The pharmaceutical industry is undergoing a radical change in its ways of working. Directives from the US Food and Drug Administration have raised questions about understanding the relationships between formulation and manufacturing that result in controlled product performance. These relationships, however, can rarely be precisely quantified, and the formulation and manufacturing must be carried out in a design space that is both multidimensional in nature and difficult to conceptualize. Attempts to investigate these problems through the use of experimental design have generated large amounts of data, but processing these data remains a challenge. This article introduces two new implementations of data-mining software packages (INForm and FormRules, Intelligensys, Stokesley, UK) that are specifically tailored for the pharmaceutical formulator or process engineer to generate understandable rules and to model and optimize the process and formulation.  

Both INForm and FormRules rely on advanced computing techniques such as neural networks, fuzzy logic, and genetic algorithms. 

 are mathematical constructs that are capable of learning, for themselves, the relationships within data. No assumptions need be made about the functional form of these relationships because the neural network simply tries out a range of models to determine one that best fits the known data. In recent years, artificial neural networks (ANNs) have increasingly and successfully been used to model complex behavior in problems such as those found in pharmaceutical formulation and processing (1, 2).  

 can be implemented to allow a formulator's objectives to be described in a linguistically intuitive way. Traditional "crisp" logic means that values must be either "true" (1) or "false" (0). Fuzzy logic, based on the theory of fuzzy sets, allows the membership in each set to take a value between 0 and 1. For example, if a tablet disintegration time of <300 s is desired, then a value >300 s will have a desirability of <100%, with the desirability decreasing as the disintegration time increases. This gives a formulator considerable control over an optimization process.  

As the name implies, genetic algorithms use an evolutionary approach to finding the best solutions. To do this, a measure of fitness is set up, using the desired values for each property together with its importance relative to other properties. The optimization starts with a random trial population, and the fitness of each member in the population is assessed. New solutions are generated from the fittest members, using mathematical operations that are analogous to reproduction and mutation, and their fitness is assessed. In this way, the population evolves so that ultimately the fittest solution is the one that best meets a formulator's specified needs. If there are constraints on the ingredients or processing conditions —for example, if a particular combination of ingredients must sum to 100% —then these can be implemented easily by penalizing the fitness of nonconforming solutions. 

Combining these technologies allows the development of useful and powerful methodologies. For example, using neural networks for modeling together with genetic algorithms for optimization (as is done within the INForm software system) allows a user to develop a formulation or process to meet stringent, often conflicting, objectives. New methodologies such as neurofuzzy logic (implemented in the FormRules system) have evolved. This combines the ability of neural networks to "learn" from data with fuzzy logic's capacity to express complex concepts simply, allowing a formulator and process engineer to gain an understanding of the underlying rules governing the formulation and process.  			

This article examines the applications of the software systems to formulation and processing using data taken from three examples in the literature. 			

The first case concerns a roller-compaction process for acetaminophen tablets using data published by Turkoglu 

(3). Because acetaminophen has poor flow and compression characteristics, a prior agglomeration process is generally used. In Turkoglu's study, both the formulation and the process conditions were changed. In the formulation, the binder was one of three possibilities: hydroxypropyl methyl cellulose (HPMC, Methocel, Dow Chemical Co., Midland, MI), polyethylene glycol, or carbomer (Carbopol, Noveon, Cleveland, OH). The percentage of binder and the amount of microcrystalline cellulose (MCC) that was added were varied. One or two passes through the roller compactor was allowed, and 42 different experiments were measured, 30 of which were used to develop the models. These 30 experiments were used in FormRules, to determine which inputs were most important and to investigate how they affect the measured properties: crushing strength of the tablets, friability, ejection force, and disintegration time. With the default training parameters, good models (as assessed by an analysis of variance statistics, which all showed the value of R

  > 0.9) were obtained for each property.

Table I: Effect of binder type and binder addition on ejection force and crushing strength.			

The models showed that both ejection force and crushing strength of the tablets depended on all four input variables. In addition, the data mining discovered an interaction between the binder type and binder concentration for both properties, as illustrated for the ejection force in Figure 1. Rules for both properties are summarized in Table I, where it is clear that, regardless of concentration, ejection force is low when polyethylene glycol is used. Data mining has also highlighted that when Carbopol is used, the crushing strength is low, regardless of the concentration. Therefore, the data-mining exercise can provide useful pointers for the choice of binder, depending on the properties that are required of the finished tablet.  

Figure 1: Graphical representation of model for ejection force.			

The other submodels developed during data mining showed that both the ejection force and the crushing strength tend to increase when the number of compaction passes was increased from one to two and that increasing amounts of MCC will also increase the ejection force and crushing strength. These effects were less marked, however, and the major contribution to these two properties comes from the nature of the binder and its concentration. Models for the other properties showed that disintegration time did not depend significantly on the number of compaction passes, and the friability depended only on the binder type and binder concentration. 

Modeling and optimization for spheronization

Hard gelatin capsules containing spheronized pellets are now a recognized delivery system for controlling or modifying the release of drugs in the body. Uniformity of pellet shape is essential both for coating and for subsequent filling into capsules. Baert 

 (4) investigated a binary mixture of Avicel PH 101 and water. Their work looked at how changing the Avicel:water ratio, the spheronization time, and the spheronization speed affects the yield and roundness of the spheronized particles. Forty-five different combinations of formulation and process conditions were investigated, by using three spheronizer speeds, five spheronization times, and three Avicel:water ratios. These data points were used in FormRules so that the key relationships could be extracted. The results showed that good models could be found for both roundness and yield. Had a poor model been obtained, it would indicate that an uncontrolled variable was affecting the results. Here, that was not the case, and all important variables had been measured.   

Figure 2: Graphical representation of model for roundness.			

In the published data, the roundness of the particles was defined as the ratio of the largest to the smallest diameter of the spheres (

the aspect ratio), so that a low value of roundness represents an approximately spherical particle. Our data-mining study shows that roundness is determined by all three input variables: Avicel:water ratio, spheronization speed, and spheronization time. As Figure 2 shows, all three variables contribute to the same submodel. This model indicates that there is an interaction between the variables, and it can clearly be seen when the model is expressed in rule form. The rules governing roundness, which were mined automatically from the data, are summarized in Table II. Values in parentheses are the confidence levels that indicate how high or low a particular value is. 

Table II: Effect of spheronization speed, spheronization time and avicel:water ratio on pellet roundness.			

