inspection process, characterized in that, include: Receive a random inspection task, which includes at least two process steps; Based on a pre-set PFMEA knowledge base, the potential failure modes and their risk values ​​corresponding to the process steps are determined. Based on the risk value, an execution plan is generated through the critical chain, wherein the execution plan includes a buffer duration, the size of which is dynamically calculated and determined based on the risk value; During the execution of the execution plan, the progress is monitored and the buffer time consumption rate is calculated; based on the risk values ​​of all currently executed process stages, a dynamic early warning threshold is calculated. When the buffer duration consumption rate is greater than or equal to the warning threshold, a warning is triggered and the corresponding countermeasures for the potential failure mode are executed.

2. The food safety sampling inspection process management method according to claim 1, characterized in that, The steps to generate an execution plan include: Identify the logical relationships and resource constraints among the various process stages in the sampling task; Based on the aforementioned logical relationships and resource constraints, the critical chain is determined; For each process step in the critical chain, determine its pessimistic and aggressive projected project durations; Based on the pessimistic duration, aggressive duration, and risk value of all process links in the critical chain, the buffer time is calculated. The output contains the critical chain and buffer duration.

3. The food safety sampling inspection process management method according to claim 2, characterized in that, The buffer duration is calculated based on the pessimistic duration, aggressive duration, and risk value of all process stages on the critical chain, including: For each process step on the critical chain, calculate the difference between its pessimistic and aggressive durations, and use this as the first safe time for that step. For each process step, an adjustment factor is calculated based on its corresponding risk value and a preset risk sensitivity coefficient. Multiply the first safety time of each process step by the corresponding adjustment factor to obtain the second safety time of that step; The buffer duration is calculated based on the second safety time of all process steps on the critical chain.

4. The food safety sampling inspection process management method according to claim 3, characterized in that, The risk sensitivity coefficient is determined through the following steps: Based on the type of the sampling task, a preset coefficient value is used as the risk sensitivity coefficient; The types of sampling inspection tasks include routine supervision sampling inspection, special sampling inspection and emergency sampling inspection, and the coefficient values ​​of the three are pre-configured to increase sequentially.

5. The food safety sampling inspection process management method according to claim 3 or 4, characterized in that, The risk sensitivity coefficient is calibrated through the following steps: Obtain actual execution data for historical spot check tasks; Based on the comparison between the actual execution data and the historical sampling task plan data, the optimized risk sensitivity coefficient is calculated through regression analysis.

6. The food safety sampling inspection process management method according to claim 1, characterized in that, The monitoring progress and calculation of buffer time consumption rate include: Collect real-time progress data for each stage of the process; Based on the real-time progress data, the amount of buffer time consumed is determined, and the buffer time consumption rate is calculated.

7. The food safety sampling inspection process management method according to claim 1, characterized in that, The dynamic early warning threshold is calculated based on the risk values ​​of all currently executed process stages, including: The risk index is calculated based on the set of risk values ​​for all currently executing process stages. Based on the preset mapping relationship, the early warning threshold corresponding to the risk index is determined.

8. The food safety sampling inspection process management method according to claim 1, characterized in that, The method further includes an optimization step, which includes at least one of the following: After the sampling inspection task is completed, the frequency score of the risk value used to calculate the potential failure mode in the PFMEA knowledge base is updated according to the actual failure situation. Based on the actual buffer duration consumption rate, the parameters of the buffer duration calculation model are calibrated, including the risk sensitivity coefficient.

9. The food safety sampling inspection process management method according to claim 8, characterized in that, The process of updating the frequency score in the PFMEA knowledge base used to calculate the risk value of potential failure modes based on actual failure scenarios includes: Record the number of times each potential failure mode occurs during the execution of the sampling task; The frequency score of the failure mode is adjusted based on the number of occurrences after statistical analysis.

10. The food safety sampling inspection process management method according to claim 1, characterized in that, The triggering of the early warning and the execution of the corresponding response measures for the potential failure mode include resource scheduling steps: When an alert is triggered, it indicates that multiple process stages are competing for the same resource; According to a preset priority rule, the multiple process steps are sorted, wherein process steps belonging to the critical chain have a higher priority than non-critical chain process steps, and under the same conditions, process steps with higher risk values ​​have a higher priority. Based on the sorting results, the resources are allocated to the process steps with the highest priority.

11. A food safety sampling inspection process management system, characterized in that, include: The task receiving module is used to receive sampling inspection tasks, which include at least two process steps. The risk analysis module is used to determine the potential failure modes and their risk values ​​corresponding to the process steps based on a preset PFMEA knowledge base. The plan generation module is used to generate an execution plan through the critical chain based on the risk value, wherein the execution plan includes a buffer duration, the size of which is dynamically calculated and determined based on the risk value; The monitoring and early warning module is used to monitor the progress and calculate the buffer time consumption rate during the execution of the execution plan; and to calculate a dynamic early warning threshold based on the risk value of the current execution process stage. The response execution module is used to trigger an early warning and execute the corresponding countermeasures for the potential failure mode when the buffer duration consumption rate is greater than or equal to the early warning threshold.

Images