Method for filling pharmaceutical compositions

Abstract

A method is provided for reducing the amount of material used in a container. The method includes obtaining historical filling weight data showing the actual unit-by-unit filling weight volume for multiple containers. The method also includes generating a simulated filling weight distribution based on an analysis of the historical filling weight data. The method further includes randomly sampling the simulated filling weight distribution to select a subset of the simulated filling weights and performing a quality check on the subset of the simulated filling weights. The method further includes generating a corrected distribution of the simulated filling weights by removing at least the subset samples that failed the quality check from the simulated filling weight distribution. The method also includes converting the corrected distribution of the simulated filling weights into a distribution of filling volumes and calculating a distribution of deliverable volumes based on the distribution of filling volumes.

Description

【Technical Field】 【0001】 Cross - reference to Related Applications Priority is claimed to U.S. Provisional Patent Application No. 63 / 466,440, filed May 15, 2023, the entire contents of which are incorporated herein by reference. 【0002】 This application generally relates to filling a container with a predetermined amount of material, and more specifically to reducing the amount of excess material used when filling a container. 【Background Art】 【0003】 Container filling processes are often tightly controlled to dispense an accurate amount of material (e.g., fluid, formulation, etc.) into a container (e.g., syringe, cartridge, auto - injector, etc.). For example, a “prefilled” syringe often contains a certain amount of formulation determined based on specifications (e.g., deliverable volume of the formulation, deliverable amount of the formulation, etc.) indicated on the formulation label. The filling target of a related container filling process often depends on the type of material being dispensed into the container and the type of the container. 【0004】 As a specific example, the filling target for a syringe filling process often depends on the specified deliverable volume for the formulation being dispensed into the syringe during the filling process. The filling target is usually larger than the specified deliverable volume to ensure that an amount of drug sufficient for patient administration can be withdrawn. First, a “hold - up volume” is often added to the deliverable volume to account for the formulation that may remain in the syringe after a related injection. Further, an “excess amount” of the formulation is often added to the deliverable volume and the hold - up volume to account for other variables (e.g., filling nozzle data, variations in the filling process, pressure variations, temperature variations, etc.). Thus, the filling target often equals the deliverable volume plus the hold - up volume plus the excess volume. 【0005】 As materials become more difficult to obtain (for example, due to supply problems, high demand, or high costs), unnecessary material usage becomes unacceptable. Accordingly, as materials become more difficult to obtain, it becomes more desirable to reduce the filling targets for the container filling process. 【0006】 Therefore, there is a need for more efficient (e.g., less wasteful) methods for material use in container filling processes. [Overview of the project] [Means for solving the problem] 【0007】 Embodiments described herein relate to methods for reducing the amount of excess material used when filling containers. 【0008】 As described herein, a method for reducing material usage in a container includes obtaining historical fill weight data showing the actual unit-by-unit fill weight volume for a plurality of containers. The method also includes generating a simulated fill weight distribution based on an analysis of the historical fill weight data. The method further includes randomly sampling the simulated fill weight distribution to select a subset of the simulated fill weights and performing a quality check on the subset of the simulated fill weights. The method further includes generating a modified distribution of the simulated fill weights by removing at least the subset samples that failed the quality check from the simulated fill weight distribution. The method also includes converting the modified distribution of the simulated fill weights into a distribution of fill volume and calculating a distribution of deliverable volume based on the distribution of fill volume. The method further includes displaying on a display either or both of (i) the distribution of deliverable volume and (ii) one or more quality performance metrics derived from the distribution of deliverable volume. 