A full examination of the rules shows that when the spheronization time lies in the mid-to-high range of examined values, then the spheronization speed has its greatest influence when the Avicel:water ratio is high. If the spheronization speed is low and the Avicel:water ratio is high, the particles will be markedly aspherical. If the Avicel:water ratio is low (

 the water content is high), however, then the spheronization speed has only a small effect. Different behavior is observed when the spheronization time is short. In that case, the spheronization speed is important for all values of Avicel:water ratio. In this way, the data mining exercise highlights which inputs must be controlled most closely and also points out which combination of conditions will lead to the most spherical shapes. 

The data mining exercise produces even more interesting results for the yield, because it was found that spheronization time did not contribute significantly to the model. There is an interaction between the spheronization speed and the Avicel:water ratio, as shown in the three-dimensional plot shown in Figure 3. Lower Avicel:water ratios and lower speeds led to the highest yields. 

Figure 3: Contribution of the ratio of Avicel to water and spheronization speed to yield.			

Using the INForm software system, it is also possible to determine the process conditions and formulation that lead to optimum yield and sphericity. The aim of the optimization is to produce a high yield of particles that have a roundness value close to one. Supplementary to this is the assumption that if spheronization time could be decreased without sacrificing yield or sphericity, then processing could be sped up. Within the INForm software system, the objectives could be assigned a relative importance. In this case, yield was assigned a value of 8, and roundness was taken to have importance of 10 (on a 0–10 scale). Conditions leading to a roundness value >1.05 were assumed to be unacceptable when setting up the fitness criterion for the genetic algorithm optimization.  

These optimization objectives could be achieved using a formulation containing 42.5% Avicel at a spheronization speed of 950 rpm and a time of 30 min. When the spheronization speed was fixed at only 5 min, reasonable results could still be achieved using a similar formulation with a spheronization speed of 870 rpm, so that yield and roundness did not need to be sacrificed significantly when the spheronization time is reduced. 

Another process that has been investigated using neural computing is tablet coating. Mitchell looked at the effect of changing batch size, drum speed, spray rate, atomizing air pressure, inlet air temperature, inlet air relative humidity, exhaust temperature, and bed temperature on the color and color variation of a tablet coating using a coater (HCT-20 Mini Hi-Coater, Vector Corp., Marion, IA) for which 49 different experiments were carried out (5). As she noted, the inlet air relative humidity, the exhaust temperature, and the bed temperature could not be controlled with the HCT-20 machine. It was clear from her results, however, that the tablet-bed temperature had a significant effect on the properties of the finished coating, because leaving this parameter out of the models gave a considerably poorer fit to the experimental data. The tablet-bed temperature depends on the inlet-air temperature and on the volumetric airflow rate, but there is limited control over these factors  in the HCT-20.  

Mitchell's data have been examined using neural networks (embodied in the INForm software package) using 45 experiments for training and the remaining 4 for model validation. Using all eight input variables, good models, as assessed by analysis of variance statistics, could be obtained for all properties except gloss. Reasonably good models could be obtained when only the five directly controllable variables were used. 

Mitchell also reported an optimization of the process conditions using the neural network models. The optimization was performed using genetic algorithms to evolve the fittest solutions, where the criterion of fitness involved fuzzy logic to specify the process goals. The aim was to have the color around 80, the color variation as low as possible to indicate a uniform film. It is possible within the INForm software system to select which of the properties is the most important by weighting the properties on a 0–10 scale, and Mitchell chose to treat color and color variation as equally important. She also allowed all the input variables, including the batch size, to participate in the optimization. This gave a set of optimization conditions for the HCT-20 coater, suggesting that the optimum batch size was 309 g, with drum speed of 11 rpm, and spray rate of 2 g/min.  

Table III: Optimum process conditions for the HCT-20 coater.			

Nonetheless, it may be more convenient to assume a specific batch size and to determine the best operating conditions for that. From the optimization performed here for various batch sizes, the best conditions are summarized in Table III. At larger batch sizes, the optimum values for drum speed and atomizing air pressure are at the low end of the experimentally used range, thereby suggesting that a further decrease, if feasible, would be beneficial.  

The ability to fix specific input parameters and to specify the relative importance of each of the properties, gives a high degree of control over the optimization, allowing a formulator or process engineer to determine exactly the best operating conditions for their equipment. A sensitivity analysis function is built into the software system so that the robustness of the process to small variations in the process conditions can be assessed.  

Benefits in process development: better products faster

Neural computing and advanced computing systems have a proven track record in pharmaceutical formulation (1, 2). Their implementations in INForm and FormRules allow formulators and process engineers to extract maximum value from their formulation and processing data, thereby improving both the efficiency and the effectiveness of their product development cycle. Using these approaches, key relationships can be discovered rapidly, expressed succinctly, and communicated clearly. Because the methods are entirely data-driven, they can be applied to many formulation and processing problems. The only requirement is the availability of data of reasonable quality and quantity.  

Forward-thinking pharmaceutical companies are already adopting these technologies as part of a decision to support tool kits for their product formulators and process engineers, and all indications are that such advanced computing techniques will be used routinely in the future.  

 is the product director at Intelligensys Ltd., Springboard Business Centre, Ellerbeck Way, Stokesley, TS9 5JZ, UK, 

 is the chief scientist, Intelligensys Ltd, Springboard Business Centre, Ellerbeck Way, Stokesley TS9 5JZ, UK. 

*To whom all correspondence should be addressed.

Keywords: advanced computing, data mining, formulation, fuzzy logic, genetic algorithms, neural networks, software 

 1. E.A. Colbourn and R.C. Rowe, "Neural Computing Boosts Formulation Productivity," 

 2. E.A. Colbourn and R.C. Rowe, "Neural Computing and Formulation Optimization," in 

Encyclopedia of Pharmaceutical Technology 

 "Modelling of a Roller Compaction Process using Neural Networks and Genetic Algorithms," 

 "Studies of Parameters Important in the Spheronization Process," 

 5. K. Mitchell, "The Scale-Up, Modelling, and Optimisation of Aqueous Film Coating Processes," PhD Thesis, University of Bradford, UK (2003).  

Articles

Large amounts of data are currently being generated in an attempt to understand and improve formulation, process, and manufacturing efficiency. This task requires novel data mining software systems tailored specifically for the formulator and process engineer.

Advanced Neural Computing Software Systems: Data Mining in Processing and Formulation

A global plastic manufacturer operates one of its plants at 104% of its theoretical maximum production rate. A cement company achieves 32% reduced product variability and improved process stabilization. Why are these industries' productivity values relevant to the pharmaceutical industry, and how are they attaining those results? 

