Improvement of Current Pharmaceutical Manufacturing Systems Utilizing Practices from Outside Industries

Current Good Manufacturing Practices (cGMP' s) were developed to keep the public safe and ensure quality products. Technological evolution of the industry resulted in cGMP requirements which could not keep pace, and ultimately had a restrictive effect on daily operations and process improvements for pharmaceutical manufacturers. In response, the FDA has recently released new Guidances for Industry, PAT-A Framework for Innovative Pharmaceutical Development, Manufacturing, and Quality Assurance and Pharmaceutical cGMP's for the 21st Century-A Risk-Based Approach, to encourage innovation in the pharmaceutical industry. The current focus is to understand the product, the manufacturing process and operations rather than the previous focus on product testing and release. Other industries have developed innovative approaches to competitive improvement and it is important to learn from their successes and failures . The aim of this study is to identify critical areas for improvement in current pharmaceutical manufacturing practices and to explore solutions to these problems using principles found outside of the pharmaceutical industry. A facility producing pharmaceutical products under cGMP's will be evaluated to gain a "baseline" understanding of current manufacturing practices. Critical and problematic areas will be identified as well as potential opportunities to incorporate external industry practices, with a focus on the Toyota Production System®, to improve the manufacturing process. A ProModel® simulation was developed in as a proof of concept to evaluate proposed improvements to the system. This simulation was run with a number of theoretical changes to the scenario to evaluate the effects and interactions of the proposed improvements. These areas of improvements include: Shrinking traditional batching rules to create a more semi continuous production system, cross training personnel on equipment and responsibilities, and the addition of parallel machines at the equipment bottleneck. Results from the simulation experimentations indicated that significant improvements to product throughput time can be achieved. While all three individual factors studied were found to be significant, the greatest gains were achieved using a combination of batch rule changes and cross training of personnel. Changing the batching rules was found to have the greatest effect on reducing throughput times, due to the drastic reduction of time spent waiting to batch in the system. Cross training resulted in increased capacity in the bottleneck machine with availability during all hours of operation. While parallel machining did achieve improved throughput times when compared to the current state of the system, the additional machine was not justified when compared with the scenario yielding superior results. Lean manufacturing and other alternative techniques should be explored in the pharmaceutical industry to improve current systems and utilized early in the development of new systems. This case study represents a number of possible opportunities for specific areas of improvement as well as suggesting an overall change in manufacturing mindset. The pharmaceutical industry can learn a great deal from outside industries where competitiveness is required to ensure solvency. Acknowledgements I would like to thank my major professor Dr. Valerie Maier-Speredelozzi for all of her help in my studies and research in the Industrial and Manufacturing Engineering department. Without her guidance my research and development as an industrial engineer would not have been possible. I would also like to thank my longtime advisor from the Biomedical and Pharmaceutical Sciences department, Dr. Thomas Needham, for his pharmaceutical expertise and advice. A special thanks to the owners of XYZ Pharma for allowing me free reign over their company and their cooperation to gather information for this study. I would like to thank my parents for supporting all of my academic endeavors, no matter how ridiculous they may seem.

theoretical changes to the scenario to evaluate the effects and interactions of the proposed improvements. These areas of improvements include: Shrinking traditional batching rules to create a more semi continuous production system, cross training personnel on equipment and responsibilities, and the addition of parallel machines at the equipment bottleneck.
Results from the simulation experimentations indicated that significant improvements to product throughput time can be achieved. While all three individual factors studied were found to be significant, the greatest gains were achieved using a combination of batch rule changes and cross training of personnel.
Changing the batching rules was found to have the greatest effect on reducing throughput times, due to the drastic reduction of time spent waiting to batch in the system. Cross training resulted in increased capacity in the bottleneck machine with availability during all hours of operation. While parallel machining did achieve improved throughput times when compared to the current state of the system, the additional machine was not justified when compared with the scenario yielding superior results.

Introduction
Chapter one aims to orient the reader with a brief history of both general manufacturing and pharmaceutical manufacturing. The history discussed is not intended as a comprehensive review of manufacturing, rather it will give the reader perspective with regard to the origins and ongoing evolution of the state of manufacturing. This chapter will also describe the general study objectives of this thesis.

Brief History of Manufacturing
In 1776 Adam Smith published "An Inquiry into the Nature and Causes of the Wealth of Nations" and introduced the concept of the division of labor. Smith felt that by dividing complex jobs which were previously performed by skilled artisans (carpenters, smiths, etc .. . ) into simplified tasks, greater levels of efficiency could be achieved. His experiments first assessed the production of metal pins by unskilled laborers using traditional craft methods, which required all aspects of manufacturing. He then compared these laborers to a group which had the manufacturing process divided into one or a few simple tasks. Each laborer was responsible for their individual area of pin manufacturing and he found a large increase in productivity. He attributed the increased output to three reasons: 1. An increase in the operator's dexterity due to simple and repetitive tasks. 2.
Minimization of travel time because the product moved and the people did not and 3. 1 Tools used in the manufacturing can be optimized based on the operator's superior knowledge of the process (Smith, 1776).
The introduction of interchangeable parts for the manufacturing of muskets by Honore Blanc in the late 1700's also signaled an important milestone in the evolution of manufacturing.
By creating interchangeable parts through the standardization of templates and equipment, he replaced artisan craftsman and enabled unskilled laborers to be trained with manufacturing tools to generate greater production numbers at lower costs. This new method was adopted in many other industries throughout the world and facilitated the rapid growth of mass production (Gillespie, 1998;Hopp, 2001).
In 1913, Henry Ford introduced a continuously moving assembly line which allowed workers to remain in their workstations to add parts to cars as they passed by. Parts were supplied to workers through conveyors and timed to keep workers well stocked with necessary components. This new method drastically reduced automobile assembly times, lowered vehicle costs, and propelled the Model T to a dominant share of the market at the time (The Henry Ford Museum, 2003). The expensive, single function machinery limited flexibility and required many buffers of supplies, workers and storage space in order to maintain a smooth operation.
Additionally, changing between products was an expensive and time consuming process due to system rigidity, and resulted in fewer options for consumers (Womack, 1990).
Ford's breakthrough changed the face of the automobile industry, and Toyoda Kiichiro of Toyota Motors, was eager to learn from the automotive giant. In 2 1933 Kiichiro stated, "We shall learn production techniques from the American method of mass production. But we will not copy it as is. We shall use our own research and creativity to develop a production method that suits our own country's situation" (Ohno, 1988). The Toyota Production System was founded on the principle of waste elimination, utilizing autonomation and just-in-time principles.
Under the guidance of Taiichi Ohno, this system evolved greatly from its inception to the present day concept of "World Class Manufacturing," with an overriding goal of continual and rapid improvement while maintaining flexibility (Schonberger, 1986). These concepts have been termed "lean production" (Womack, 1990) and have resulted in the commitment to improve industry flexibility and competitiveness.

