Critical process parameter analysis method, critical process parameter analysis device, and critical process parameter analysis program

Abstract

The present invention addresses the problem of identifying critical process parameter candidates in a manufacturing process for manufacturing a product. A critical process parameter analysis method according to one aspect of the present invention includes: variation evaluation steps (S2, S3) for evaluating, for each operation with respect to a subject system subjected to a plurality of operations included in a manufacturing process of a product, the magnitude of variation in system input inputted from outside the subject system on the basis of operation parameters set in order to perform the respective operations, and the magnitude of variation in system output outputted from the subject system in accordance with the respective operations; and a critical process parameter candidate identification step (S4) for identifying critical process parameter candidates on the basis of the evaluation results obtained by evaluating the magnitude of variation in the system input and the magnitude of variation in the system output for each operation.

Description

Critical process parameter analysis method, critical process parameter analysis device, and critical process parameter analysis program

[0001] The present invention relates to a method for analyzing critical process parameters, which identifies candidates for critical process parameters that have a high degree of influence on key indicators that reflect important quality characteristics of a product in a manufacturing process of the product.

[0002] In order to establish a manufacturing process that can consistently supply products with the intended functions, it is common to identify critical process parameters that affect the product's critical quality characteristics. For example, Patent Literature 1 discloses a method for manufacturing medical products using a quality by design approach. The quality by design approach sets a target product quality profile and critical quality characteristics of the final product, identifies critical process parameters that affect the performance and quality of the final product based on risk assessment and multivariate experiments, and determines the ranges of the critical quality characteristics and critical process parameters required to ensure the target product quality profile.

[0003] International Publication No. 2019-044559

[0004] Here, even when the operations in each process of manufacturing a product are performed by, for example, a device (machine), the input values ​​of the inputs performed by the device cannot be completely controlled, and fluctuations may exist in the input values. If such fluctuations, i.e., instability in the input values, exist, there is a possibility that they may affect the final product, but the technology described in Patent Document 1 does not take such fluctuations into consideration.

[0005] Furthermore, in the manufacture of cell preparations, polymers, etc., outliers may occur in the properties of the target substance after the reaction, the causes of which are difficult to identify, and it is necessary to extract critical process parameters while taking into account the instability of the target substance's properties after the reaction.

[0006] An object of one aspect of the present invention is to provide a critical process parameter analysis method and a critical process parameter analysis device for identifying candidates for critical process parameters in a manufacturing process for manufacturing a product.

[0007] In order to solve the above-mentioned problems, a critical process parameter analysis method according to one aspect of the present invention includes: a variation evaluation step of evaluating, for each of a plurality of operations included in a product manufacturing process, a target system that is the target of the operations and includes a response field that shows a response corresponding to each of the operations, the magnitude of variation of a system input that is input from outside the target system based on operation parameters that are set for performing each of the operations, and the magnitude of variation of a system output that is output from the target system in response to each of the operations; and a critical process parameter candidate identification step of identifying critical process parameter candidates from the operation parameters based on the evaluation results of the magnitude of variation of the system input and the magnitude of variation of the system output for each of the operations.

[0008] In order to solve the above problems, a critical process parameter analysis device according to one aspect of the present invention comprises: a variation evaluation unit that evaluates, for each of a plurality of operations included in a product manufacturing process, a response field that indicates a response corresponding to each of the operations, the magnitude of variation in a system input input from outside the target system based on operation parameters set for performing each of the operations, and the magnitude of variation in a system output output from the target system in response to each of the operations; and a critical process parameter candidate identification unit that identifies critical process parameter candidates from among the operation parameters based on an evaluation result obtained by evaluating, for each of the operations, the magnitude of variation in the system input and the magnitude of variation in the system output.

[0009] According to one aspect of the present invention, it is possible to identify candidates for critical process parameters in a manufacturing process for manufacturing a product.

[0010] FIG. 1 is a flowchart showing an example of processing of a critical process parameter analysis method according to embodiment 1 of the present invention. FIG. 2 is a diagram showing an example of a manufacturing process for manufacturing a cell preparation according to embodiment 2 of the present invention. FIG. 3 is a diagram for explaining control difficulty evaluation for each operation and task identification according to embodiment 2 of the present invention. FIG. 4 is a diagram for explaining control difficulty evaluation and response variability evaluation for each task according to embodiment 2 of the present invention. FIG. 5 is a table summarizing task difficulty for each sub-step according to embodiment 2 of the present invention. FIG. 6 is a table summarizing response variability for each sub-step according to embodiment 2 of the present invention. FIG. 7 is a table showing variability indexes for each sub-step according to embodiment 2 of the present invention. FIG. 8 is a block diagram showing the configuration of a main part of a critical process parameter analysis device according to embodiment 3 of the present invention. FIG. 9 is a conceptual diagram showing an example of a situation in which the critical process parameter analysis device is applied.

[0011] [Embodiment 1] Hereinafter, one embodiment of the present invention will be described in detail. In this embodiment, a method for analyzing critical process parameters for identifying candidates of critical process parameters that affect important quality characteristics of a product in a manufacturing process of the product will be described.

[0012] 1 is a flowchart showing an example of the processing of the critical process parameter analysis method according to this embodiment. As shown in FIG. 1, the extraction method according to this embodiment includes an operation identification step S1, a control difficulty evaluation step S2 (variation evaluation step), a response variability evaluation step S3 (variation evaluation step), and a critical process parameter candidate identification step S4.

[0013] The action identification step S1 is a step of identifying, for each step, a plurality of actions included in each step of the manufacturing process. Specifically, in the action identification step S1, for example, if the action performed in each step is an action to be performed by an apparatus to execute the step, each action performed by the apparatus may be identified as an action in the step. Alternatively, in the action identification step S1, for example, if the action performed in each step is an action performed by a person, each action performed by the person may be identified as an action in the step. In the action identification step S1, it is preferable to identify the action as finely as possible, for example, "opening a door" or "grabbing a container."