These companies demonstrate the benefits of measuring, analyzing, and controlling processes to achieve business benefits using principles that life-sciences companies are pursuing today under the process analytical technology (PAT) umbrella. Of course, the life-sciences industry has understood the returns that stem from quality assessment and adaptive development. So, why are experienced industry executives asking, "Haven't we been doing this for 20 years in our industry?"  

Make no mistake: the challenges in the pharmaceutical industry are definitely unique because of the industry's highly sophisticated regulations and complex processes such as cell culture and antibiotic fermentations. Still, there have been lessons already learned and problems that have already been solved in consumer products, refining, metals, and a host of other industries. With the US Food and Drug Administration's PAT initiative, the pharmaceutical industry is now heavily engaged in evaluating its own measurement, analysis, and control capabilities and it is looking to other industries for proven technologies and approaches that may be able to transfer to pharmaceutical operations.  

PAT goes far beyond closed-loop control, which has been a standard operating procedure in other industries for a long time, and it isn't completely new to life sciences, especially in biotechnology and fermentation processes. Unfortunately, much of the past emphasis on applying good chemical engineering principles to drug manufacturing was put aside during the past several years as the industry redirected its attention to Y2K, business-system implementations, and 21 

 Part 11. The result may have enhanced compliance, but it did little to achieve the goal of pharmaceutical manufacturing, which is to make high-quality products.  

FDA's PAT initiative allows pharmaceutical manufacturers to refocus on improving their processes for making drug products more efficiently and with less variability. Moreover, it allows companies to focus on determining which variables are most critical to the final desired product and where controls should be inserted and implemented into the development cycle. The success of PAT is applying the process-monitoring tools needed to analyze each of the critical product attributes and then effectively and accurately controlling them. Detecting errors or process deviations and correcting them while the product is being made is more cost-efficient and can help justify flexible regulatory paths for innovations in manufacturing as well as postapproval changes.  

Applying principles associated with the PAT initiative that have been identified by FDA, including process understanding and real-time releases, requires a change in technology and techniques currently used in many pharmaceutical operations. Life-sciences companies have undertaken numerous pilot projects to determine:  					

how to measure what must be measured (sensor development); 

how to validate those measurements to identify meaningful process-parameter relationships;

how to integrate advanced strategies to achieve reliable, predictable operations. 

Defining and implementing an architecture that incorporates all of these elements is critical to achieving real-time releases of product that leverages PAT principles and brings to fruition the business benefits that are the driving forces behind these investments. 

FDA identifies the tools for supporting innovation as multivariate tools for design, data acquisition, and analysis; process analyzers; process-control tools; and continuous-improvement and knowledge-management tools (see Figure 1). 

 Figure 1: Tools and parameters to support and implement process analytical technology principles during pharmaceutical product development.			

This article examines the specific elements of measurement, analysis, and control and what other industries have done to address these challenges in successful implementations using concepts that can be applied directly to the life-sciences industry. 

What type of sensors must be in place to measure which process parameters? Many companies are undergoing sensor development pilots, realizing that if a parameter can't be measured, it can't be analyzed or controlled. Some sensing problems can be solved with existing analytical instrumentation, but others will require new technology to measure the parameter of interest in a nondestructive, real-time manner.  

Food, beverage, drinking-water, and wastewater industries are finding that multielement sensors (MEMS) applications can help them identify trace contaminants, degradation patterns in fluids, and subtle patterns in chemical fingerprints correlated to quality. MEMS can simultaneously provide five or more process readings on one sensor, including, for example, pH, oxidation, viscosity, and temperature. MEMS units also can withstand sterilization and harsh conditions. There is clearly a trend toward digital sensors, smart sensors (

 local processing and lab-on-a-chip devices) and distributed sensor networks. Within the sensor arena, considerable effort is expended on biosensors and MEMs devices such as microfluidics, and new sensors for biological applications appear almost weekly. 

What data are necessary for analyzing critical process parameters, and what are the tools that will be needed to conduct these analyses? FDA has identified the tools and concepts supporting innovation as "multivariate tools for design, data acquisition, and analysis; process analyzers; process-control tools; and continuous-improvement and knowledge-management tools." There are several off-the-shelf tools on the market that have been proven in pharmaceutical and other industries and that use a range of multivariate techniques to analyze process data. But tools such as these will not magically provide the answers about which parameters are critical control parameters or how to control a process.  

Good process knowledge must be applied to determine which data sets to include, how to reconcile disparate data—such as time-stamped, tag-based data; sample-based analytical data; and raw-material parameters—and how the data should be cleansed and preprocessed. For example, in a batch or fed-batch operation, what is the best way to unfold the data to capture the maximum information content while producing valuable results? The bottom line on data analysis is that there are many algorithms and tools available, and applying good process knowledge during the analysis will help achieve the best results. 

What about closed-loop control and real-time releases? As in other industries, implementing the elements of PAT is only valuable if it ultimately helps a pharmaceutical company achieve its business goal. Process-control tools, the third of the PAT tools identified by FDA, help realize the potential of accurate sensor measurements and analysis.  

For example, a cement manufacturer achieved more consistent product quality using on-line, closed-loop multivariate control in its milling operations. Using a multivariable predictive controller and soft sensors, the mill optimizer ensured smooth handling of process nonlinearities across multiple process units. In abnormal conditions, the optimizer switched to a rule-based control mechanism to restore stability. The result was a 30% reduction in product variability and improved process stabilization. 

An alphaolefins plant operated at 105% of its theoretical maximum production rate as a result of optimized closed-loop control. The oxidation rate in the batch processing is now carefully controlled with multiple setpoints to achieve consistent quality with virtually no rework or waste—far improved over the 30% rework before closed-loop control. Batch processes such as these have parallels in pharmaceutical batch processes.  

For a process that involves one product on a dedicated process line, solving the measure, analysis, and control problem could be done with one set of sensors, one analysis tool, and one data source. The reality is that most facilities are not dedicated sites, and each process line may produce many products. In this environment, one of the fundamental challenges with implementing advanced measurement, analysis, and optimization is that many different and disparate data sources can arise. On the path to realizing PAT's potential, industry leaders have proposed structured architectures for implementing PAT tools and concepts within manufacturing environments. Integrating PAT into the architecture of a plant provides a foundation for continuous improvement in operations, FDA's fourth PAT tool. 

To become more competitive and profitable in the global economy, life-sciences companies have had to restructure their operations to cut costs and improve operational efficiencies. Many have embraced FDA initiatives such as PAT that promote advances in automation control and information technology to make process-improvement efficiencies possible and reduce business risk.  