Brief History of Pharmaceutical Manufacturing
In 1938, The Food and Drug Administration (FDA) enacted the Federal Food, Drug, and Cosmetic Act and subsequently developed the Code of Federal Regulations (CFR), which contained current Good Manufacturing Practices (cGMP's). These practices were intended to keep the public safe and ensure quality products (FDA, 2006a). The pharmaceutical industry developed rapidly in response to the great need for penicillin and medical supplies during World War II. Following the war, manufacturers began researching and developing new drugs to expand the industry. Drug discovery evolved over the next 50 years from traditionally random occurrences to a highly focused methodology. Screening of target receptors and functional chemical structures makes it possible to identify potential successful therapeutic opportunities from a large pool of candidates (Ratti, 2001 ). While drug discovery experienced rapid growth with relatively little regulatory oversight, the manufacturing areas of the industry were under the strict guidance of the FDA's cGMP's. These regulations resulted in a strong commitment to quality and industry adoption of Total Quality Management (TQM) (Ghobadian, 1995), Statistical Process Control (SPC), Six Sigma (Swarbick, 2006), and other quality programs.
Rapid technological evolution of the industry resulted in cGMP's which could not keep pace, and ultimately had a restrictive effect on daily operations and process improvements for pharmaceutical manufacturers. The industry fell behind other less regulated manufacturing industries, where innovative practices and procedures were critical for global competition.

Study Objectives
The aim of this study is to identify critical areas for improvement in current pharmaceutical manufacturing practices and to explore solutions to these problems using principles found outside of the pharmaceutical industry. The study will be divided into two phases: Current State: 1. The evaluation of a manufacturing facility currently producing pharmaceutical products under current Good Manufacturing Practices is to gain a "baseline" understanding of the state of the art of manufacturing.
Using this knowledge to develop an accurate model of the system will identify critical and problematic areas.

Study Significance
Currently, there is little publicly available information about improving processes through innovational approaches. The National Institute for Pharmaceutical Technology and Education (NIPTE) has identified that while manufacturing is a significant expense to industry, there is very little industry and federal research funding in this area as compared to drug discovery. The lack of interest in funding manufacturing and processing research can be attributed to both the pharmaceutical industry and the FDA. The pharmaceutical industry is focused on bringing their product to the market place as rapidly as possible prior to the expiration of the patent in order to recoup the large investment made during the development process. The FDA is a regulatory agency concerned with product safety and has fostered a long tradition of discouraging risk taking involved in technological changes. As a result, industry has little incentive to invest dollars geared at the advancement of product/process development and manufacturing technologies/practices. Additionally, the National Institute of Health (NIH) is focused on funding predominately discovery related basic research, while the National Science Foundation has focused on manufacturing industries that are critical to the economy but struggling or threatened (Couts, 2006). This research will serve as a model for pharmaceutical processing in order to encourage novel approaches to manufacturing, through specific and system wide solutions. Ideally this research will serve small to medium sized pharmaceutical manufacturers who are considering or implementing changes in their processes, where simply improving their current system will be inadequate to meet the future needs of manufacturing. Through experimental and statistical modeling, conclusions can be drawn about altering the traditional mentality of manufacturers to improve production. Additionally, because the pharmaceutical industry is so diverse, research must be conducted in all aspects of the field. While powder based processing (tablets, capsules, etc ... ) and biotechnology has received wide attention, smaller sectors have not been explored in depth. Contract manufacturing can be considered a specialty area of the pharmaceutical industry requiring flexible personnel, equipment and production capabilities. This research should also be considered by other regulated industries such as food, cosmetics, nutraceutical, etc. for viable methods for improvement.

t.5 Chapter Review
Manufacturing has evolved over the past two hundred years and has created different levels for operational standards throughout industries. It is important to understand how the pharmaceutical industry was born and its evolution, to better grasp the current state of pharmaceutical manufacturers. Identifying developmental differences in outside industries will encourage greater fundamental understanding of why pharmaceutical manufacturing is unique.

Tools Utilized by Other Industries
Chapter two will give the reader an overview of the current state of lean manufacturing and commonly used lean tools in order to familiarize the reader with methods used in outside industries. Value Stream Mapping® (VSM) is discussed as an important method to identify problematic areas. System configuration is addressed with regard to flexibility and responsiveness through parallel machining.

Current State of Lean Manufacturing
Toyota is the developer and leader in the use of lean manufacturing, which has been widely adopted with applicability to any industry. Prior to beginning improvement, there must be an understanding of the current system. In order to gain a baseline understanding of the current system, VSM has been used as a critical first step in the process. The process involves the investigation and recording of all activities involved with manufacturing of a product from raw materials to finished  "l. Product flow is the path(s) the product take through production, before being shipped to the customer. 2. Information flow is how information is shared and communicated during the production process. 3. Material flow deals with how incoming material is moved and replenished and in what quantities during production" (Hall, 2001 ).
To create an accurate depiction of the three flows in the map, there must be accurate data from the production floor. "Gemba" is defined by Liker (2004) as the "actual place" and is the first step used by Toyota® when solving a problem. The concept reinforces the need for first hand knowledge and challenges conventional management methods of system reports and computer analysis. While many tools can be utilized by personnel, it is important to have a firm grasp on the actual situation. In order to illustrate the benefits of a VSM, an example is given. The EPA collected the following pieces of data to create their current state map: • Cycle time: The average time a specific routing spends from the point of release at the beginning of the routing until it reaches an inventory point at the end of the routing (Hopp, 2001).
• Change Over Times: The total amount of time it takes to change machinery, tools, personnel, etc ... from the last unit of good production at normal line efficiencies to the first good production unit of another product at normal line efficiency (Henry, 2001).
• Total Lead Time: The EPA used throughput time, (composed of processing time, setup time, move time, plus waiting time) (Askin, 1993) plus delivery time to calculate the total time from the release of the product to the shop floor to the time of delivery.
• Number of operators at each station: Additional information about manpower requirements versus actual staffing.
• Inventory levels: The amount and locations of the three inventory types is considered. Raw materials are used to make components and sub assemblies, work-in-process are materials and components waiting to be completed, and finished goods are completed products.
• Uptime (Utilization): The fraction of time the system or machine is not idle for lack of parts, calculated by dividing the arrival rate of parts by the effective production rate of the machine. The effective production rate is the maximum average rate at which the workstation can process parts, while accounting for failures, setups and other pertinent factors over the planning period of interest (Hopp, 2001