[0014] The control difficulty evaluation step S2 is a step of evaluating the magnitude of fluctuations in a system input from outside the target system, which is a system that is the target of multiple operations included in a product manufacturing process and includes a response field that exhibits a response corresponding to each of the operations, based on operation parameters set for performing each of the operations. In this specification, a "response field" refers to a field that exhibits a response corresponding to each of the operations included in the product manufacturing process. For example, in the case of the production of a cellular preparation, the response field is a field in which cells react and may include cells, liquids, etc. Furthermore, in the case of the production of a chemical product that includes a polymerization reaction process, the response field is a field in which chemical substances react and may include solids, liquids, etc. In this specification, a "target system" refers to a system that is the target of multiple operations included in a product manufacturing process and includes the response field. For example, in the case of the production of a cellular preparation, the target system may be a culture bed, tank, test tube, etc. that contains cells and liquids. The fluctuations in the system input may be fluctuations in the operation output actually output to the target system.

[0015] In the control difficulty evaluation step S2, the control difficulty of each motion included in each process may be evaluated. In this case, for each motion identified in the motion identification step S1, the control difficulty is evaluated based on the magnitude of the difference between the control target of the motion parameters set for each motion and the actual control status of the motion parameters for each motion.

[0016] Specifically, first, a control target value is set for each operation. The control target value is set for an operation parameter that can affect a key indicator that reflects an important quality characteristic of the product. The operation parameter is an operation parameter that affects the environment of the response field that is the target of the operation in each process. Examples of the operation parameter include pH, oxygen concentration, and temperature. The operation parameter may also include the duration of the operation. Hereinafter, the duration of the operation will also be referred to as the operation time. The operation parameter may also include a mechanical parameter of the operation. For example, if the operation is "grabbing a container," the mechanical parameter of the operation may be the speed at which the container is grasped. The control target of the operation parameter is set for each operation.

[0017] Next, an index representing the magnitude of the difference between the set control target value and the control situation when each of the multiple operations is actually executed (hereinafter referred to as a first difference) is calculated for each operation. Specifically, first, for each operation parameter, a range of controllable values ​​when the operation is executed is determined. Next, for each operation parameter, an index representing the difference in magnitude of the difference between the set control target value and the range of controllable values ​​when the operation is executed (hereinafter referred to as a controllable range) is calculated. Here, the controllable range is the range of values ​​that can actually be output when the operation is executed with the set control target value. Hereinafter, the index representing the difference in magnitude of the difference between the control target value and the controllable range for each operation is referred to as a control difficulty index. The control difficulty index is an index calculated based on the magnitude of the first difference.

[0018] The control difficulty index may be calculated, for example, as follows. First, the value that is included in the controllable range and that has the largest difference from the control target value is identified, and a value (hereinafter referred to as the maximum difference value) is calculated by subtracting the control target value from that value. Next, the control target value and the maximum difference value are each expressed as a power, and the control difficulty index is calculated by subtracting the exponent of the control target value from the exponent of the maximum difference value. As an example, a case will be described in which the control target for the motion parameter "velocity of grasping the container" of the motion "grabbing the container" is set to 0.1 m / s and the controllable range is 0.09 to 0.12 m / s. In this case, the maximum difference value is 0.12 - 0.1 = 0.02 m / s. When the control target value and the maximum difference value are expressed as a power of, for example, 10, the exponent is 1 x 10 －１ m / s, 2 x 10 －２ Since the control difficulty index is m / s, the control difficulty index is calculated as "-1" obtained by subtracting the exponent of the control target value, "-1", from the exponent of the maximum difference value, "-2". In the above explanation, the control target value and the maximum difference value are expressed as exponents with a base of 10, but the control difficulty index may also be calculated by expressing them as exponents with a base other than 10.

[0019] As another example, the control difficulty index may be calculated by dividing the maximum difference value by the control target value. For example, in the case of the "velocity of grasping the container" of the above-mentioned action of "grabbing the container," the control difficulty index may be calculated as 0.2, which is obtained by dividing the maximum difference value of "0.02 m / s" by the control target value of "0.1 m / s."

[0020] In the control difficulty evaluation step S2, the control difficulty is evaluated for each operation parameter using the calculated control difficulty index. Specifically, the operation parameter with the largest control difficulty index may be identified, and the control difficulty index of that operation parameter may be identified as the control difficulty index indicating the control difficulty of that operation. Alternatively, the operation parameters may be ranked in descending order of the control difficulty index.

[0021] As described above, in the control difficulty evaluation step S2, the control difficulty is evaluated for each operation parameter based on the magnitude of the first difference between the control target value set when performing each of the plurality of operations and the actual control situation. Then, the control difficulty for the plurality of operations is used to evaluate the control difficulty of the process including the plurality of operations. The control difficulty evaluation is performed for each process.

[0022] In the above example, the control target value is set to a single value. In one aspect of the present invention, when the control target value is set to a range of values, the control difficulty index may be calculated as an index indicating the difference between the scale of the range of the control target value and the scale of the controllable range when the operation is executed.

[0023] The response variability evaluation step S3 is a step of evaluating the magnitude of fluctuation in the system output output from the target system in response to each of the actions. Specifically, the response variability evaluation step S3 evaluates the response variability, which indicates the magnitude of fluctuation in the output from the target system, for each of the multiple actions. In this case, the response variability based on the fluctuation in the system output for each of the actions identified in the action identification step S1 is evaluated for each of the multiple actions. Specifically, an index indicating the magnitude of the difference (hereinafter also referred to as the second difference) between the action (input) to the target system (response field) and the response output from the target system (response field) is calculated. Hereinafter, the index indicating the magnitude of the second difference is referred to as the response variability index. For example, when the present invention is applied to a manufacturing process for producing a cell preparation, the time of the cell response can be used as an index for calculating the response variability index.