【0009】 A non-temporary computer-readable medium that stores computer-readable instructions that, when executed by one or more processors, cause one or more processors to obtain historical fill weight data representing the actual unit-by-unit fill weight volume for multiple containers. Further execution of the computer-readable instructions by one or more processors causes one or more processors to also generate a simulated fill weight distribution based on an analysis of the historical fill weight data. Further execution of the computer-readable instructions by one or more processors causes one or more processors to also randomly sample the simulated fill weight distribution to select a subset of the simulated fill weights and perform quality checks on the subset of the simulated fill weights. Further execution of the computer-readable instructions by one or more processors causes one or more processors to also generate a corrected distribution of simulated fill weights by removing at least the subset samples that failed the quality checks from the simulated fill weight distribution. Further execution of computer-readable instructions by one or more processors causes one or more processors to convert the simulated modified distribution of filling weights into a distribution of filling volumes and calculate the distribution of deliverable volumes based on the distribution of filling amounts. Further execution of computer-readable instructions by one or more processors causes one or more processors to display either (i) the distribution of deliverable volumes or (ii) one or more quality performance metrics derived from the distribution of deliverable volumes. 【0010】 A novel method is provided for reducing the amount of material used when filling containers. 【0011】 Those skilled in the art will understand that the drawings described herein are included for illustrative purposes only and do not limit the disclosure. The drawings are not necessarily to scale and instead focus on illustrating the principles of the disclosure. In some cases, various aspects of the embodiments described may be exaggerated or enlarged to facilitate understanding of the embodiments described. In the drawings, similar reference numerals throughout the various drawings refer to components that are generally functionally and / or structurally similar. [Brief explanation of the drawing] 【0012】 [Figure 1] A schematic block diagram of an example system for reducing material usage in the container filling process is shown. [Figure 2] This document shows an example of a method for generating a container filling process model. [Figure 3] Figure 2 shows a specific embodiment of the method example. [Figure 4] This example demonstrates how to generate a distribution of deliverable volume and quality performance metrics derived from that distribution using a container filling process model. [Figure 5] This document provides an example of how to operate a container filling process based on a container filling process model. [Modes for carrying out the invention] 【0013】 Those skilled in the art will understand that the elements in the figures are drawn for simplification and clarity and are not necessarily drawn to a specific scale. For example, the dimensions and / or relative positions of some elements in the figures may be exaggerated relative to others to help improve the understanding of the various embodiments of the invention. Also, common but well-understood elements that are useful or necessary in commercially viable embodiments are often omitted to make these various embodiments easier to see. Furthermore, it will be recognized that some actions and / or steps may be described or shown in a particular order of occurrence, but those skilled in the art will understand that such specificity regarding order is not actually necessary. Furthermore, it will be recognized that some actions and / or steps may be described or shown in a particular order of occurrence, but those skilled in the art will understand that such specificity regarding order is not actually necessary. Unless a different predetermined meaning is described herein, it will also be understood that the terms and expressions used herein have the ordinary technical meanings that those skilled in the art would give to such terms and expressions, as described above. 【0014】 The various concepts introduced above and discussed in more detail below may be implemented using any of the many methods, and the concepts described are not limited to any particular embodiment. Examples of embodiments are provided for illustrative purposes. 【0015】 This invention provides a method and system for reducing material usage in a container filling process. The system and method generate a container filling process model (e.g., a mathematical model, a statistical model, a machine learning model, etc.) based on historical filling data related to the operation of the filling process. The system and method generate information (e.g., the distribution of deliverable volume, quality performance metrics derived from the distribution of deliverable volume, filling targets, etc.). Using the container filling process model, it is possible to determine whether the filling target results in an acceptable risk to the container filling process (e.g., less than 3 parts per million (3 ppm), as determined using Six Sigma design) of dispensing less than a threshold amount of material into each container. 