Tools for Leaner Manufacturing
Lean manufacturing has been considered to capitalize on the advantages of both mass production and artisan crafting. Cross trained workers are defined as "workers trained over time to perform a variety of tasks within their work area" (Askin, 2002) and then often rotated throughout the positions. Utilizing cross trained employees with flexible autonomated processing equipment enables the production of a wide variety of products in high volumes. "Production is lean if it is accomplished with minimal waste due to unneeded operations, inefficient operations, or excessive buffering in operations" (Narasimhan, 2006). It is considered "lean" because it utilizes far fewer resources, such as worker effort, production/storage space, on site inventory, less equipment investments while striving to achieve defect free processing (Womack, 1990).

Waste
In order to improve any company's leanness, it is important to identify and address the seven forms of waste which were identified by Taiichi Ohno of Toyota® (Ohno, 1988): 1) Defects: Parts that are manufactured but do not meet specifications 2) Waiting: Operators/equipment that is starved for work due to a lack of materials, equipment, or pertinent information.
3) Motion: Non value added movement of personnel, inventory, and/or equipment.

4)
Over-Processing: Any extra operations that are not required to manufacture a product.
5) Over-Production: Making more product than the customer has ordered. 6) Inventory: Having excess raw materials, work-in-process, and finished goods.
13 7) Inefficiency: Using excess material to process; people wasting time, ideas, effort, and time, and under usage of capacity (Allen & Robinson, 2001).
The following lean techniques were developed in the pursuit of achieving zero wastes.

Autonomation
Autonomation (Jidoka) in Japanese has two distinct meanings: I) A change from a manual process to a machine process. 2) Automatic control of defects or automation with a human mind (Monden, 1998 Machinery stoppage requires immediate attention to repair and improve the system. If the machine can not be fixed in a set period of time, five minutes for example, then all members of the production team become involved with a repair solution because the entire production line is stopped. The cartoon below, figure 3, is a simple example but shows two scenarios, the top portion represents a traditional automated 14 machine which will produce whether the products are good or defective. In the lower portion, there is the addition of a stop button which the machine can activate itself in addition to a siren on top of the machine alerting the operator that there is a problem. Ohno said, "A problem early in the process always results in a defective product later in the process. This will stop the production line or change a plan whether you like it or not" (Ohno, 1988).  (Hirano, 1987)

Error-Proofing
The term "Baka-Y oka" (idiot proofing) was first used by Shigeo Shingo, and evolved into, "Poka-Yoka" (fool-proofing) (Shingo, 1985) and is now widely recognized as error-proofing. This technique places various checking devices on equipment and tools to remove the potential for error and ultimately the creation of 15 defects. A simple example can be seen in figure 4a, where an electrical socket could potentially be plugged into the incorrect charge. Figure 4a. Electrical connection before error-proofing (QS Consult, 2006) This error could cause an electrical short and would result in a defective connection and an unusable product. When error-proofing is implemented, the part is redesigned to prevent the erroneous connection from being made. Figure 4b below depicts how a simple modification can achieve this result and remove the potential for error from the operator or manufacturing equipment. Error-proofing is used extensively throughout the Toyota Production System m conjunction with autonomation to minimize the production and passage of defective parts to the next process and to the ultimate consumer (Ohno, 1988). 16

Just-In-Time
Just in time (JIT) systems require a holistic approach to ensure accurate production, ordering, and stock quantities to ensure that the right parts needed in an assembly are available at the exact time they are needed and only in the amount needed (Ohno, 1988). While this may seem intuitive, it can be difficult to create such a system for a complex product where there may be many components and subassemblies. Additionally, JIT systems are reliant on flexible internal and third party component suppliers to cooperate with variability in ordering and delivery agreements. This approach requires synchronization throughout the company and transition from traditional "push" systems to a more flexible "pull" system. Push systems utilize a central planning system to determine and dictate production schedules, based upon sales forecasts, past sales records and customer orders. The demand forecast is used by production control as the basis of how many components to purchase, how much raw material to buy, at what rate to produce and when to place orders (Ploss!, 1967). A pull system regulates materials and production through the withdrawal of only the necessary parts at the needed times from a previous (upstream) process which has only produced enough units to replace the parts that have been withdrawn. Company-wide JIT implementation results in the elimination of unnecessary inventory thereby removing the need for warehouse and storage space, while simultaneously decreasing inventory holding costs and improving the ratio of capital turnover (Monden, 1998). Organization wide implementation of JIT can be difficult, but research has shown the positive impacts of the system on other operational areas such as organization, human resources, quality management, information systems, technology development, and manufacturing strategy (Matsui, 2006). This indicates a synergistic impact on a highly linked system and an overall contribution to competitive performance improvements.

5 Kanban
Kanban (sign board) is a communication tool to convey information about picking up or receiving the production order (Ohno, 1988). Kanbans vary in form (cards, bins, electronic signs, etc . ... ) but they all circulate between the processes to control the amount of production to ensure that only what is needed is produced.
Withdrawal Kanbans specify the type and amount of product required for the next process to withdraw from the previous process, and a production-ordering Kanban specifies the kind and quantity of product which the preceding process must produce (Monden, 1998). This system is vital for a pull system to maintain just-in-time manufacturing and prevent overproduction or underproduction of necessary components. Figure 6 is a simplified example of the use of Kanbans to communicate both withdrawal (retrieval) and production in the Toyota Production System®.  While the Kanban system appears simple, it is important to realize that understanding and calculating the optimal number of Kanbans required is difficult.
A system's robustness must be studied to gain an understanding of the future environment where unexpected events such as breakdowns occur. If the Kanban system is not optimized and flexible to handle variations, production will suffer and the benefits of the just-in-time system will not be realized (Kleijen, 2003).