The potential benefits of PAT for the life-sciences industry more than offset the immediate challenges in measuring, analyzing, and controlling a company's processes. By leveraging the successes achieved in other industries where quality is paramount and margins drive efficiency, a company may find an unexpected source of new solutions that apply to life sciences in the new world of PAT.  

 is the life-sciences industry marketing lead at Rockwell Automation, 1201 South Second St., Milwaukee, WI 53204-2496, tel. 414.382.4329

Keywords: automation, control, manufacturing, process analytical technology

This article examines what other industries have done to address the challenges associated with successfully implementing the principles and concepts of process analytical technology, particularly the specific elements of measurement, analysis, and control. The author explores how other industries' approaches can be applied directly to pharmaceutical manufacturing.

Leveraging PAT: What Life-Sciences Companies Can Learn from Other Industries

Global competition in the pharmaceutical industry and increasing demands by governments and citizens for affordable medicines have focused the industry's attention on manufacturing efficiency. In this new era, improvements in process and product development approaches and the streamlining of manufacturing operations can have a profound effect on the bottom line. Process simulation and scheduling tools can play an important role in this endeavor. The function of such tools in the development and manufacture of active pharmaceutical ingredients (APIs) has been reviewed (1–8). This article focuses on the role of process simulation and scheduling tools in the development and manufacture of pharmaceutical products.  

During process development, process simulation software is used to perform the following tasks: 					

represent the entire process on the computer;

calculate demand for utilities as a function of time;

assess the environmental impact and so forth.

The availability of a good computer-based model improves the understanding of the entire process by team members and facilitates communication. What-if and sensitivity analyses are greatly facilitated by such tools. The objective of such studies is to evaluate the effect of critical parameters on various key performance indicators (KPIs) such as production cost, cycle times, and plant throughput. If there is uncertainty for particular input parameters, sensitivity analysis can be supplemented with Monte Carlo simulation to quantify the effect of uncertainty. Cost analysis, especially capital cost estimation, facilitates decisions related to in-house manufacturing 

 outsourcing. An estimation of the cost-of-goods identifies expensive processing steps and such information guides R&D work in a judicious manner.  

When a process is ready to move from development to manufacture, process simulation facilitates technology transfer and process fitting. A detailed computer model provides a thorough description of a process in a format that can be readily understood and adjusted by the recipients. Process adjustments are commonly required when a new process is moved into an existing facility that does not contain equipment that is ideally sized for the new process. The simulation model is used to adjust batch sizes, determine the cycling of certain steps (for equipment that cannot handle a batch in one cycle), estimate recipe cycle times, and so forth.  

Production scheduling tools play an important role in large- and clinical-scale manufacturing alike. They are used to generate production schedules on an on-going basis and in a manner that does not violate constraints related to the limited availability of equipment, labor resources, utilities, or inventories of materials. Production-scheduling tools close the gap among enterprise resource planning (ERP) and materials requirements planning (MRP) II tools and the plant floor (9). Production schedules generated by ERP/MRP II tools typically are based on coarse process representations and approximate plant capacities. As a result, solutions generated by these tools may not be feasible, especially for multiproduct facilities that operate at high capacities. This situation often leads to late orders that require expediting or to large inventories to maintain customer responsiveness. Lean manufacturing principles such as just-in-time production, low work-in-progress (WIP), and low product inventories cannot be implemented without good production scheduling tools that can accurately estimate capacity.  

Commercially available simulation and scheduling tools

Process simulation programs, also known as process simulators, have been in use in the chemical and petrochemical industries since the early 1960s. Established simulators for those industries include Aspen Plus and HYSYS (Aspen Technology, Inc., Cambridge, MA), ChemCAD (Chemstations, Inc., Houston, TX), and PRO/II (SimSci-Esscor, Inc., Lake Forest, CA).  

These simulators have been designed to model primarily continuous processes and their transient behavior for process control purposes. Most pharmaceutical products, however, are produced in batch and semicontinuous mode. Such processes are best modeled with batch process simulators that account for time-dependency and sequencing of events. The first simulator specific to batch processes was Batches from Batch Process Technologies, Inc. (West Lafayette, IN). It was commercialized in the mid-1980s. All of its operation models are dynamic and simulation always involves the integration of differential equations over a period of time. In the mid-1990s, Aspen Technology introduced Batch Plus, a recipe-driven simulator that targeted batch pharmaceutical processes. Around the same time, Intelligen, Inc. (Scotch Plains, NJ) introduced SuperPro Designer. The initial focus of SuperPro was on the production of APIs (synthetic and biosynthetic) and specialty chemicals. Over the years, its scope has been extended to include process modeling for the production of pharmaceutical and consumer products.  

Discrete-event simulators also have found applications in the pharmaceutical industries, especially in the modeling and debottlenecking of packaging operations. Established tools of this type include ProModel (ProModel Corporation, Orem, UT), Arena and Witness (Rockwell Automation, Inc., Milwaukee, WI), Extend (Imagine That, Inc., San Jose, CA), and FlexSim (FlexSim Software Products, Inc., Orem, UT). The focus of models developed with such tools is usually on the minute-by-minute time-dependency of events and the animation of the process. Material balances, equipment sizing, and cost analysis tasks are usually out of the scope of such models. Some of these tools are quite customizable and third-party companies occasionally use them as platforms to create industry-specific modules. For instance, BioPharm Services, Ltd. (Bucks, UK) has created a module with emphasis on biopharmaceutical processes that runs on top of Extend.  

Microsoft's MS Excel is another common platform for creating models for pharmaceutical processes that focus on material balances, equipment sizing, and cost analysis. Some companies have even developed models in Excel that capture the time-dependency of batch processes. This is typically done by writing extensive code (

macros and subroutines) in Visual Basic for Applications that comes with Excel. K-TOPS from Alfa Laval Biokinetics, Inc. (Philadelphia, PA) belongs to this category.  

In terms of production scheduling, established tools include Optiflex (i2 Technologies, Inc., Irving, TX), SAP APO (SAP AG, Walldorf, Germany), ILOG Plant PowerOps (ILOG SA, Gentilly, France), and Aspen SCM (formerly Aspen MIMI) (Aspen Technology, Inc., Cambridge, MA). Their success in the pharmaceutical industry, however, has been rather limited so far. Their primary focus on discrete manufacturing (as opposed to batch chemical manufacturing) and their approach to scheduling from a mathematical optimization viewpoint are some reasons for the limited market penetration.  