【0016】 Figure 1 shows a schematic block diagram of system example 100 for reducing material usage in a container filling process. System 100 includes a container filling process model generation (VFPMG) device 125, a history container filling database 120, and a container filling process 105. The container filling process 105 is implemented using a container filling module 106 and a nozzle module 110 for filling multiple nozzles 111. Each nozzle 111 is configured to dispense a predetermined amount of material 116 (e.g., a pharmaceutical product) into its respective container 115 (e.g., a handheld auto-injector, pre-filled syringe, pre-filled cartridge, pre-filled auto-injector, vial, etc.). 【0017】 The container filling module 106 and the nozzle module 110, when executed by the processing unit 127, can be stored in the memory unit 128 as a set of computer-readable instructions that cause the processing unit 127 to communicate container filling data and nozzle data between the container filling process 105, the VFPMG device 125, and the history filling database 120. For example, the processing unit 127 can send filling target data to the container filling process 105 and receive correlated container filling data from the container filling process 105. 【0018】 The history container filling database 120 includes filling data from in-process measurements of each filling nozzle 111 (e.g., filling target data, container filling volume data, deliverable volume data, hold-up volume data, excess volume data, etc.). The history container filling database 120 may also include nozzle data, container data, formulation data, material data, etc. For example, the deliverable volume of a formulation is often specified on the relevant formulation label. The deliverable volume of the formulation is provided by a delivery mechanism such as a handheld autoinjector or syringe. The deliverable volume typically corresponds to information specified on the formulation label, as well as the drug type, drug concentration, etc. The deliverable volume of a formulation may be, for example, 1.0 mL, and the concentration of the formulation may be, for example, 100 mg / mL. 【0019】 The hold-up volume is determined by the material properties of the relevant material (e.g., material density, material viscosity, etc.) and the characteristics of the delivery mechanism (e.g., volume and / or dispensing opening dimensions of a syringe, auto-injector, etc.). 【0020】 Overfill weight is based on the specified deliverable volume (e.g., the deliverable volume specified on the formulation label) and the threshold of increased risk that the relevant container filling process will not dispense the specified deliverable volume, based on a specific filling target. The risk that the corresponding container filling process will not meet relevant standards (e.g., container filling process standards, Six Sigma standards, etc.) is inversely proportional to the overfill weight. 【0021】 The VFPMG device 125 includes a user interface 126, a processing unit 127 (such as one or more processors, etc.), and a memory unit 128 (such as a non - transient computer - readable medium, etc.). The memory unit 128 can store, as a set of computer - readable instructions that, when executed by the processing unit 127, cause the processing unit 127 to generate a container filling process model, a container filling process model generation module 129 and a deliverable volume module 130. For example, the processing unit 127 may use Monte Carlo simulation to generate a virtual fill weight distribution for a large number of lots (such as more than 50, more than 1000, more than 1 million, etc.) of a given formulation. 【0022】 The system 100 can implement an algorithm that uses fixed rules (such as the risk that the container filling process does not dispense at least a threshold amount of material into each container is acceptable, etc.). In addition to, or instead of, the fixed - rule - based model, the system 100 may implement various techniques related to the training (and optionally verification and / or certification) and / or use of one or more neural networks or other non - machine - learning (ML) systems to reduce the material usage of the container filling process. The system 100 can also be used to test / validate a non - ML system for reducing material usage. 【0023】 To facilitate explanation, in this specification, for system 100, one or more container filling processes 105 are trained and verified using history container filling data from history filling data 120, and then, using the trained / verified neural network, for example, it is described as determining the risk that the container filling process does not dispense a minimum amount of material in each container based on a specific filling target. However, it should be understood that this need not necessarily be the case. For example, the training / verification may be performed by another system, and system 100 may then use the trained neural network (e.g., during commercial production). In some embodiments, some or all of the history filling data used for training and / or verification closely reproduces important aspects of commercial line equipment stations (e.g., containers, materials, container filling processes, nozzles, etc.) using one or more offline (e.g., laboratory-based) "simulation stations", thereby expanding the training and / or verification library without causing excessive downtime of the commercial line equipment. 