Kaizen
Kaizen (good change) was defined as continuing improvement in personal life, home life, social life and working life. When this definition is applied to the workplace, the kaizen philosophy is to create an environment of ongoing improvement driven by company wide involvement at all levels (Imai, 1986). The organization must create a sense of belonging to encourage employee commitment in order to facilitate the exchange of knowledge and ideas without fear of reprisal (Recht, 1998). Continuous improvement requires a concerted effort to identify areas of waste and propose solutions to those problems. Employees are encouraged to make suggestions on a regular basis, and these suggestions create the basis for continuous improvement.  innovation has evolved and is now known as a Kaizen event or a Kaizen blitz. The focus of the event is to improve a specified process or area in a short time period, 20 through the Kaizen ideals of company wide cooperation. One expert describes the blitz as a week long event where a group composed of the process manager and personnel outside the department analyzes the current state of the process or area of focuses, address the problem, and implements the changes (Liker, 2004). By using both approaches, continuous improvement can be sustained both short and long term.
"Ss" is a tool which can be used during a Kaizen event to reduce hidden wastes in the plant through a cleanup activity, and is comprised of:  Reliable product deliveries. S. Improved worker safety and 6. Higher equipment availability (Hirano,199S).

System Configurations
The statement "An hour lost at a bottleneck is an hour out of the entire system" (Goldratt, 2004) embodies the "Theory of Constraints". It is essential to identify a system' s bottlenecks and focus on improving them in order to realize the full potential of the system. In order to address bottlenecks and meet customer demands, manufacturers must be flexible, especially when changes occur with little notice.
Machine configurations ranging from serial to hybrid to parallel have been studied to determine their impacts on key metrics. Research investigating performance was defined through the following metrics: productivity, quality, convertibility and scalability to evaluate each configuration of equipment (Maier-Speredelozzi, 2002). Findings indicated that a completely parallel configuration of machines yielded the greatest performance improvements, with hybrid systems yielding greater performance measurements than serial systems. While this work is encouraging, parallel machining is not a "cure-all" approach, as research has also shown that other configurations have their own advantages. Quality analysis has shown that serial machining had the smallest standard deviation, and different hybrid configurations had the smallest mean deviation and greatest convertibility (Zhong, 2000). The complexity and needs of the system must be addressed for each individual facility, equipment and product. Additionally, the cost of parallel machining must be considered, especially for equipment with long setup and change over times.

Z.4 Chapter Review
The basic concepts of selected lean manufacturing tools and other practices used in outside industries were addressed in this chapter. The overall goal of manufacturing lean is the elimination of waste, as this principle is applicable to every industry. It is important to identify where practices are applicable in the pharmaceutical industry.

Pharmaceutical Manufacturing Background
Chapter 3 will address the current state of pharmaceutical regulations by the FDA through cGMP's. This chapter will discuss Process Analytical Techniques Additionally, regulations regarding over the counter (OTC) products will be discussed as an appealing area for advancing manufacturing based research. Finally, opportunities for pharmaceutical manufacturers to take advantage of the new regulations will be discussed.

Current State of Pharmaceutical Regulations for Manufacturing
Current pharmaceutical leaders have recognized the shortcomings of cGMP' s; examples include dubious and repetitive product testing, validation procedures and extensive documentation for well understood process changes.
Inflexible regulations have hindered process improvements in the past and became the object of industry and FDA focus for change (Hussain, 2002 (Joneckis, 2006) to encourage innovation in the pharmaceutical industry.
The FDA defines PAT as: A scientific, risk-based framework intended to support innovation and efficiency in phannaceutical development, manufacturing, and quality assurance. The framework is founded on process understanding to facilitate innovation and risk-based regulatory decisions by industry and the Agency. The framework has two components: (1) a set of scientific principles and tools supporting innovation and (2) a strategy for regulatory implementation that will accommodate innovation (FDA, 2004a).
The definition for PAT presented does not define the overall strategy of the initiative, but serves as an introduction to the regulatory practice. A practical definition of PAT is given as "systems for continuous analysis and control of manufacturing processes based on real-time measurements, or rapid measurements during processing, of quality and performance attributes of raw and in-process materials and processes to assure acceptable end product quality at the completion of the process" (Hussain, 2002). These revised guidance documents have created a new era for the pharmaceutical industry. The documents shift away from conventional thinking, where process changes are discouraged due to the substantial financial and scientific resources required to fulfill FDA documentation requirements. The new focus is to understand the product, the manufacturing process, and operations. This approach has been described as the "design space'', defined by the FDA as: . .. the multidimensional combination and interaction of input variables and process parameters that have been demonstrated to provide assurance of quality. Working within the design space is not considered as a change. Movement out of the design space is considered to be a change and would normally initiate a regulatory postapproval change process.

26
Design space is proposed by the applicant and is subject to regulatory assessment and approval (FDA, 2006a).
Within the design space, the control space has been described as a: "Multidimensional space that encompasses process operating parameters and component quality measurements that assure process or product quality. It is a subset of the design space" (Desai, 2006). Exploration and understanding of these areas will lead to the identification of critical parameters, as well as metrics and methods to capture their impact on the process, enabling quality management through a risk based approach. A new control strategy aims to minimize risks associated with failures when critical and non-critical process parameters fall outside the control space but remain within the design space.
The FDA has recognized, similarly to outside industries, that quality must be built into the design of the product, and that it cannot be achieved through testing or inspection alone. Throughout the development and product life cycle, changes in formulation and manufacturing practices offer opportunities to gain greater knowledge of product characteristics and performance under various conditions.
Inclusion and analysis of all relevant experimental and experiential information can be used to create, support, and expand the control and design spaces. While absolute operational and process understanding is impossible, it is important to recognize what is known and operate within those parameters. Figure 8 depicts a conceptual representation of the importance of experimental data to provide the framework for creating and supporting the control and design spaces as defined by the manufacturer and approved by the FDA.