[0024] The response variability index may be calculated by calculating an exponent when the action input (action input value) to the target system and the response output value from the target system are expressed as a power, and subtracting the exponent of the response output value from the target system from the exponent of the action input. For example, when the present invention is applied to a manufacturing process for producing a cellular preparation, if "death" is set as the cell response output due to the action (input) to the target system, the time required to perform the action (hereinafter referred to as the action time) is 10 seconds, and the time until the cell shows a death phenotype in response to the action input (i.e., apoptosis occurs) (hereinafter referred to as the cell response time) is 1 hour (i.e., 3600 seconds). In this case, when the action time and the cell response time are expressed as a power of base 10, they are each 1 x 10 １ seconds, 3.6×10 ３ Therefore, the response variability index is calculated as "-2" obtained by subtracting the cellular response time exponent "3" from the operation time exponent "1." In the above explanation, the operation time and cellular response time are expressed as exponents with a base of 10, but the response variability index may also be calculated by expressing them as exponents with a base other than 10.

[0025] In the response variability evaluation step S3, the response variability is evaluated for each process using the calculated response variability index. Specifically, the action with the highest response variability index among the actions included in each process may be identified, and the response variability index of that action may be identified as the response variability index indicating the response variability of that process. Alternatively, the actions may be ranked in descending order of response variability index.

[0026] As described above, in the response variability evaluation step S3, the response variability is evaluated for each of the plurality of movements based on the second difference between the movements. Then, the response variability for the plurality of movements is used to evaluate the response variability of the process including the plurality of movements. The response variability evaluation is performed for each process.

[0027] The critical process parameter candidate identification step S4 is a step of identifying critical process parameter candidates from among operation parameters included in each operation of the product manufacturing process based on the control difficulty and response variability. Hereinafter, each of the multiple processes included in the product manufacturing process is referred to as a subprocess. In the critical process parameter candidate identification step S4, a variability index indicating the degree of influence on the critical index is calculated for each subprocess using the control difficulty index of each subprocess calculated in the control difficulty evaluation step S2 and the response variability index of each subprocess calculated in the response variability evaluation step S3. The calculated variability index is an index indicating the variability of the product manufacturing system. The method for calculating the variability index is not particularly limited. For example, the variability index may be calculated by linearly adding the control difficulty index and the response variability index with a predetermined weighting. As a specific example, for each operation parameter, the variability index may be calculated by linearly adding the control difficulty index calculated in the control difficulty evaluation step S2 and the response variability index calculated in the response variability evaluation step S3 with a 1:1 weighting. In the critical process parameter candidate identification step S4, the operational parameters whose calculated variability index is equal to or greater than a predetermined value are identified as critical process parameter candidates.

[0028] As described above, the critical process parameter analysis method of this embodiment calculates a system variability index based on a control difficulty index indicating the stability / instability of the input value of each operation and a response variability index indicating the instability of the output from the target system (response field) due to the input of each operation. Then, based on the calculated variability index, operation parameters that have a high degree of influence on the critical index are automatically identified as critical process parameter candidates. With the above configuration, it is possible to automatically identify critical process parameter candidates while reflecting the stability / instability of the input value of each operation and the instability of the output from the target system (response field) due to the input of each operation.

[0029] In one aspect of the critical process parameter analysis method of the present invention, when the method is applied to a manufacturing process in which the operation in each sub-process is predetermined, the operation specifying step S1 can be omitted.

[0030] In this embodiment, the control difficulty and response variability are calculated for each operation, but the critical process parameter analysis method of the present invention is not limited to this. For example, a series of operations in which the target system shows the same response may be defined as an operation group, and the control difficulty and response variability may be calculated for each operation in the operation group. This makes it possible to reduce the number of targets for identifying critical process parameter candidates.

[0031] [Embodiment 2] Another embodiment of the present invention will be described below. For ease of explanation, components having the same functions as those described in the above embodiment will be denoted by the same reference numerals, and their description will not be repeated. In this embodiment, a method for analyzing critical process parameters in a manufacturing process for producing a cell preparation will be described as a specific example of the method for analyzing critical process parameters described in Embodiment 1.

[0032] Fig. 2 is a diagram showing an example of a manufacturing process for producing a cell preparation. In this embodiment, a method for analyzing critical process parameters in a manufacturing process for a cell preparation, which is composed of the sub-processes shown in Fig. 2 , namely, a thawing / seeding process, an amplification / culture process (cell culture process), a medium replacement process, a recovery / passaging process, a detachment process, a dispensing process, and a freezing process, will be described, but other processes may also be included.

[0033] <Action Identification Step> In the critical process parameter analysis method of this embodiment, first, multiple actions included in each partial process are identified for each partial process (action identification step). For example, in the recovery / passage process, actions such as "move to the incubator," "grab the outer door of the incubator," and "open the outer door of the incubator" are identified. <Control Difficulty Evaluation Step for Each Action> Next, the control difficulty of each action included in each partial process is evaluated. Note that the following explanation will be given using the recovery / passage process as an example. Figure 3 is a diagram for explaining the control difficulty evaluation for each action in the recovery / passage process and the identification of the work described below. Note that Figure 3 shows some of the actions in the recovery / passage process.

[0034] In this step, first, operation parameters (process parameters) are set for each operation identified in the operation identification step. Operation parameters are parameters that serve as inputs to the response field (i.e., cells) and can affect important indicators (e.g., cell viability) that reflect important quality characteristics of the cell preparation to be manufactured. In the example shown in Figure 3, "pH," "O2 (oxygen concentration)," "temperature," "liquid phase," "operation time," and "operation" are set as operation parameters.

[0035] Next, a control target value is set for each operation parameter for each operation. In the example shown in Fig. 3, a control target value is set for each operation parameter: "pH", "O2 (oxygen concentration)", "temperature", "liquid phase", "operation time", and "operation".