【0024】 However, in some embodiments, system 100 includes two or more computers that are either co-located with each other or remote from each other. In these distributed embodiments, the operations described herein regarding processing unit 127 and memory unit 128 may be divided among multiple processing units and / or memory units, respectively. 【0025】 The processing unit 127 may include one or more processors, each of which may be a programmable microprocessor that executes software instructions stored in the memory unit 128 to perform some or all of the functions of the container filling process 105 as described herein. The processing unit 127 may include, for example, one or more central processing units (CPUs). Alternatively, or in addition, some of the processors in the processing unit 127 may be other types of processors (e.g., application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), etc.), and some of the functions of the system 100 as described herein may instead be implemented in hardware. 【0026】 The memory unit 128 may include one or more volatile and / or non-volatile memories. The memory unit 128 may include one or more of any suitable memory types, such as read-only memory (ROM), random access memory (RAM), flash memory, solid-state drives (SSDs), and hard disk drives (HDDs). Together, the memory unit 128 can store one or more software applications, data received / used by those applications, and data output / generated by those applications. 【0027】 The memory unit 128 stores software instructions for various modules that, when executed by the processing unit 127, perform various functions for the purpose of training, verifying, and / or certifying one or more mathematical models, statistical models, artificial intelligence (AI) neural networks, etc. Specifically, in the embodiment shown in Figure 1, the memory unit 128 stores instructions for the container filling process model generation module 129 and the deliverable volume module 130. In other embodiments, the memory unit 128 may omit one or both of modules 129 and 130, and / or include one or other modules. In addition, or alternatively, one or both of modules 129 and 130 may be implemented by a different computer system (e.g., a remote server connected to system 100 via one or more wired and / or wireless communication networks). Furthermore, the functions of either or both of modules 129 and 130 may be divided between different software applications and / or computer systems. As just one example, in an embodiment in which system 100 accesses a web service to train and use one or more container filling process models, the software instructions for the container filling process model generation module 129 may be stored on a remote server. 【0028】 The container filling process model generation module 129 includes software for training (generating and / or updating) one or more models using historical filling data 120. The historical filling data 120 may be stored in the memory unit 128 or in another local or remote memory (e.g., memory linked to a remote library server). In addition to training, the module 129 may optionally implement / run the models by applying filling data acquired by the container filling process 105 (or another container filling system) to the models after certain preprocessing has been performed on the historical filling data, as described later. In various embodiments, the models trained and / or run by the module 129 can determine the risks associated with a container filling process that fail to dispense at least a threshold amount of material (e.g., a filling volume) into each container, based on a specific filling objective. 【0029】 Module 129 may run the trained model for verification, certification, and / or inspection purposes during commercial production. In one embodiment, for example, Module 129 is used solely for training and verifying the model, and the trained model is then transferred to another computer system (for example, using another module similar to Module 129) for certification and inspection during commercial production. In some embodiments in which the container filling process model generation module 129 trains / runs multiple models, Module 129 includes separate software for each model. 【0030】 In some embodiments, the deliverable volume module 130 controls / automates the operation of the container filling process 105 so that historical filling data 120 can be generated with little or no human interaction. The deliverable volume module 130 can cause a given container filling process 105 to capture container filling data by sending commands or other electronic signals (e.g., generating pulses on a control line). The container filling process 105 can send the container filling data to the VFPMG device 125, which can store the filling data in the memory unit 128 for local processing. In alternative embodiments, the container filling process 105 may be locally controlled, in which case the deliverable volume module 130 may have fewer functions than those described herein (e.g., only handling the retrieval of historical filling data) or may be completely omitted from the memory unit 128. The VFPMG device 125 may implement a container filling data module 106 and / or a nozzle data module 110 to implement local control. 