27
, , I , , , , , , Figure 8. Hierarchy of manufacturing control strategy (Low, 2006) , , , The FDA has given the pharmaceutical industry an opportunity to enter the 21st century of manufacturing by expanding the control space of their operations while still staying within safe operating conditions in the design space. PAT has excited many in the pharmaceutical industry, who see the potential to continuously improve processes as they occur in other manufacturing industries.
This drastic change in the regulatory mindset has raised many questions regarding feasibility and practicality. PAT has become an industry "buzzword" with much of the knowledge only attainable through consultants, and/or is guarded closely within the industry. Current seminars and workshops address very specific ideas of using novel techniques for limited areas, such as specific assays used during processing to verify quality (Tyler, 2006). Many consultants propose vague "buzzword" laden approaches to improvements without scientific background and appear to be more oriented to selling their services. Adoption of PAT has been slow due to a lack of technical knowledge and trepidation over regulations, as PAT is currently only a guidance document and not a requirement.
PAT is still in its early stages of use, and there is wide spread skepticism and uncertainty of how the FDA will regulate this new area. While the long term benefits of manufacturing improvements are clear, it is difficult to make the argument for a sizable investment in innovation on a product currently being developed for fear of clinical failure and the uncertainties of FDA product approvals.
Further, if a product will be approved, companies want to avoid delaying or jeopardizing the approval process due to changes in manufacturing approaches.
Additionally, there are concerns over technological limitations, which prevent online process monitoring and understanding for complex products such as protein drugs (Glaser, 2006). Others have addressed some of these concerns in outside industries and it is important to learn from their successes and failures.

Examples of Current PAT Use
Near infrared (NIR) spectrometry has been studied extensively for the use of online monitoring of pharmaceutical products.
This non-invasive and nondestructive technique has been studied for a number of different potential uses during manufacturing. This method can identify and differentiate between differently formulated products even if they appear identical, across a wide variety of dosage forms (tablets, solutions, capsules, etc) (Medendorp, 2006). NIR has been studied throughout the manufacturing process, from raw material qualification and 29 quantification, in line monitoring processes and final packaging identification systems. Some monitoring applications include moisture content, active pharmaceutical ingredient (API) concentrations, blending homogeneity, coating thickness and tablet hardness (Swarbick, 2006).
Raman spectroscopy utilizes an energy source, such as a laser or ultra violet (UV) light, to excite molecules from their original states into an excited state, and measures the vibrational level required for the shift (Raman, 1928). This tool has been used in the identification, monitoring of homogeneity, and quantification of different drug products. Polymorphs, defined as a drug substance existing as two or more crystal chemical structures which may have different chemical and physical properties than the original structure (FDA, 2002), have been identified as an opportunity for processing improvements. Continuous monitoring of polymorphic transformation secondary to high sheer granulation in formulations for encapsulation and the solubilization of API in inactive ingredients (excipients) has been studied (Jayawickrama, 2006 There are a number of other methods of utilizing analytical and process instrumentation to identify acceptable processing and product specifications. A few examples are included here: Mass spectroscopy has been proposed to monitor moisture content of drug products during lyophilization to ensure acceptable processing conditions to yield less than 1 % water content (Wiggenhom, 2005). Cell density probes have been utilized to monitor the fermentation process of protein drugs, to reduce potential quality issues, process variability and meet the desired target density with fewer deviations (Low, 2006).
Pharmaceutical manufacturing requires that strict processing and formulation specifications must be met prior to the release of the product. Currently, manufacturers must "test in" quality by sampling a number of randomly selected finished products prior to their release, known as end product testing. By monitoring important metrics during production with a demonstrated correlation with end product quality, one could conceivably avoid end product testing and utilize inprocess (parametric) release testing. In 1987, The FDA approved parametric release only for terminally heat sterilized drug products. Parametric release is defined as "a sterility release procedure based upon effective control, monitoring, and documentation of a validated sterilization process cycle in lieu of release based upon end-product sterility testing. All parameters within the procedure must be met before the lot is released (FDA, 1987)." This regulation only applies to certain large volume parenteral (L VP) products and requires a supplemental new drug application for approval. Currently several manufacturers have gained approval for parametric release using moist heat sterilization (autoclave) methods, but it has not been adopted widely due to the extensive documentation and validation procedures required (Stevens-Riley, 2005). Chemical, ionic, and thermal methods of terminal sterilization for many drug products to avoid end product sterility testing and allow parametric release have been discussed by members of the United States Pharmacopeia (USP) for FDA approval {Tirumalai, 2005).

Over-The-Counter Product Regulations
There are more than 100,000 over-the-counter (OTC) products and 800 active ingredients, encompassing over 100 therapeutic classes currently for sale in the US (FDA, 2000). There are two methods to get an OTC product on the market to sell to the general public without a prescription. Similar to new prescription drugs, a New Drug Application (NDA) for a specific product must be approved prior to marketing which may require clinical studies, mandated FDA review timelines and postapproval maintenance (Christ!, 2006). An alternative method is to be in compliance with an FDA approved drug monograph. Monographs specify active ingredients which must comply with standards that have been found "generally recognized as safe and effective" (GRASE), and must comply with the monograph with regard to the formulation, labeling, and testing requirements (FDA, 2006d similarities between NDA and monograph products include: safety and efficacy standards, manufacturing and GMP inspections, labeling requirements and advertising regulations (Hilfiker, 2006).
OTC monograph products represent an opportunity to investigate new manufacturing techniques. Formulation, dosage form, concentration requirements and labeling are the key monograph factors (Hilfiker, 2004), while manufacturing processes are generally left to the manufacturers. This flexibility allows manufacturers to individualize specific processes to create their own version of the product.
During the approval process for a prescription drug, clinical or bioequivalence studies are required and the manufacturing processes used during this period must remain the same after approval. If there is a significant change to the production process, it must be validated and documented in a supplemental document. The financial and time constraints cause pressure to gain an expedient approval, and can later result in non optimal processes which are not changed for fear of negative FDA action. As a manufacturing model, it would be more feasible to use an OTC monograph product because it has been characterized and studied.
While monographed products still require validation for changes during manufacturing, there is a greater flexibility and information available during the initial production process planning.