[0036] Next, the exponent value when each set control target value is expressed as a power of base 10 is calculated. In this embodiment, the exponent value when expressed as a power of base 10 is called a scale. For example, suppose that the control target value of the operation parameter "pH" of the operation "move from incubator to front of safety cabinet" (hereinafter referred to as operation 1) is set to "7". In this case, when "7" is expressed as a power of base 10, it is calculated as "7 x 10 ０ ", the scale of the control target value of the operation parameter "pH" in operation 1 is calculated to be "0." In this case, the calculated scale of the control target value corresponds to "0" in the "Actual pH Scale" column in the table shown in FIG. 3.

[0037] Next, the exponent value is calculated when the maximum difference between the value included in the controllable range of values ​​of each operation parameter when the operation is executed (i.e., the controllable range) and the control target value is expressed as a power of base 10. In other words, the scale of the maximum difference value is calculated. For example, for the operation parameter "pH" of operation 1, if the controllable pH range is "6.9 to 7.2", the maximum difference value is "0.2 (= 7.2 - 7)", and when "0.2" is expressed as a power of base 10, it is "2 x 10 -１Therefore, the scale of the maximum difference value is calculated as "-1." The calculated scale of the maximum difference value corresponds to "-1" in the "Control pH Scale" column in the table shown in FIG.

[0038] Next, a value obtained by subtracting the scale of the control target value from the scale of the maximum difference value is calculated. For example, for the operation parameter "pH" of operation 1, the scale of the maximum difference value is "-1" and the scale of the control target value is "0", so "-1 (=-1-0)" is calculated. The calculated value corresponds to "-1" in the "Index difference (controlled pH scale - actual pH)" column in the table shown in FIG. 3. The value obtained by subtracting the scale of the control target value from the scale of the maximum difference value is an index indicating the difference between the magnitude of the control target value of the operation parameter and the magnitude of the controllable range of the operation parameter when the operation is executed, i.e., an index of the difficulty of control of the operation parameter.

[0039] <Task Identification Step> Next, in each sub-step, a series of actions that cause the target system to show the same response is identified as a single action group. In the following description, this action group is referred to as a "task." In this step, first, for each action, the input to the response field (hereinafter referred to as a response field input) is identified. For example, in the case of action 1, an inertial force is input to the response field due to mechanical action, and metabolic activity changes with changes in the environment (i.e., pH, oxygen concentration, temperature, etc.). Therefore, "inertial force" and "metabolic activity" are identified as response field inputs. In this case, the identified response field input corresponds to "inertial force (N)" in the "Cellular response input variable (cell characteristic)" column and "metabolic activity" in the "Cellular response input (environment)" column in the table shown in FIG. 3 .

[0040] Next, the response of the response field due to the identified response field input is identified. For example, in the case of operation 1, the response exhibited by the cell due to "inertial force" is identified as the "natural death switch," and the response exhibited by the cell due to "environment" is identified as "metabolic activity." In this case, the identified responses correspond to "natural death switch" in the "Major cellular response (cell characteristic)" column and "metabolic activity" in the "Major cellular response (environment)" column in the table shown in FIG. 3.

[0041] Next, the response output of the response field is identified for each action. For example, in the case of action 1, the cell dies when the natural death switch is turned on by inertial force, so "death" is identified as the response field output. Also, because the metabolic activity of the cell decreases due to changes in the environment, "decreased metabolic activity" is identified as the response output. In this case, the identified response field output corresponds to "death" in the "Cell response output (cell characteristics)" column and "decreased metabolic activity" in the "Cell response output (environment)" column in the table shown in FIG. 3.

[0042] Next, for each action, the exponent value is calculated when the time from inputting the action until the response field (cell) outputs a response (hereinafter referred to as the response time) is expressed as a power of base 10. In other words, the scale of the response time is calculated. Note that when there are multiple responses from the response field (cell), the scale of the response time is calculated using the response time of the response with the shorter response time. For example, for action 1, it is assumed that the time until the response with the shorter response time, either death or decreased metabolic activity, is output is 3600 seconds. In this case, when "3600" is expressed as a power of 10, it is "3.6 x 10 ３ ", the response time scale is calculated as "3." In this case, the calculated response time scale corresponds to "3" in the "Cell response time scale" column in the table shown in FIG. 3.

[0043] Next, a series of actions that cause the target system to respond identically are grouped together as a single action group (i.e., task). Specifically, a series of actions that have the same cellular response output and cellular response time scale are identified as a single task. For example, in the example shown in FIG. 3 , the action of "moving from the incubator to the front of the safety cabinet" (i.e., action 1) and the action of "introducing into the safety cabinet" (hereinafter referred to as action 2) are grouped together as the same action because the cellular response outputs from the target system are both "death" and "decreased metabolic activity" and the cellular response time scale is both "3." Furthermore, in this embodiment, actions that are performed in environments with different cleanliness levels are grouped together as different tasks. For example, actions 1 and 2 are performed outside the safety cabinet, while the action of "opening the container lid" shown in FIG. 3 (hereinafter referred to as action 3) is performed inside the safety cabinet, so the cleanliness of the environment in which action 3 is performed is different from that of actions 1 and 2. Therefore, although the cellular response output from the target system and the cellular response time scale are the same as those of operations 1 and 2, operation 3 is not grouped as the same task as operations 1 and 2.

[0044] <Step of Evaluating the Difficulty of Controlling Each Operation> Next, the difficulty of controlling each operating parameter is evaluated for each operation. Figure 4 is a diagram for explaining the evaluation of the difficulty of controlling each operation in the recovery and passaging process and the evaluation of response variability, which will be described later. Note that Figure 4 shows some of the operations in the recovery and passaging process.