【0031】 A container filling process model (e.g., a mathematical model, a statistical model) can be combined with Six Sigma (DFSS) design to optimize deliverable volume in release testing for commercial filling operations. The container filling process model uses historical filling weight data and release testing deliverable volume data to predict the distribution of filling volume and deliverable volume. The container filling process model determines the percentage of containers that fall below the specification limit for deliverable volume. The model then simulates the decrease in filling weight to optimize the deliverable volume during release testing so that a sufficient margin against the specification limit is maintained. 【0032】 Figure 2 shows a method 200 for generating a container filling process model (e.g., a mathematical model, a statistical model, an AI model, etc.), which can be implemented by a processor (e.g., the processing unit 127 in Figure 1) executing, for example, at least part of a container filling process model generation module 129 and / or deliverable volume module 130. In particular, the container filling process model generation module 129 can accept a hold-up volume (HUV) distribution input (block 225). The HUV distribution input may represent filling weight data from in-process measurements of multiple (innumerable) previously filled containers (samples), for example, represented by historical filling data 120. 【0033】 Filling weight data from in-process measurements can include a correlation with each individual nozzle 111. For example, historical filling weight data (showing the actual filling weight in past processes) may be collected for each filling nozzle (e.g., for 16 different nozzles, for one nozzle) across multiple lots or batches of a particular formulation. 【0034】 The container filling process model generation module 129 can determine the total deliverable volume for the entire lot (block 230). For example, the container filling process model generation module 129 can multiply the filling value for each container (e.g., filling volume, filling weight, etc.) by the total number of containers in each lot. 【0035】 The container filling process model generation module 129 can receive filling nozzle data input (block 235). For example, the container filling process model generation module 129 can receive the distribution of the mean and standard deviation of the amount of material discharged from each nozzle. The historical data 120 can include, for example, this received data. 【0036】 The container filling process model generation module 129 can generate a filling weight distribution (block 237). Historical data 120 can include, for example, this filling weight distribution data. For example, the processing unit 127 can generate data 236 showing the filling weight distribution generated by performing a Monte Carlo simulation. The filling weight distribution can include a correlation with the material discharged from each nozzle 111. Based on this data, the statistical (Monte Carlo) model predicts a vast number (e.g., countless) of filling weight-related results for each nozzle. 【0037】 The container filling process model generation module 129 can isolate and inspect a subset 238 (e.g., 2 percent) of samples from the distribution of filling weights 2. The container filling process model generation module 129 can perform a filling weight data quality check on the subset 238 (block 240). For example, the container filling process model generation module 129 can virtually perform quality control checks that reflect what is physically done on the production line, such as random sorting / selection, and then checking 2% of the randomly sorted / selected sample / simulation values ​​for each nozzle. The filling weight distribution may include correlations of the amount of material released by each nozzle 111. 【0038】 The container filling process model generation module 129 can generate a rejection control signal 245 and inspect any nozzle-related data that fails the quality check (block 246). For example, if a single sample is below the limit, a non-conforming (NC) counter can be triggered and the sample can be virtually removed. 【0039】 The container filling process model generation module 129 can combine nozzle-related data that have passed quality checks by aggregating simulation data attributable to those nozzles (block 247). The container filling process model generation module 129 can convert material weight into filling volume from the aggregated simulation data attributable to the nozzles that have passed quality checks (block 248). For example, the container filling process model generation module 129 can divide the weight of the material (mg) by the density of the liquid containing the material (mg / mL). In this example, if the filling weight of the formulation is 1.080 g and the density of the liquid containing the material is 1080 mg / mL, the deliverable volume of the formulation is 1.0 mL. 