Opportunities for Pharmaceutical Manufacturing
There is considerable opportunity to investigate the implementation of current practices and knowledge found outside the pharmaceutical industry for 33 incorporation into novel processes in line with the FDA's cGMP's and PAT

regulations. The concept of the incorporation of external industries has been
proposed (Crosby, 2006), but research focused on these principles and their effects on manufacturing operations and pharmaceutical product development has not been explored. It is important to approach improvements to manufacturing while still complying with cGMP and FDA regulations. Each product and manufacturer should asses whether changes in their regulatory requirements are necessary if changes are made.

4.1 Batch vs. Continuous Processing
One of the most apparent areas that could be improved in the pharmaceutical industry is a change from batch manufacturing to a continuous manufacturing approach. Currently the metal industry has utilized continuous manufacturing for metal matrix composite wires, among other products. Originally the wire was manufactured by batching, but long processing times and extended heat exposure was detrimental to the final product. The process was redesigned addressing the critical parameters required for a quality product, and thorough experimentation was performed in order to identify the optimal operating parameters of the new process.
The new process yielded lower throughput times and lower exposure times which had resulted in heat damage to the metal and an unacceptable product (Blucher, 2001).
Some areas of pharmaceutical production have considered continuous production methods. Tablet, transdermal patch and biotechnology product manufacturing have been studied to determine the potential advantages of continuous manufacturing practices. Granulation is a method to increase powder flow and compressibility (Ansel, 1999), and is often a bottleneck in tablet manufacturing.
Some continuous production methods which have been considered to overcome this problem include: fluid bed agglomeration, spray-drying, extrusion, instant agglomeration, roller compaction and semi-continuous granulation. Continuous transdermal patch manufacturing using hot melt extrusion principles has been studied (Wolff, 2000). Spray drying techniques, used extensively in the dairy industry, have been used in the biopharmaceutical industry but have been increasingly displaced by lyophilization, a batch process. Ease of automation, lower material handling, in-process controls, flexible manufacturing and avoidance of scale-up are some advantages of a continuous system (Vervaet, 2005). It is important to consider that continuous practices are not a "one size fits all" application, and individual products must be critically evaluated. This methodology can be applied early on in the development phase, when considering formulations, but it is critical to evaluate the formula's performance during the manufacturing process, in vitro and in vivo. If production .volumes are kept constant, continuous manufacturing would require significantly smaller space and equipment, potentially yielding a several fold cost reduction (Cini, 2006).
Currently there are technical and regulatory barriers which must be addressed prior to the commercial institution of continuous manufacturing. Few manufacturing systems can be easily adapted to a continuous process without significant capital investment. Publications in this area are sparse, and narrowly focused on a single product, requiring expansive experimentation to gain the information required to change. Additionally, some products are made in small quantities and production goals can be met in a single batch. The FDA has given manufacturers theoretical guidance, and the industry is unsure of the regulatory agent's interpretations of the new documents. Manufacturers should not overlook previously discussed lean techniques which could be evaluated and applied to the pharmaceutical industry now for results in the near future.

4. 2 Lean Techniques
Currently, setup times are very long for equipment and result in large batch processing to preserve their economy of scale. components, or creating one way assemblies of parts can simplify the setup and change over tasks, while reducing the chance of an erroneous setup resulting in damage to the machine or the product.
A pull system could be used to mcrease product flow through the manufacturing facility and movement towards one piece flow. Predominantly the formulation aspect of pharmaceutical products requires a batch process to generate the bulk drug. The following stages could utilize a semi continuous approach to decrease batch sizes similar to Toyota® production method. Batch splitting or mini batching could be explored to increase flow and flexibility in the remaining operations. Kanban cards could be used to facilitate the pull system, especially since the master batch record currently travels with the product throughout processing and acts as a production Kanban. Retrieval Kanbans could be used to help create greater visualization of where equipment and material should be placed to support the flow of production.

Designfor Manufacturing
Design for Manufacturing (DfM) principles have been studied extensively for products such as wiring harnesses, electronics, machining of metal components, powder metal processing, plastic injection molding, sheet metalworking, casting, and final assembly (Bralla, 1986;Boothroyd, 2002;Otto, 2001 ). There are opportunities to develop similar principles for the development of drugs and medical device 37 products. Traditional development methodologies are concerned only with creating a safe and efficacious product. The product is then given to the manufacturing department and is expected to be produced rapidly, despite the fact that manufacturability was not considered at all during the design process. In reality, the transition from development to manufacturing is a difficult and complex process.
Using principles from DfM to guide experimentation during formulation, identification of factors that are critical to the manufacturability of the product will support expansion of the design space to allow more flexible manufacturing.
Examples of experimental focal points include raw material properties, variation of formulation characteristics and delivery device components.

4. 4 Simulation
Simulation has been used throughout many industries to gain insight into current conditions and hypothetical future states. Mathematical models and simulations have been used in operations research for many years to solve operational, warehousing, transportation, distribution and logistical problems (Chung, 2004

Other Manufacturing Tools
Parallel and flexible machining could also be used in many of the processes after the bulk drug has been formulated. Filling bulk product into its container, cartoning of filled drug product and boxing of finished goods are all strong candidates for parallel machining. Utilizing flexible routing and reconfigurable machines can reduce throughput time and minimize the effects of downtimes. Also, resources required for production may be shared, and have a synergistic effect on production rate and setup times.
Online and in line sensing methodology is continuously evolving and has the potential to change the way processes are monitored. Unfortunately the current limitations are sensitivity issues and equipment can not accurately analyze all aspects of the manufacturing processes. As the sensing and analytical technology improves, manufacturers can rely greater on processing parameters and less on end product tests.

Chapter Review
The current state of the regulatory influences guiding the pharmaceutical industry was addressed in this chapter. New guidance documentation and the opportunities which have arisen for the industry were addressed. The opportunities addressed may not apply to all pharmaceutical manufacturers, and it should be noted that there is varying commitment to lean manufacturing throughout the industry.
particularly small and medium companies with more limited resources may need to explore these techniques. 40 Chapter 4

Methodology and Procedures
Chapter four will describe the methods and procedures used to gam an understanding of a product throughout the current system of the company used in this study. Collected empirical data are defined as model inputs to accurately portray the product in the system. A simulation model will be employed to allow visualization of the process and for identification of problematic areas. A theoretical future state will be created in ProModel® software and compared against the original current state to identify which changes to the system yield a significant impact.