SchedulePro (Intelligen, Inc.) is a new finite-capacity scheduling tool that focuses on the scheduling of batch and semicontinuous chemical and related processes. It is a recipe-driven tool with emphasis on the generation of feasible solutions that can be readily improved by a user in an interactive manner.  

Examples that illustrate the benefits from the use of simulation and scheduling tools in the production of pharmaceutical products follow.  

Modeling and analysis of a tablet manufacturing process

Common forms of pharmaceutical products include tablets, capsules, ointments, creams, and solutions in syringes and vials. Their preparation involves mixing the API with various excipients that increase the product's shelf life and facilitate the delivery of the API. The manufacturing of injectable products involves the filling of syringes or vials under aseptic conditions.  

A tablet manufacturing process will be used as a representative example to demonstrate the use of process simulation and scheduling tools in the development and manufacture of finished pharmaceutical products. To model an integrated process using SuperPro Designer, a user starts by developing a flowsheet that represents the overall process. Figure 1, for instance, displays part of the flowsheet of a tablet manufacturing process. The flowsheet is developed by putting together the required unit procedures and joining them with material flow streams. Next, the user initializes the flowsheet by registering the various materials that are used in the process and specifying operating conditions and performance parameters for the various operations.  

 Figure 1: A flowsheet for a pharmaceutical tablet manufacturing process.			

Most pharmaceutical processes operate in batch or semi-continuous mode. In contrast, petrochemical and other industries that handle large throughputs use continuous processes. In 

, a piece of equipment performs the same action all the time. In 

, on the other hand, a piece of equipment goes through a cycle of operations. For instance, a slurry preparation step (P-1 in V-101) includes the following operations (see Figure 2): sanitize, charge USP water, charge sucrose, charge API, agitate, transfer to mill recirculation vessel, and flush (clean equipment). 

Figure 2: The operations associated with the first unit procedure of Figure 1.			

 In the SuperPro system, the set of operations that make up a processing step is called a unit procedure (as opposed to a unit operation). The individual tasks contained in a procedure (

charge, heat, agitate) are called operations. A unit procedure is represented on the screen with a single equipment-looking icon. In essence, a 

 is the recipe that describes the sequence of actions required to complete a single processing step. Figure 2 displays the dialog through which the recipe of a vessel unit procedure is specified. On the left-hand side of that dialog, the program displays the operations that are available in a vessel procedure; on the right-hand side, it displays the registered operations. The significance of the unit procedure is that it enables a user to describe and model the various activities of batch processing steps in detail. 

For every operation within a unit procedure, the simulator includes a mathematical model that performs material and energy balance calculations. Based on the material balances, it performs equipment-sizing calculations. If multiple operations within a unit procedure dictate different sizes for a certain piece of equipment, then the software reconciles the different demands and selects an equipment size that is appropriate for all operations. The equipment is sized large enough so that it will not be overfilled during any operation, but no larger than necessary (to minimize capital costs). If the equipment size is specified by the user, then the simulator ensures that the vessel is not overfilled. In addition, the tool checks to ensure that the vessel contents will not fall below a user-specified minimum volume (

 a minimum stir volume) for applicable operations. 

In addition to material balances, equipment sizing, and cycle-time analysis, the simulator can be used to carry out cost-of-goods analysis and project economic evaluation. The sections that follow provide illustrative examples of the previously described operations. 

Having developed a good model using a process simulator, the user may begin experimenting on the computer with alternative process setups and operating conditions. This process may reduce the costly and time-consuming laboratory and pilot-plant effort. Of course, the garbage-in, garbage-out principle applies to all computer models. If critical assumptions and input data are incorrect, so will be the outcome of the simulation. Consequently, a certain level of model verification is necessary. In its simplest form, a review of the results by an experienced engineer can play the role of verification. 

 The objective of this example is to illustrate how batch process simulators can be used to model, visualize, and analyze fine pharmaceutical processes. It focuses on a process for manufacturing pharmaceutical tablets in an existing facility for which equipment sizes are known. Such tools, however, also can be used to size equipment during the design of new facilities.   

The entire flowsheet of the process is shown in Figure 1. A batch begins by charging 705 L of USP water into a 1500-L mixing tank (V-101), then adding 200 kg of sucrose and 400 kg of API. The suspension is agitated thoroughly for 8 h. Then, the suspension is transferred into another 1500-L tank (V-102) that feeds the nanomill (NM-101). The role of the nanomill is to homogenize the suspension thoroughly and reduce the API particles to nanometer scale. This step is required because this specific API is insoluble in water. The suspension is pumped through the nanomill twice during a 22-h period. After homogenization is completed, the material is transferred into a 2000-L tank (V-103) where 70 kg of an excipient and 50 kg of a flavoring agent are added along with 100 L of USP water. The material is agitated thoroughly for 5 h. Then, the material is transferred into another 2000-L tank (V-105) that feeds the granulator. A stabilizer solution is prepared in a 500-L tank (V-104) by dissolving 80 kg of the stabilizer into 180 L of USP water. The stabilizer solution is combined with the homogenized solution in V-105.  

In preparation for the granulation–drying step, mannitol is added into the granulator bowl (GRN-101). Then, the suspension is sprayed into the granulator's chamber at a rate of 120 kg/h. Almost all of the water is removed (final water content is 0.005% w/w). The granulated and dried material is removed from the granulator and stored into several 50-L mobile containers (MC-101). The granulator handles a batch of homogenized material in two cycles because its limited bowl volume can hold only ;1350 L (or 600 kg) of bulk solids. 

The mobile tanks are moved into the tablet press room. The tablet press (TBP-101) makes 0.5-g tablets at a rate of 250,000 tablets/h. The processing of a batch is completed in approximately 9.6 h. The tablets are collected in a storage bin (DB-101).  

Then, a tablet coater (TB-101) coats the tablets with a material that gives them a sweet taste and a blue color. A batch is processed in four cycles because the coater can handle as much as ;300 kg of tablets per batch. The coating solution is prepared ahead of time in a 150-L mixing tank (V-106). Approximately 10 mg of coating solution is required per tablet. A coating cycle takes about 6 h. Warm air is pumped through the drum of the coater during the coating process to vaporize the water of the coating solution. The coated tablets are stored in drums (not shown in the flowsheet) and taken to the packaging area. 

All the unit procedures up to the granulator (P-7/GRN-101) require a sanitization operation before the main processing activities and a clean-in-place (CIP) operation after the main processing activities.  