【0040】 The container filling process model generation module 129 can generate a distribution of deliverable volumes based on the container filling process model. Alternatively, the container filling process model generation module 129 can generate quality performance metrics (e.g., the risk that the relevant container filling process will not meet the specified deliverable volume) derived from the distribution of deliverable volumes based on the container filling process model. 【0041】 The deliverable volume of the formulation may be, for example, 1.0 mL, and the concentration of the formulation may be, for example, 100 mg / mL. The deliverable volume of the formulation may be, for example, 1.0 mL, and the filling weight of the formulation may be, for example, 1.080 g. A reduction of 0.015 g in container filling weight can be verified to result in an acceptable risk using the container filling process model generated by Method 200. 【0042】 Figure 3 shows a specific embodiment of Method 200 of Figure 2. In Figure 3, Method 300 for generating container filling data uses a container filling process model (e.g., a mathematical model, a statistical model, an AI model, etc.). Method 300 can be implemented by a processor (e.g., processing unit 127 in Figure 1) executing, for example, at least a portion of the container filling process model generation module 129 and / or deliverable volume module 130. In particular, the container filling process model generation module 129 can accept a hold-up volume (HUV) distribution input (block 325). The HUV distribution input may represent filling weight data from in-process measurements of multiple (innumerable) previously filled containers (samples), for example, represented by historical filling data 120. The HUV distribution input may represent filling weight data from in-process measurements, for example. The filling weight data from in-process measurements may include a correlation with each respective nozzle 111. Historical filling weight data (showing the actual filling weight in past processes) is collected for each filling nozzle (for example, for one nozzle for 16 different nozzles) across multiple lots or batches of a particular drug. 【0043】 The container filling process model generation module 129 can determine the total deliverable volume for the entire lot (block 330). For example, the container filling process model generation module 129 can multiply the filling value for each container (e.g., filling volume, filling weight, etc.) by the total number of containers in each lot. 【0044】 The container filling process model generation module 129 can receive filling nozzle data input (block 335). For example, the container filling process model generation module 129 can receive the distribution of the mean and standard deviation of the amount of material released from each nozzle. The historical data 120 may include data representing the distribution of the mean and standard deviation of the amount of material released from each nozzle. 【0045】 The container filling process model generation module 129 can generate a filling weight distribution (block 337). The historical data 120 may include data representing the filling weight distribution. For example, the processing unit 127 can generate data 336 based on the filling weight distribution generated by performing a Monte Carlo simulation. The filling weight distribution may include a correlation with the material discharged from each nozzle 111. Based on this data, a statistical (Monte Carlo) model can predict a vast number (e.g., countless) of filling weight-related results for each nozzle. 【0046】 The container filling process model generation module 129 can isolate and inspect a subset 338 (e.g., 2 percent) of samples from the filling weight distribution. The container filling process model generation module 129 can perform a filling weight data quality check on the subset 338 (block 340). For example, the processing unit 127 can virtually perform a quality control check that reflects an in-process control (IPC) filling weight check, where first, 2% of the samples / predictions for each nozzle (e.g., 16 nozzles, 1 nozzle, etc.) are randomly selected / sampled and checked (block 341). The filling weight distribution may include correlations of the amount of material released by each nozzle 111. The processing unit 127 can virtually remove samples that fall below the IPC threshold limit (block 342). The processing unit 127 can perform standard deviation control (block 343). 【0047】 The container filling process model generation module 129 can generate a rejection control signal 345 and inspect any nozzle-related data that fails the quality check (block 346). For example, if a single sample is below the limit, a non-conforming (NC) counter can be triggered and the sample can be virtually removed. 【0048】 The container filling process model generation module 129 can combine nozzle-related data that have passed quality checks by aggregating simulation data attributable to those nozzles (block 347). The container filling process model generation module 129 can convert material weight to the filling volume of the aggregated simulation data attributable to the nozzles that have passed quality checks (block 348). For example, the container filling process model generation module 129 can divide the material weight (mg) by the material density (mg / mL). For example, the deliverable volume of the formulation may be 1.0 mL, and the filling weight of the formulation may be 1.114 g. 