Phase I -Current State of Pharmaceutical Manufacturing
Phase I will evaluate a manufacturing facility that produces pharmaceutical products under current Good Manufacturing Practices to gain a "baseline" understanding of current manufacturing systems. This phase will develop a model of the system to identify critical and problematic areas, to identify potential opportunities for processing and formulation improvements.

1 Evaluation of the Current System
The current manufacturing system will be evaluated by "walking the line" and "gemba" (go and see) lean manufacturing techniques (Meier, 2001) to create an accurate current Value Stream Map®. VSM inputs will include the collection of direct measurements, historical data and empirical information. Manufacturing data will be collected through direct and indirect methods, including time and motion 41 studies and the measurement of factors such as Work In Process (WIP) throughout the system, productivity, rework/scrap rates and other critical manufacturing indicators (Hopp, 2001 ). An AutoCAD® drawing of the layout will be used to measure distances traveled by personnel and materials. Detailed employee shifts with breaks will be collected to measure total work hours in a day. Delivery times and quantities will be documented to monitor the flow of materials into and out of the production facility.
Direct methods: • Measuring throughput time, cycle time and change over/setup times.
• Counting materials, components, Work in Process (WIP) and finished goods throughout the system.
• Machine utilization/uptime and operators required.
• Counting defects, rework and scrap rates.
Indirect Methods: • Quality Control Master Batch Records.
• Other documentation recorded during the manufacturing process.

Simulation Modeling
S\·srems Mo<icli11g gives an overview of the steps required to create an accurate simulation and how experimentation can be performed to gain insight into "real world" problems in the current system. ProModel® software is a tool for simulating and analyzing diverse production systems including assembly lines, transfer lines, flexible manufacturing systems, job shops and Just-In Time systems. Resource utilization, production capacity, efficiency, inventory levels, etc. are some of the elements which can be modeled (PROMODEL, 2003). This software will allow experimentation with numerous variables, such as equipment, personnel, and layout with only conceptual alterations to the manufacturing facility. Data collected will be analyzed to identify trends, which enhance production metrics, (throughput, Work in Process, cycle time, etc.) (Hopp, 2001) and those that are detrimental to efficient production. Simulation will allow visualization and will incorporate detailed information about the system while closely conforming to the individual aspects, (Askin, 1993) facilitating experimentation with theoretical scenarios to identify problematic areas and potential failures. Phase one will result in the characterization and creation of a simulation model of the existing manufacturing system to identify the current manufacturing state.

Statistical Methods
In order to identify the most critical parameters after phase I, statistical analysis will be used to determine factors of interest which will be tested during phase II. A quality by design (QbD) method using a Box-Behnken design has been recently employed by Zidan et al. (2006), and will serve as a model. Critical factors 43 will be identified in phase I and ranked according to perceived impact and an appropriate statistical design will be chosen based on the number and type of factors.
The top factors, approximately three or four, will then become the statistical focus of the phase II simulation, with each of the factors containing appropriate levels for thorough study. Examples of potential critical areas include identification of bottleneck workstations, problem areas (defect/scrap rates) and unreliable operations/operators.
The significance of interactions between independent and dependent factors will be evaluated through analysis of variance (ANOV A), with graphs to enhancing visualization where appropriate (NIST, 2006). Minitab® Statistical software, Minitab Inc. Pennsylvania, USA (or equivalent) will be used for all statistical analysis and generation of appropriate tables, graphs and charts.

Phase II -Future State of Pharmaceutical Manufacturing
Phase II will utilize the information gained in phase I to compare current drug processing with a model of a theoretically and experimentally based "improved" system. Improved systems will address the critical processing parameters identified in phase one, and employ concepts adapted from alternative industries to gauge feasibility and potential outcomes.

4.2.J Development of the Future State Model
To address the identified critical parameters, manufacturing tools from across all industries will be utilized including The Toyota Production System®, flexible manufacturing system designs, continuous processing methods, in-line product testing and additional manufacturing tools. New processing and techniques for 44 manufacturing will be analyzed to determine feasibility, using simulation to understand the impacts. Experimentation comparing alternative layouts and theoretical flexible systems to enhance production while potentially minimizing space and expenses will be performed using simulation (Askin, 1993).
A design of experiment using a factorial design was chosen in order to identify which interactions and factors achieve statistical significance. Table 1 describes the 2 3 factorial design employed to compare the different simulation models. There are three factors where the current state for each factor is denoted as zero, and the future state is denoted as 1. Therefore, run 1 depicts the current state of the system and run 8 is the most radically changed future state. In order to evaluate which areas of the future state are significant, six other intermediate simulations will be run, where some factors will be modified while others remain the same.

Chapter Review
This chapter describes the methodology that will be used to evaluate a pharmaceutical manufacturer and a specific product. Using simulation and lean 45 tools, identification of problem areas and solutions will be evaluated. Additionally, by using a statistical design, interactions between factors can be evaluated for significance.

Case Study
Chapter five will describe a pharmaceutical company which was selected as a case study for manufacturing improvements. The current manufacturing processes of a product chosen as the focal point of the study will be described and evaluated.
A value stream map will guide the reader through the process and indicate where value and non value added time is expended on the product. The data collected and used to prepare the value stream map and the ProModel® simulation model will be described in this chapter.

Background of the Pharmaceutical Company Selected for Study
The pharmaceutical company chosen for this study will be referred to as XYZ Pharma in order to maintain their anonymity. While the company produces a few proprietary and prescription products, it primarily serves as an OTC contract manufacturer, producing over 100 human and veterinary products. Management currently describes that they are slightly behind schedule to meet customer deadlines, but feel that they are one of the more reliable and capable contract manufacturers in their market area.

5.1.J Facility
The facility has the capacity to produce oral products such as liquids and suspensions, and topical products including gels, creams, liquids, adhesives and more. This study will focus on issues affecting topical gels and their manufacturing.