[0045] In this step, first, for each task, a control target value is set for each operation parameter, as shown in FIG. 4 . Each operation parameter is an operation parameter set when evaluating the control difficulty of each operation. At this time, if the control target values ​​differ between multiple operations included in one task, the control target value set for the operation with the largest variation in operation parameter is set as the control target value for that task. Next, an exponent value is calculated when each set control target value is expressed as a power of 10. In other words, a scale is calculated for each set control target value. For example, the scale of the control target value calculated for "pH" corresponds to the value in the "Actual pH Scale" column in the table shown in FIG. 4 .

[0046] Next, an index indicating the difference between the magnitude of the control target value of the operation parameter and the magnitude of the controllable range of each operation parameter when performing the task (i.e., the controllable range), i.e., a control difficulty index, is calculated for each operation parameter. Specifically, first, the exponent value when the maximum difference between the control target value and the controllable range of the operation parameter is expressed as a power of 10, i.e., the scale of the maximum difference value, is calculated. For example, the scale of the maximum difference value calculated for "pH" corresponds to the value in the "Control pH Scale" column in the table shown in FIG. 4.

[0047] Next, a control difficulty index is calculated by subtracting the scale of the control target value from the scale of the maximum difference value. For example, the value calculated for "pH" corresponds to the value in the "Index Difference (Control pH Scale - Actual pH)" column in the table shown in Figure 4.

[0048] Next, the control difficulty of each sub-process is evaluated for each task. This will be explained with reference to FIG. 5. FIG. 5 is a table summarizing the task difficulty of each sub-process using the control difficulty index calculated by the above method. As shown in FIG. 5, for example, the "thawing and seeding process" is summarized into seven tasks, and the maximum value of the control difficulty index for the operation parameters of each task is evaluated as "0" for one task, "-1" for five tasks, and "-2" for one task. The task with the highest control difficulty is evaluated as "warming the frozen sample."

[0049] <Response variability evaluation step> Next, the response variability index for each task is evaluated. Specifically, first, the actual task time scale is identified for each task. The actual movement time scale for each task is calculated as an exponent value when the total movement time of a series of movements included in each task is expressed as a power of base 10. The calculated value corresponds to the value in the "Actual movement time scale" column in the table shown in FIG. 4.

[0050] Next, the response variability index for the task is calculated by subtracting the cellular response time scale for the task from the calculated actual action time scale. The calculated response variability index corresponds to the value in the "Actual action time scale - Cellular response time scale" column in the table shown in Figure 4.

[0051] Next, the response variability is evaluated for each task. This will be explained with reference to FIG. 6. FIG. 6 is a table summarizing the response variability of each sub-step using the response variability index calculated by the above method. As shown in FIG. 6, for example, for the "thawing and seeding step," one task has the maximum response variability index of "1," four tasks have the maximum value of "0," and two tasks have the maximum value of "-1." The task with the highest response variability is evaluated as "warming the frozen sample."

[0052] <Critical process parameter candidate identification step> Next, critical process parameter candidates in the manufacturing process for producing a cell preparation are identified. Below, two patterns will be described: a method of calculating a variability index for each operation to identify critical process parameter candidates, and a method of calculating a variability index for each action to identify critical process parameter candidates.

[0053] In the method for calculating a variability index for each task and identifying candidate critical process parameters, first, for each partial process, a numerical value is calculated for each operation parameter by linearly adding, with a predetermined weighting, the controllability index calculated in the controllability evaluation step for each task and the response variability index calculated in the response variability evaluation step. In this embodiment, an example is described in which the variability index is calculated by linearly adding, with a weighting of 1:1, the controllability index and the response variability index. For example, for the task "introducing container into safety cabinet" in the "recovery / passaging process," the highest controllability indexes among the operation parameters are "-1" for "pH," "task time," and "operation," and the response variability index is "-2." Therefore, the variability index is calculated as 1 x (-1) + 1 x (-2) = -3.

[0054] Next, operational parameters whose calculated variability index is equal to or greater than a predetermined value are identified as candidate critical process parameters. FIG. 7 is a table showing the variability index of each partial process. In this embodiment, operational parameters whose variability index is equal to or greater than 0 are identified as candidate critical process parameters. The example shown in FIG. 7 shows that, for the "recovery / passaging process," the "temperature," "operation time," and "operation speed" in the "addition of detachment solution" step, and the "number of cells seeded" in the "seeding" step, are identified as candidate critical process parameters.

[0055] In the method for calculating a variability index for each operation and identifying key process parameter candidates, for each partial process, a variability index is calculated for each operation parameter by linearly adding, with a predetermined weighting, the control difficulty index calculated in the control difficulty evaluation step for each operation and the response variability index calculated in the response variability evaluation step.Then, operation parameters whose calculated variability index is equal to or greater than a predetermined value are identified as key process parameter candidates.

[0056] In addition, when identifying key process parameter candidates by a method of calculating a variability index for each task and identifying key process parameter candidates, it is only necessary to calculate a controllability index for each task, and there is no need to calculate a controllability index for each action.

[0057] This embodiment also includes a method for teaching product manufacturing design using a guidebook or the like that includes the above-mentioned critical process parameter analysis method.

[0058] [Embodiment 3] Another embodiment of the present invention will be described below. The critical process parameter analysis methods described in embodiments 1 and 2 may be executed by one or more computers. In this embodiment, a critical process parameter analysis device that executes the critical process parameter analysis method described in embodiment 1 will be described. Figure 8 is a block diagram showing the configuration of the main parts of the critical process parameter analysis device 1 in this embodiment. Figure 9 is a conceptual diagram showing an example of a situation in which the critical process parameter analysis device 1 is applied.

[0059] The critical process parameter analysis apparatus 1 is an apparatus for identifying critical process parameter candidates from among operation parameters set for performing operations included in a product manufacturing process. As shown in Fig. 8, the critical process parameter analysis apparatus 1 includes a control unit 10, a storage unit 20 for storing various data used by the critical process parameter analysis apparatus 1, an input unit 31 for receiving input operations to the critical process parameter analysis apparatus 1, and an output unit 32 for displaying various information.