【0049】 The container filling process model generation module 129 can generate a total deliverable volume population output (block 356). The container filling process model generation module 129 can generate an output of the percentage of units below the drug volume (DV) lower limit (LSL) (block 357). The container filling process model generation module 129 can generate an output of the percentage of lots with nonconformities (block 358). 【0050】 The deliverable volume module 130 can generate a distribution of deliverable volumes based on a container filling process model. Alternatively, the processing unit 127 can run the deliverable volume module 130 to generate, for example, quality performance metrics derived from the distribution of deliverable volumes based on a container filling process model. 【0051】 The deliverable volume of the formulation may be, for example, 1.0 mL, and the concentration of the formulation may be, for example, 100 mg / mL. The deliverable volume of the formulation may be, for example, 1.0 mL, and the filling weight of the formulation may be 1.080 g. A reduction of 0.015 g in container filling weight can each be verified to result in an acceptable risk using the container filling process model generated using Method 300. 【0052】 Figure 4 shows a method 400 for generating a distribution of deliverable volume and a quality performance metric derived from the distribution of deliverable volume, based on a container filling process model (e.g., a statistical model). Method 400 can be implemented by a processor (e.g., processing unit 127 in Figure 1) executing, for example, at least a portion of the container filling process model generation module 129 and / or the deliverable volume module 130. In particular, the container filling process model generation module 129 can receive historical filling weight data (block 410). The HUV distribution input can represent, for example, filling weight data from in-process measurements. The filling weight data from in-process measurements can include a correlation with each nozzle 111. The historical filling weight data (showing the actual filling weight in past processes) is collected for each filling nozzle (e.g., for 16 different nozzles, for 1 nozzle, etc.) across multiple lots or batches of a particular drug. 【0053】 The container filling process model generation module 129 can generate a simulated distribution of filling weights (block 415). The container filling process model generation module 129 can select a subset of the simulated distribution of filling weights (block 420). 【0054】 The container filling process model generation module 129 can perform quality checks on a subset of simulated filling weights (block 425). The container filling process model generation module 129 can generate a corrected distribution of simulated filling weights (block 430). The container filling process model generation module 129 can convert the corrected distribution of simulated filling weights into a distribution of filling volumes (block 435). 【0055】 The container filling process model generation module 129 can calculate the distribution of filling volume (block 440). The deliverable volume module 130 can display the distribution of filling volume and one or more quality performance metrics (block 445). 【0056】 Figure 5 shows an example method 500 for operating a container filling process based on a container filling process model (e.g., a statistical model). Method 500 can be implemented by having a processor (e.g., processing unit 127 in Figure 1) execute, for example, at least a portion of the container filling module 106 and / or nozzle module 110. In particular, the container filling module 106 can receive filling target data (block 510). 【0057】 The nozzle module 110 can fill containers based on filling target data (block 515). The container filling module 106 can receive actual filling weight data (block 520). 【0058】 The container filling module 106 can generate a feedback filling target (block 525). The nozzle module 110 can fill the container based on the feedback filling target (block 530). 【0059】 Systems, methods, apparatus, and their components have been described in terms of exemplary embodiments, but they are not limited to these exemplary embodiments. The detailed description should be interpreted as illustrative only, and does not describe all possible embodiments of the invention, as it would be impractical, if not impossible, to describe all possible embodiments of the invention. Many alternative embodiments can be implemented using either the current art or art developed after the filing date of this patent, and these should still be included within the claims defining the invention. 【0060】 Those skilled in the art will understand that a variety of modifications, variations, and combinations can be created from the embodiments described above without departing from the scope of the present invention, and that such modifications, variations, and combinations should be interpreted as falling within the scope of the concept of the present invention.