Table I provides information about raw-material requirements for the entire process. A batch consists of approximately 2.4 million product tablets that have a total mass of around 1224 kg. The API is approximately 33% of the final tablet mass. Notice the large amount of hot USP water, which is consumed for equipment cleaning.  

Process scheduling and resource tracking. 

Figure 3 displays the equipment occupancy chart for four consecutive batches (each color represents a different batch). The recipe-scheduling summary dialog is shown on the top right hand corner of the chart. The recipe batch time is approximately 102 h, which is the total time between the start of the batch's first step and the end of the batch's last step. Because most of the equipment items are used for shorter periods within a batch, however, a new batch is initiated every 30 h, which is known as the recipe cycle time. The minimum possible recipe cycle time is 29.6 h. Multiple bars of the same color on the same line represent reuse (sharing) of equipment by multiple procedures or operations. The single CIP skid (top line in Figure 3) is the only shared equipment in this process. White space between procedure bars represents idle time. White space within a procedure bar represents waiting time. For instance, the white spaces in the procedure bars of V-103 and V-104 represent waiting for cleaning because of the constraints imposed by the single CIP skid. This type of chart is an invaluable tool for visualizing cycle times and scheduling bottlenecks. 

Figure 3: Equipment occupancy chart for four consecutive batches.			

Figure 4 displays the operations Gantt chart which provides more detailed scheduling information. Notice, for instance, the duration of the TRANSFER-OUT-1 operation in P-6 (V-105) that feeds material to the granulator (P-7/GRN-101). The granulator handles a batch in two cycles because of its limited bowl volume. The duration of TRANSFER-OUT-1 in P-6 is from the start of the first granulation operation in P-7 to the end of the second granulation operation in P-7. In reality, there are two shorter transfer-out operations (synchronized to the two granulation operations in P-7), but for simplicity, SuperPro represents that with a single transfer-out operation that has the correct total duration.  

Process scheduling in the context of a simulator is fully process driven and the effect of changes can be analyzed in a matter of seconds. For instance, the effect of an increase in batch size (that affects the duration of charge, milling, granulation, and other scale-dependent operations) on the recipe cycle time and the number of batches can be evaluated instantly. Because of many interacting factors involved with even a relatively simple process, simulation tools that allow users to describe their processes in detail and to perform what-if analyses quickly can be extremely useful.  

Another characteristic of batch processing is the variable demand for resources (

labor, utilities, and raw materials) as a function of time. For instance, Figure 5 displays the demand for purified water (USP water) for eight consecutive batches. This demand includes USP water consumed for the CIP operations as well as USP water used for preparing the product mixture. The red lines represent the instantaneous demand whereas the green line represents the cumulative demand and corresponds to the 

-axis on the right-hand side. The blue line corresponds to daily demand (the averaging period can be adjusted by a user). High-purity water is a common potential bottleneck in pharmaceutical processes. It is frequently used by multiple processing steps simultaneously in activities such as solution preparation and equipment cleaning. If not enough instantaneous (or cumulative) capacity is available, one or more process steps may be delayed, possibly with severe consequences. During the design of new facilities or retrofit of existing ones, the information provided by such charts is used for sizing utility systems.  

Figure 5: USP water demand for eight consecutive batches.			

In addition to the instantaneous demand of resources, the simulator provides the means to track the volumetric usage of all vessels throughout the batch cycle. This allows a user to track maximum working volumes over time and ensure that the minimum stir volume is always met at any relevant point in a process. The volume content of vessels is also used in sizing new vessels and calculating the capacity usage of existing vessels. 

Cost analysis and project economic evaluation are important for several reasons. For a new product, if the company lacks a suitable manufacturing facility with available capacity, it must decide whether to build a new plant or outsource the production. Building a new plant is a major capital expenditure and a lengthy process. To make a decision, management must have information about the capital investment required and time to complete the facility. When production is outsourced, a cost-of-goods analysis serves as a basis for negotiation with contract manufacturers. A sufficiently detailed computer model can be used as the basis for the discussion and negotiation of the terms. Contract manufacturers usually base their estimates on the requirements of equipment use and labor per batch, which is information provided by a good model. SuperPro performs thorough cost analysis and project economic evaluation calculations. It estimates capital as well as operating cost. The cost of equipment is estimated using built-in cost correlations that are based on data derived from several vendors and literature sources. The fixed capital investment is estimated based on equipment cost and using various multipliers, some of which are equipment specific (

installation cost) and others are process specific (

cost of piping, buildings). The approach is described in detail in the literature (7). The rest of this section provides a summary of the cost analysis results for this example process.  

Table II shows the key economic evaluation results for this project. The fixed capital investment for a manufacturing facility of this size is nearly $145 million. The annual operating cost (assuming the facility is dedicated to a single product) is nearly $87 million, resulting in a unit production cost of $0.14/tablet. Assuming a product selling price of $0.25/tablet, the facility generates annual revenues of $160 million and has an attractive return on investment. 

Table II: Key economic evaluation results.			

Table III provides a breakdown of the manufacturing cost, including (two left columns) and excluding (two right columns) the cost of API. A purchasing price of $500/kg was assumed for the API, resulting in an annual API cost of $52.2 million. When the cost of the API is considered in the cost-of-goods, the cost of raw materials becomes the dominant cost (63% of total). If the cost of API is ignored, the facility overhead becomes the dominant item of the manufacturing cost (75% of total). In the latter case, the unit production cost drops to $0.05/tablet. That is representative of the incremental cost of contract manufacturers.  

This article discusses the role of process simulation and scheduling tools in the development and manufacture of pharmaceutical products.

The Role of Process Simulation and Scheduling Tools in the Development and Manufacture of Pharmaceutical Products

Process analytical technology (PAT), according to the US Food and Drug Administration's Web site (

www.fda.gov/cder/OPS/PAT.htm#Introduction

a system for designing, analyzing, and controlling manufacturing through timely measurements (

 during processing) of critical quality and performance attributes of raw and in-process materials and processes with the goal of ensuring final product quality. It is important to note that the term analytical in PAT is viewed broadly to include chemical, physical, microbiological, mathematical, and risk analysis conducted in an integrated manner. 