[0060] The control unit 10 controls each unit of the critical process parameter analysis apparatus 1. The control unit 10 includes an operation parameter information acquisition unit 11, a variation evaluation unit 12, a critical process parameter candidate identification unit 13, and an output control unit 14.

[0061] The operation parameter information acquisition unit 11 acquires information about the operation parameters. The information about the operation parameters includes the number of partial processes included in the product manufacturing process, the operations included in each partial process, the operation parameters of each operation, the control target values ​​set for the operation parameters in each operation, the controllable range of each operation parameter, the operation time of each operation, and the response time of the response field due to each operation. The operation parameter information acquisition unit 11 may acquire the information about the operation parameters by inputting the information about the operation parameters via the input unit 31. Alternatively, the operation parameter information acquisition unit 11 may acquire the information about the operation parameters from an information processing terminal 5 owned by a user via a network 9, as shown in FIG. 9 .

[0062] The operation parameter information acquisition unit 11 may acquire information predicted using a prediction means such as a computer simulation as at least a part of the information related to the operation parameters. In this case, the fluctuation evaluation unit 12, which will be described later, may evaluate the magnitude of fluctuation in a system input input from a target system including a response field to the target system and the magnitude of fluctuation in a system output output from the target system in response to the operation, based on information including virtual data obtained using the prediction means. The prediction using the prediction means may be performed by the operation parameter information acquisition unit 11 itself.

[0063] The fluctuation evaluation unit 12 evaluates, for each action, the magnitude of fluctuation in a system input input from the target system including a response field and the magnitude of fluctuation in a system output output from the target system in response to the action. The fluctuation evaluation unit 12 includes a control difficulty index calculation unit 12A and a response variability index calculation unit 12B.

[0064] The control difficulty index calculation unit 12A calculates a control difficulty index indicating the magnitude of fluctuation in the system input. Specifically, the control difficulty index calculation unit 12A uses the information acquired by the operation parameter information acquisition unit 11 to calculate, for each operation parameter of an operation included in a partial process, a control difficulty index indicating the difference in magnitude between a control target value set by a user and a controllable range when the operation is executed. The control difficulty index calculation method uses the method described in embodiment 1. The control difficulty index calculation unit 12A outputs the calculated control difficulty index to the key process parameter candidate identification unit 13.

[0065] The response variability index calculation unit 12B calculates a response variability index that indicates the magnitude of variation in the system output that is output from the target system in response to an operation. Specifically, the response variability index calculation unit 12B uses the information acquired by the operation parameter information acquisition unit 11 to calculate a response variability index that indicates the magnitude of the difference between the operation (input) to the target system and the response output from the target system (response field) for each operation included in a partial process. The response variability index calculation method uses the method described in embodiment 1. The response variability index calculation unit 12B outputs the calculated response variability index to the critical process parameter candidate identification unit 13.

[0066] The critical process parameter candidate identifying unit 13 identifies critical process parameter candidates from among the operation parameters included in the product manufacturing process using the evaluation results (i.e., the control difficulty index output from the control difficulty index calculation unit 12A and the response variability index output from the response variability index calculation unit 12B) obtained by evaluating the magnitude of fluctuation in the system input and the magnitude of fluctuation in the system output for each operation. Specifically, the critical process parameter candidate identifying unit 13 calculates a variability index for each operation by linearly adding the largest control difficulty index among the operation parameter control difficulty indexes and the response variability index of the operation with a predetermined weighting. Then, the critical process parameter candidate identifying unit 13 identifies critical process parameter candidates from among the operation parameters included in the product manufacturing process based on the calculated variability index. For example, the critical process parameter candidate identifying unit 13 may identify operation parameters whose variability index is equal to or greater than a predetermined value as critical process parameter candidates.

[0067] The output control unit 14 controls the output unit 32 to output the important process parameter candidates identified by the important process parameter candidate identifying unit 13 .

[0068] The variation evaluation unit 12 and the critical process parameter candidate identification unit 13 may perform the above operations using a critical process parameter candidate analysis application 21 equipped with the critical process parameter analysis program, which is stored in advance in the storage unit 20.

[0069] In the critical process parameter analysis apparatus 1 according to one aspect of the present invention, information including the identified operational parameters that are candidates for critical process parameters may be provided to an external terminal device via a communication network (for example, the Internet).

[0070] [Embodiment 4] Another embodiment of the present invention will be described below. The critical process parameter analysis methods described in embodiments 1 and 2 can also be used to change and / or improve critical process parameters. In this embodiment, as an example, a configuration will be described in which the critical process parameter analysis methods described in embodiments 1 and 2 are applied to change and / or improve candidate critical process parameters.

[0071] 7, "temperature" in the operation of "addition of detachment solution" in the "recovery and passaging step" is identified as a candidate critical process parameter. The reason for this is that the detachment solution used in the "recovery and passaging step" (hereinafter referred to as detachment solution A) is a detachment solution that exerts its desired effect by being added at room temperature and then transferred to a temperature condition of 37°C, and therefore it is necessary to change the temperature when using the detachment solution.

[0072] Assume that there is a stripping solution that can be used in a certain environment, specifically, stripping solution B, which is a stripping solution that is added at room temperature and exhibits the desired effect at room temperature. In this case, it is possible to change the stripping solution used in the above-mentioned "adding stripping solution" operation from stripping solution A to stripping solution B. Therefore, the critical process parameter candidates when the stripping solution is changed from stripping solution A to stripping solution B are identified using the critical process parameter analysis method described in embodiments 1 and 2. Unlike the case where stripping solution A is used, the use of stripping solution B eliminates the need to change the temperature. Therefore, the controllability index for the "temperature" in the above-mentioned "adding stripping solution" operation is reduced, resulting in a smaller variability index. Therefore, it is possible to show that the "temperature" in the "adding stripping solution" operation in the "recovery / passaging step" is not a candidate critical process parameter. If the use of stripping solution B requires a procedure and / or operation that is not necessary when stripping solution A is used, it is possible to show whether or not a parameter related to the procedure and / or operation is identified as a candidate critical process parameter. This allows the manufacturing process designer to confirm changes in the candidate critical process parameters resulting from changing the stripper solution from stripper solution A to stripper solution B, and to change and / or improve the candidate critical process parameters in consideration of the confirmed results.