Claims

[Claim 1] A method of filling a container with a pharmaceutical product, To obtain historical filling weight data showing the actual filling weight volume per unit for multiple containers, Based on the analysis of the aforementioned historical filling weight data, a simulated filling weight distribution is generated. The distribution of the simulated filling weights is randomly sampled to select a subset of the simulated filling weights, Performing quality checks on the subset of the simulated filling weights, At a minimum, a corrected distribution of simulated filling weights is generated by removing the subset of samples that failed the quality check from the simulated filling weight distribution, Converting the simulated corrected distribution of filling weight into a distribution of filling volume, Based on the distribution of the packing volume, the distribution of deliverable volume is calculated, (i) displaying on the display either the distribution of deliverable volume or (ii) one or both of the quality performance metrics derived from the distribution of deliverable volume, Methods that include... [Claim 2] The method according to claim 1, wherein the analysis of the historical filling weight data includes Monte Carlo analysis. [Claim 3] The method according to claim 1 or 2, wherein the historical filling weight data includes a correlation between each of the plurality of containers and each nozzle associated with filling the container, and the analysis of the historical filling weight data includes data indicating the filling weight volume for each nozzle. [Claim 4] The method according to claim 3, wherein the quality check identifies at least one nozzle that is non-compliant. [Claim 5] The method according to claim 4, wherein the quality check identifies all containers associated with the at least one nozzle as non-conforming, and the modified distribution of simulated filling weights excludes containers associated with the at least one nozzle. [Claim 6] The method according to any one of claims 1 to 5, wherein the quality check includes determining whether the container filling level is below a minimum filling level threshold. [Claim 7] The method according to any one of claims 1 to 6, wherein the quality check includes determining the standard deviation of the filling level and determining whether the standard deviation exceeds a filling level threshold. [Claim 8] The method according to any one of claims 1 to 7, wherein calculating the distribution of deliverable volumes includes generating the distribution of deliverable volumes based on (i) the distribution of the filling volume and (ii) historical deliverable volume data or historical hold-up volume data. [Claim 9] In real-time operation of the production line, and based on the distribution of deliverable volume, adjust the target weight of the filling process of the production line. The method according to any one of claims 1 to 8, further comprising: [Claim 10] A system for filling a container with a pharmaceutical product, comprising one or more processors and at least one memory which stores computer-readable instructions that, when executed by the one or more processors, cause the one or more processors to carry out the method according to any one of claims 1 to 9. [Claim 11] When executed by one or more processors, the one or more processors will To obtain historical filling weight data showing the actual filling weight volume per unit for multiple containers, Based on the analysis of the aforementioned historical filling weight data, a simulated filling weight distribution is generated. Randomly sampling the distribution of the simulated filling weights to select a subset of the simulated filling weights, Performing quality checks on the subset of the simulated filling weights, At a minimum, a corrected distribution of simulated filling weights is generated by removing the subset of samples that failed the quality check from the simulated filling weight distribution, Converting the simulated corrected distribution of filling weight into a distribution of filling volume, Based on the distribution of the packing volume, the distribution of deliverable volume is calculated, (i) displaying on the display either the distribution of deliverable volume or (ii) one or both of the quality performance metrics derived from the distribution of deliverable volume, A non-temporary computer-readable medium that stores computer-readable instructions for performing a certain action. [Claim 12] The non-temporary computer-readable medium according to claim 11, wherein the analysis of the historical filling weight data includes Monte Carlo analysis. [Claim 13] The non-temporary computer-readable medium according to claim 11 or 12, wherein the historical filling weight data includes a correlation between each of the plurality of containers and each nozzle associated with filling the container, and the analysis of the historical filling weight data includes data indicating the filling weight volume for each nozzle. [Claim 14] The non-temporary computer-readable medium according to claim 13, wherein the quality check identifies at least one nozzle that is non-compliant. [Claim 15] The non-temporary computer-readable medium according to claim 14, wherein the quality check identifies all containers associated with the at least one nozzle as non-conforming, and the modified distribution of simulated filling weights excludes containers associated with the at least one nozzle. [Claim 16] The non-temporary computer-readable medium according to any one of claims 11 to 15, wherein the quality check includes determining whether the filling level is below a minimum filling level threshold. [Claim 17] The non-temporary computer-readable medium according to any one of claims 11 to 16, wherein the quality check includes determining the standard deviation of the filling level and determining whether the standard deviation exceeds the filling level threshold. [Claim 18] The non-temporary computer-readable medium according to any one of claims 11 to 17, wherein calculating the distribution of deliverable volumes includes (i) generating the distribution of deliverable volumes based on the distribution of the filled volume and (ii) historical deliverable volume data or historical hold-up volume data. [Claim 19] The non-temporary computer-readable medium according to any one of claims 11 to 18, wherein further execution of the computer-readable instructions by the one or more processors causes the one or more processors to adjust the target weight of the filling process of the production line in real-time operation of the production line and based on the distribution of the deliverable volume.

Images