Many tools are available for PAT. This article focuses on the knowledge discovery available from the creation and operation of a virtual plant environment. Such an environment can involve the synergistic use of model predictive control, first-principal models, neural networks, and principal component analysis, as described in 

New Directions in Bioprocess Modeling and Control 

The virtual plant is a relatively new concept that should not be confused with existing simulation methods for process design, configuration checkout, and operating training systems. Most existing batch-process simulations are off-line and noninteractive, and most real-time dynamic-process simulations were designed originally for continuous processes. These real-time process simulations can develop severe numerical errors or even fail for the extreme conditions of batch operations. They also may require interfaces for communication of input–output, inventory controls, and coordination with the control system of speedup, slowdown, pause, and resume. The control-system engineer is probably most familiar with tieback simulations because these have been used predominantly for configuration checkout and operating training systems. The process response in these tiebacks is mimicked by the trial-and-error adjustment of ramp rates triggered by the opening and closing of valves or the turning on or off of pumps. 

The first key feature that distinguishes a virtual plant from process simulators is its ability to use the actual configuration, historian, displays, and advanced-control tool set of the real plant without translation, emulation, special interfaces, or custom modifications. The configuration database from the real plant can be exported and then imported and downloaded into a personal computer or a control-system computer just as if it were an actual hardware controller. Files for operator graphics, process-history charts, and data history from the real plant can be copied to the computer for the virtual plant so that the user has the entire control system of the real plant on a computer (see Figure 1).  

 Figure 1: Virtual plant with imported configuration from actual plant and embedded advanced control tools and process simulation.			

Most dynamic, high-fidelity, process-simulation software allows users to build a basic control strategy or sequence inside the simulation environment. Nonetheless, simulation developers tend to have a process rather than a control background and focus. It is unrealistic to expect the process and batch-control capability offered by simulation software to be in the same realm as the control capability of a distributed control system's (DCS) software developed from hundreds of years of experience by process-control experts. The overall control functionality in process simulators is primitive compared with the capabilities offered in the modern DCS, which offers capabilities such as sequential function charts and basic function blocks, a batch manager, and advanced control tools such as multivariable model predictive control. These are almost nonexistent in simulators. Duplicating even a simplified version of a control system in a dynamic simulation is a large effort. At best, a user may end up with two control systems with no assurance of how well they match up and with no way to automatically manage changes between them.  

Consequently, most simulation software now offers a standard or custom open-process control interface. For the simulations with material balances to run independently of the DCS for development and testing, however, the simulations still must have internal pressure and level loops set up to prevent volumes from running dry, overflowing, overpressurizing, or developing flow reversals from pressure-gradient reversals that can lead to fatal numerical errors. Tables must be mapped that transfer control from these internal simulation loops to the DCS loops and initialize the proper controlled variables, set points, and manipulated variables. DCS loops that do not have a counterpart in the simulation still must have their controller outputs initialized. Using standard blocks for split-ranged control, velocity limiting, signal characterization, and signal selection makes proper initialization problematic for external simulations. 

The virtual plant also can simultaneously stop and start the execution and restore and replay simulated conditions of all control and simulation modules. Emerging is the ability to replay actual plant data-history files at high speeds for adaptation and testing of the process modules without a connection to the actual plant. 

Most simulations used for control-system checkout and operator training become obsolete after startup. The investment is lost. The virtual plant offers a better chance of keeping the control system up-to-date simply by importing the most current configuration and enabling the simulation to match process changes better by a nonintrusive automatic adaptation when running the virtual plant in real-time, on-line, and in a read-only mode. An innovative use of model predictive control has been demonstrated to adjust process model parameters (manipulated variables) automatically to provide a better match between key process variables in the actual plant (set points) and in the virtual plant (controlled variables). 

In a virtual plant, everything is done in the same configuration environment used for the actual control system. The focus can be more on the application than on learning the inevitable undocumented features and tips and techniques associated with any new simulation software and interface. In summary, a virtual plant offers the following advantages: 					

The control system and graphics do not need to be duplicated, emulated, or translated.

Special data interfaces and tables and initialization issues are avoided.

All batch, basic, and advanced control tools can be tested readily.

Controls and simulation can run in unison at the same real-time multiplier.

Controls and simulation scenarios can be saved, restored, and played back.

Actual plant data can be played back at high speeds for testing and adaptation.

Simulations can handle extreme conditions of batch operations and failures.

Simulations can incorporate dynamics important for tuning and performance.

Controls and simulation can stay up-to-date and have a longer life cycle.

Engineers can work in the same environment and focus on the application.

The most familiar use of a virtual plant is for testing and training. For the checkout of batch sequences and the training of operators, it is important to be able to simulate batch phases repetitively and rapidly. The ability to stop, start, save, restore, and replay scenarios and record operator actions is critical. For first-pass testing and familiarization of sequences and graphics, an automated tieback simulation may be sufficient, but to test and learn the interaction and performance of control strategies and the process, the higher fidelity of dynamics offered by process modules is important. It opens the door for upgrading the process and control skills of the technology, maintenance, and configuration engineers who will support the operations. 

Before the configuration starts at the beginning of a project, the process modules can be used to evaluate control strategies and advance control tools. In the past, this was achieved with off-line dynamic simulations. Today, ready access to an industrial tool set for basic and advanced control and simulations adapted to benchtop or pilot-plant runs offers the opportunity for rapid prototyping that can lead to control definitions with better detail and potential performance. Benchtop or pilot-plant systems with a miniature version of the industrial DCS that greatly facilitate the development and scale-up of the control system now are available (2). Benchtop systems with all the functionality of the main manufacturing systems are scarce, however, because the expertise (and, more important, the interest) to configure, maintain, and engineer these systems is not traditionally in the development groups of these types of companies. 

The virtual plant can be demonstrated with industrial-batch, basic, and advanced control systems used in a benchtop or pilot-plant system. The process modules can be adapted by means of a connection to a benchtop or pilot-plant system or by high-speed playback of process data from experimental runs. An opportune time to take advantage of the PAT initiative is during the research and process-development phase. The use of benchtop and pilot-plant systems with an industrial DCS can improve project schedule, cost, and effectiveness because it allows process control to be designed into the plant at an early stage and puts the biochemist, process technology, and configuration engineer on the same page (2). Nonetheless, the best return on investment for PAT is realized by the eventual implementation of advanced process analysis and control in large-scale manufacturing processes. 

A virtual plant allows manufacturers to try out and tune innovative strategies such as effective switching of the controller output for set-point response optimization. Advanced control tools such as adaptive control, autotuning, model predictive control (MPC), neural networks (NN), principal component analysis (PCA), and partial least squares or projection to latent structure (PLS) can be demonstrated, adjusted, and evaluated faster than in real time. 