[0073] The above example describes a situation in which a change in the stripping solution in the "recovery and passaging process" is considered, but the present invention is not limited to this situation. For example, by mechanizing a previously manual operation and changing the operation speed to a constant value, it is possible to show that the "operation speed" is no longer a candidate for a critical process parameter. In this way, by applying the critical process parameter analysis method, it is possible to confirm how changes in each process will affect the candidate critical process parameters before actually making the changes.

[0074] As described above, when at least some elements in a manufacturing process are changed, the critical process parameter analysis method described in the first and second embodiments can be applied to both the cases before and after the change of the elements. Then, changes in the critical process parameters due to the change in the manufacturing process can be confirmed, and it can be determined whether or not to make the change. This allows the manufacturing process designer to confirm changes in the candidate critical process parameters due to the change in the manufacturing process before actually making the change in the manufacturing process. Therefore, the manufacturing process designer can change and / or improve the candidate critical process parameters in consideration of the confirmed results.

[0075] [Example of Implementation by Software] The functions of the critical process parameter analysis apparatus 1 (hereinafter referred to as the "apparatus") can be realized by a program that causes a computer to function as the apparatus, and a program that causes a computer to function as each control block of the apparatus (particularly, each unit included in the control unit 10).

[0076] In this case, the device includes a computer having at least one control device (e.g., a processor) and at least one storage device (e.g., a memory) as hardware for executing the program. The functions described in each of the above embodiments are realized by executing the program using the control device and storage device.

[0077] The program may be non-transitory and may be recorded on one or more computer-readable recording media. The recording media may or may not be included in the device. In the latter case, the program may be supplied to the device via any wired or wireless transmission medium.

[0078] Furthermore, some or all of the functions of the control blocks can be realized by logic circuits. For example, an integrated circuit in which a logic circuit that functions as each of the control blocks is formed is also included in the scope of the present invention. In addition, the functions of the control blocks can also be realized by, for example, a quantum computer.

[0079] The present invention is not limited to the above-described embodiments, and various modifications are possible within the scope of the claims. Embodiments obtained by appropriately combining the technical means disclosed in different embodiments are also included in the technical scope of the present invention.

[0080] [Summary] A critical process parameter analysis method according to a first aspect of the present invention includes the following steps: a variation evaluation step of evaluating, for each of a plurality of operations included in a product manufacturing process, a response field that indicates a response corresponding to each of the operations, the magnitude of variation in a system input that is input from outside the target system based on operation parameters that are set for performing each of the operations, and the magnitude of variation in a system output that is output from the target system in response to each of the operations; and a critical process parameter candidate identification step of identifying critical process parameter candidates from among the operation parameters based on the evaluation results of the magnitude of variation in the system input and the magnitude of variation in the system output for each of the operations.

[0081] A key process parameter analysis method according to a second aspect of the present invention is the method of the first aspect, wherein the fluctuation of the system input may be a fluctuation of an operational output that is actually output to the target system.

[0082] A key process parameter analysis method according to Aspect 3 of the present invention may be configured in the above-described Aspect 2, wherein the variation evaluation step evaluates the difficulty of control based on the magnitude of a difference between a control target value set for the operation parameter when performing each of the plurality of operations and an actual control state of the operation output as a magnitude of variation of the system input.

[0083] The critical process parameter analysis method according to Aspect 4 of the present invention may be configured in the above-mentioned Aspect 3 to calculate, in the variation evaluation step, a control difficulty index indicating the control difficulty for each of the operation parameters.

[0084] A key process parameter analysis method according to Aspect 5 of the present invention may be the method according to Aspect 4, wherein the control difficulty index is an index indicating a difference between the size of the range of the control target value and the size of the controllable range when the operation is performed.

[0085] A sixth aspect of the present invention provides the method for critical process parameter analysis in any one of the first to fifth aspects, wherein the fluctuation in the system output is a fluctuation in a response output from the target system.

[0086] A critical process parameter analysis method according to a seventh aspect of the present invention may be configured in any one of the first to sixth aspects to evaluate, for each of the operations, a response variability based on a degree to which the operation affects a key indicator that reflects an important quality attribute of the product, as a magnitude of variation in the system output.

[0087] The critical process parameter analysis method according to Aspect 8 of the present invention may be configured in the above-mentioned Aspect 7 to calculate, for each of the operations, a response variability index indicating the response variability in the variability evaluation step.

[0088] A critical process parameter analysis method according to a ninth aspect of the present invention may be the method according to the eighth aspect, wherein the response variability index is an index indicating the magnitude of a difference between the operational output and a response output of the target system due to an input to the target system.

[0089] A critical process parameter analysis method according to Aspect 10 of the present invention may be configured in the above-mentioned Aspect 8 or 9, wherein the critical process parameter candidate identifying step identifies the critical process parameter candidate based on a variability index calculated by linearly adding, with a predetermined weighting, a control difficulty index indicating the controllability and a response variability index indicating the response variability.

[0090] A critical process parameter analysis method according to an eleventh aspect of the present invention may be configured in any one of the first to tenth aspects, wherein the plurality of operations are operations to be executed by a predetermined device in order to execute the partial process.

[0091] A critical process parameter analysis method according to Aspect 12 of the present invention may be configured in any one of Aspects 1 to 11 above, wherein the product is a cell preparation, and the manufacturing process includes a cell culture process.

[0092] A critical process parameter analysis method according to Aspect 13 of the present invention may be configured in the above-mentioned Aspect 12, wherein the critical indicator includes a survival rate of the cells.

[0093] A critical process parameter analysis method according to Aspect 14 of the present invention may be configured in any one of Aspects 1 to 11 above, wherein the product is a chemical product, and the manufacturing process includes a polymerization reaction process.

[0094] A critical process parameter analysis device according to a fifteenth aspect of the present invention comprises: a variation evaluation unit that evaluates, for each of a plurality of operations included in a product manufacturing process, a response field that indicates a response corresponding to each of the operations, a magnitude of variation in a system input input from outside the target system based on operation parameters set for performing each of the operations, and a magnitude of variation in a system output output from the target system in response to each of the operations; and a critical process parameter candidate identification unit that identifies a critical process parameter candidate from among the operation parameters based on an evaluation result obtained by evaluating, for each of the operations, the magnitude of variation in the system input and the magnitude of variation in the system output.

[0095] A critical process parameter analysis program according to a sixteenth aspect of the present invention is a critical process parameter analysis program for causing a computer to function as the critical process parameter analysis device according to the fifteenth aspect, and is a critical process parameter analysis program for causing a computer to function as the variation evaluation unit and the critical process parameter candidate identification unit.

[0096] The education method according to aspect 17 of the present invention is an education method for product manufacturing design that utilizes a guidebook or the like that includes the critical process parameter analysis method according to any one of aspects 1 to 14 above.

[0097] REFERENCE SIGNS LIST 1 Critical process parameter analysis device 5 Information processing terminal 9 Network 10 Control unit 12 Variation evaluation unit 13 Critical process parameter candidate identification unit S1 Operation identification step S2 Control difficulty evaluation step (variation evaluation step) S3 Response variability evaluation step (variation evaluation step) S4 Critical process parameter candidate identification step

Claims

1. a fluctuation evaluation step of evaluating, for each of a plurality of operations included in a manufacturing process of a product, a magnitude of fluctuation in a system input input from outside the target system based on operation parameters set for performing each of the operations, the magnitude of fluctuation in a system output output from the target system in response to each of the operations, the target system being a system that is the target of a plurality of operations included in a manufacturing process of a product and including a response field that shows a response corresponding to each of the operations; and a critical process parameter candidate identifying step of identifying a critical process parameter candidate from the operation parameters based on an evaluation result obtained by evaluating the magnitude of the fluctuation of the system input and the magnitude of the fluctuation of the system output for each operation. Critical process parameter analysis method.

2. The fluctuation of the system input is a fluctuation of the operational output actually output to the target system. The method for analyzing critical process parameters according to claim 1.

3. In the fluctuation evaluation step, the difficulty of control based on the magnitude of a difference between a control target value set for the operation parameter when performing each of the plurality of operations and an actual control state of the operation output is evaluated as a magnitude of a fluctuation of the system input. The method for analyzing critical process parameters according to claim 2.

4. In the fluctuation evaluation step, a control difficulty index indicating the control difficulty is calculated for each of the operation parameters. The method for analyzing critical process parameters according to claim 3.

5. The control difficulty index is an index indicating a difference between the scale of the range of the control target value and the scale of the controllable range when the operation is executed. The method for analyzing critical process parameters according to claim 4.

6. The fluctuation of the system output is a fluctuation of the response output from the target system. The method for analyzing critical process parameters according to claim 1.

7. For each of the operations, a response variability is evaluated based on the degree to which the operation affects a key indicator reflecting an important quality characteristic of the product, as a magnitude of variation in the system output. The method for analyzing critical process parameters according to claim 1.

8. The fluctuation of the system input is a fluctuation of an operational output actually output to the target system, In the variability evaluation step, a response variability index indicating the response variability is calculated for each of the actions. The method for analyzing critical process parameters according to claim 7.

9. The response variability index is an index indicating the magnitude of the difference between the operation output and the response output of the target system due to the input to the target system. The method for analyzing critical process parameters according to claim 8.

10. In the critical process parameter candidate identification step, the critical process parameter candidate is identified based on a variability index calculated by linearly adding, with a predetermined weighting, a control difficulty index indicating control difficulty based on the magnitude of a difference between a control target value set for the operation parameter when performing each of the plurality of operations and an actual control status of the operation output, and a response variability index indicating the response variability. The method for analyzing critical process parameters according to claim 8.

11. The plurality of operations are operations to be executed by a predetermined device. The method for analyzing critical process parameters according to any one of claims 1 to 10.

12. the product is a cell preparation, The manufacturing process includes a cell culture process. The method for analyzing critical process parameters according to any one of claims 1 to 10.

13. the product is a cell preparation, the production process includes a cell culture process, The key indicators include cell viability; The method for analyzing critical process parameters according to claim 7.

14. the product is a chemical product, The production process includes a polymerization reaction process. The method for analyzing critical process parameters according to any one of claims 1 to 10.

15. a fluctuation evaluation unit that evaluates, for each of a plurality of operations included in a manufacturing process of a product, the magnitude of fluctuation in a system input input from outside the target system based on operation parameters set for performing each of the operations, and the magnitude of fluctuation in a system output output from the target system in response to each of the operations, for the target system including a response field that shows a response corresponding to each of the operations; and a critical process parameter candidate identifying unit that identifies a critical process parameter candidate from the operation parameters based on an evaluation result obtained by evaluating the magnitude of the fluctuation of the system input and the magnitude of the fluctuation of the system output for each operation. Critical process parameter analyzer.

16. A critical process parameter analysis program for causing a computer to function as the critical process parameter analysis device according to claim 15, the critical process parameter analysis program causing a computer to function as the variation evaluation unit and the critical process parameter candidate identification unit.