Although actual plant operation is the best source of data, the long batch-cycle time and the desire to minimize disruptions from the introduction of perturbations severely restricts the amount of useful plant data for the development of MPC, NN, and PCA. As a general rule, five changes are needed for each process input to develop an experimental model of a process output. For MPC, this corresponds to a minimum of five step changes in each process input, at least one of which is held long enough for the process output to reach a steady condition. For NN and PCA, it means a minimum of five batches per process input in which the respective process input differs from the normal value. Developing a PCA with four inputs for the detection of an abnormal batch requires at least 20 batches with varying inputs and at least five batches with normal inputs. If the batch-cycle time is about two weeks, it would take approximately a year of plant production to have enough data. If the spectrum of variability in the inputs cannot be optimized, then it may take several years of production runs. 

It is important to keep in mind that process control transfers variability from a controlled variable to a manipulated variable. When a loop for tight control is tuned, a straight line is generated for the controlled variable, but fluctuations appear in the manipulated variable on a long-term trend. Process control does not make variability disappear. One cannot keep both a controlled variable and manipulated variable constant. In advanced control, variability is transferred from a higher-level process variable (

product-formation rate) to a lower-level process variable (

substrate flow or concentration). The level of control and transfer of variability has a profound effect on the choice of variables for NN and multivariable statistical process control.  

The profiles to be analyzed depend upon the level of control. When proportional-integral-derivative (PID) loops are added and put in automatic, the transfer of variability by feedback control means that the batch profile is seen in the controller output rather than in the controller input. For example, if dissolved oxygen is not controlled, then the dissolved oxygen shows the batch profile, and the difference between the measured and reference dissolved-oxygen profile is the PCA input. If a PID loop for dissolved-oxygen control is implemented and put in automatic to manipulate airflow, then the dissolved oxygen draws a relatively straight horizontal line (

 the process variable stays relatively close to a constant set point and thus makes a straight horizontal line on the trend plot) and airflow now shows the batch profile and is used for the PCA. If the dissolved-oxygen set point is not constant but is manipulated by MPC to control product formation rate, both the dissolved-oxygen set point and airflow show the batch profile. 

Perturbations can be automated and introduced to a virtual plant running faster than real time so that in a couple of weeks there are enough data for the identification of models for MPC, the training of NN, the development of latent variables and discriminant analysis for PCA, and to predict economic variables using PLS. The predictive ability of MPC, NN, PCA, and PLS then can be verified and evaluated by the high-speed playback of previous plant batches.  

Conventional PCA assumes that all process inputs, other than those used for the PCA, are fixed. A virtual plant running in real time and synchronized with the actual plant can predict the effect of variations in other process inputs by using the model-based and super model-based PCA algorithms. Figure 2 shows the setup of super model-based control (3). In this setup, a dynamic-error model implementing techniques such as dynamic PLS is used to predict the structured errors from an imperfect model. These structured errors are subtracted from the differences in the profiles between the actual plant and the virtual plant before being sent to the dynamic PCA for batch fault detection and PLS for fault prediction. 

Figure 2: Super model-based dynamic principal component analysis and partial least squares or projection to latent structures for fault detection and prediction. Abbreviations: PCA is principal component analysis and PLS is projected to latent structure.			

With a virtual plant, manufacturers can explore more optimal operating conditions and investigate "what-if" scenarios. These scenarios are important to identify the cause of an abnormal batch. To date, PCA for batch fault detection only identifies a batch as abnormal; logic must be added to diagnose the fault. By creating scenarios, a virtual plant can help develop these rule sets off-line faster than real time and can evaluate the rule sets by the high-speed playback of previous batches. An on-line virtual plant synchronized with the real plant can be sped up and run to batch completion to predict and analyze abnormal situations on the basis of current batch conditions. A virtual plant also can help create the predictive capability of the PLS "

 space" of economic variables from the PCA "

A virtual plant can provide inferential real-time measurements of important process outputs. The built-in material balances, in conjunction with kinetic models, can predict biomass and product concentrations and the slope of batch profiles such as biomass growth rate and product-formation rate. Oxygen-uptake rates can be used to predict metabolic rates. These predictions then are delayed so that the values are synchronized with analysis on-line, at-line, or in a laboratory. The prediction is shifted by a bias that is a fraction of the difference between the inferential and actual measurement in a fashion similar to what is done in the feedback correction of an NN for property estimation (4). The inferred growth, production, or metabolic rates can be controlled variables for optimization by MPC that manipulates substrate and dissolved oxygen or dissolved carbon dioxide concentrations. Figure 3 shows the general setup for the adaptation of the virtual plant process-model parameters by an MPC and the optimization of the batch profiles by another MPC using inferential measurements of growth and production rates from a virtual plant.  

Figure 3: Setup of a virtual plant and use of model predictive control for adaptation and optimization.			

The biomass-growth and product-formation rates for mammalian cell cultures are so slow that the signal-to-noise ratio is often too low to allow calculation of these rates from the differences between concentration measurements. An inferential measurement can have significantly less noise and more sensitivity. It is important to remember, however, to correct the inferential measurement by a change in analysis only over a sufficient time interval (

 24 hours) so that the true change in concentration is larger than the scatter in the analysis, as determined by the repeatability of the analyzer.  

A virtual plant can be run faster than real time to batch completion, thereby providing an online prediction of key performance indicators such as batch-cycle time and yield based on current batch conditions and inferential measurements. The virtual plant can accelerate the benefits from PAT by offering the ability to use process and endpoint monitoring and control, continuous improvement, and knowledge-management tools in an integrated and accelerated manner.  

In summary, the potential applications of a virtual plant are as follows:  					

testing of configuration and process interactions;

process and control education of operators, technicians, and engineers;

virtual experimentation for exploration of optimums and "what-if" scenarios;

rapid prototyping of innovative control strategies and advanced controls;

development of latent variables and reference trajectories for PCA;

development of logic for fault analysis by PCA;

online prediction of abnormal situations;

inferential real-time measurements of important concentrations;

Modeling conducted in the process development and early commercialization stages can increase process efficiency and lead to process-control improvement. A virtual plant environment, running on a PC, can incorporate the same industrial control systems and configuration expertise used in manufacturing and helps many parts of process analytical technology. It also can make possible the enlightened use of technologies such as model predictive control, first-principal models, neural networks, and multivariate statistical process control.

Pharmaceutical Technology-11-01-2006

Advanced Neural Computing Software Systems: Data Mining in Processing and Formulation

Leveraging PAT: What Life-Sciences Companies Can Learn from Other Industries

The Role of Process Simulation and Scheduling Tools in the Development and Manufacture of Pharmaceutical Products

Maximizing PAT Benefits from Bioprocess Modeling and Control