47
The company is comprised of a 25,000 sq./ft. production facility and a 75,000 sq./ft.
detached warehouse building, approximately 1/10th of a mile away. The production facility layout can be seen in Appendix II, while the warehouse building, (not shown) is not considered the primary area of study. Unlike many other manufacturers with dedicated manufacturing equipment and personnel, XYZ Pharma is a self described pharmaceutical "job-shop". At any given point they are manufacturing between five and seven different products simultaneously throughout the production facility in various formulation, filling and packaging areas. The company does not have a yearly shut down period and strives to meet customer needs in a flexible and timely manner.
Due to the high volume of products being manufactured and the need to meet deadlines, the overall maintenance of the facility is lacking. While there is an employee dedicated to cleaning and waste removal, large amount of cardboard boxes can be found throughout the facility. Additionally, empty chemical containers, usually 55 gallon drums, can be found in large numbers, 60-90 at any given time.
These barrels are removed by a recycling company which compensates XYZ Pharma at an approximate rate of $2 per barrel. Currently, the recycling company will remove all unwanted drums at the request of XYZ Pharma, which occurs occasionally every 10-14 days. Materials waiting for recycling are stored in the loading dock area, and often monopolize 75% of that area, creating obstructions during loading and unloading of trucks. Material handlers must frequently move raw materials and finished goods multiple times to access blocked pallets.

J. 2 Personnel
There are approximately 55 employees. Many employees are temporary, but permanent employees have between 5 and 15 years of experience at their position.
There are three overlapping shifts: shift 1 starts at 5:00am and ends at 2:00pm, shift 2 begins at 7:30am and ends at 4:00pm, and shift 3 begins at 2:00pm and ends at 10:30pm. Each shift is entitled to two 15 minute breaks and a 30 minute lunch period. There are currently two mechanics who service all of the filling and packaging lines in the facility, with a third mechanic in training. Two manufacturing specialists formulate all topical products. There is currently one "dishwasher" responsible for sanitizing parts and components for equipment, but he feels there is need for another full time washer. Two material handlers move most of the components and trash in the facility, but often assist in other areas when needed.
There are between 20 and 30 operators during the shifts to run the filling and packaging machines. Only a select group of operators have been trained on the newest filling machine, and their operating times are limited to the second shift (7:30am -4:00 pm). During slow periods, operators build inventories of boxes or they are sent home. There is currently a surplus of operators, who are occasionally given time off due to lack of work.

Quality Control
The quality control (QC) department functions to ensure that all shipments received and products exiting the plant meet quality specifications. The QC department consists of a laboratory based group with five analysts. All incoming raw chemical ingredients and packaging materials must be placed into a "quarantine" area. This designation identifies that the materials have not been tested to ensure that they meet the required specifications. A QC analyst must sample the quarantined material and upon successful testing, the material is released to the general production warehouse area or production facility. Raw materials will generally spend between one and two weeks under quarantine before they are released for use.
This group is also responsible for all analytical testing of the bulk formulated drug substance, the final product filled into packaging and stability studies required to determine expiry periods.
The other function of the QC department is the monitoring of all manufacturing operations and documentation. This section of the QC department is comprised of 6 inspectors and a supervisor. There are 12 total QC employees responsible for ensuring quality and proper documentation at XYZ Pharma.
Inspectors rotate around the facility to monitor a number of processes at once and must record their findings in the master batch record. The master batch record captures manufacturing and formulation data and is vital for the manufacturer to prove to the FDA that all SOP's, and cGMP's were followed. All equipment must be inspected by QC prior to machine setups to ensure that proper cleaning has occurred and there are no residual materials on the parts. To ensure proper formulation, QC is required to initial the batch record along with the manufacturing specialist after every step in the process. Tube filling is monitored by checking the weight of 5 tubes every half hour to ensure accurate fill weights. Filled tubes placed into cartons must be inspected every half hour to ensure the machine is producing a 50 legible lot number and expiration date on the cartons. This monitoring is in addition to visual and weight measurements taken by the operators and supervisors to make adjustments to the machines on a regular basis, approximately every 15 minutes.
Finally, it is the responsibility of a separate dedicated QC inspector to reconcile all of the batches manufactured. Finished goods are considered to be quarantined again until the product has been fully tested and the shipping release has been completed, which may take up to a day or two. All packaging components, raw materials, defects and finished goods must be counted. All of the inventory must be accounted for and finished goods production must be within 10% of the desired goal. The QC supervisor then must compile the master batch record with the reconciliation documents and acceptable test results from the QC laboratory group's quality tests.
When all of these documents are compiled, the shipping release sheet is created and signed by the supervisor in compliance with FDA regulations. The batch is then released, removed from the quarantine area and ready to be shipped out to the customer.

Deliveries
One driver is responsible for operating and filling XYZ Pharma's truck, which travels between the main warehouse building and the production warehouse area in the production facility with raw materials, finished goods and recyclables. If material handlers are available, they will assist the driver in loading or unloading the truck. The truck is on a continuous delivery schedule beginning at 7:30am and ending at 4:00pm. Occasionally, the truck will make deliveries or pickup materials 51 from an outside company and material will buildup in both buildings. The truck completes one trip where it is loaded and unloaded between the two buildings in one and half to two hours. These milk runs create a steady flow of materials entering and exiting the system. Currently, the truck is not always filled before it departs from the production facility loading dock. Poorly marked components for return or the unknown status of a finished good prevents the driver from independently making return decisions. If cardboard is to be recycled, it will be taken to the warehouse if the pallet has been wrapped by the pallet wrapper and placed in the loading dock or storage facility, but the driver will not actively search for additional return materials.
Instead, a warehouse supervisor must instruct the driver what to remove.
Additionally, the last truck load at 4:00pm leaving the production facility is empty and stored at the warehouse building's loading area.
The external shipping truck, operated by a freight company, arrives at the warehouse production area daily between 1 :OOpm and 3:30pm, and serves to ship finished goods inventory or deliver small packages of raw materials. Small packages of raw materials are generally less than 5Kg and are stored in a designated area of the production building for easy access by manufacturing specialists during formulation. Large quantities of incoming materials from the external shipper arrive at the warehouse building daily. Outgoing shipments of large finished good orders are removed to the warehouse building as well, where there is ample storage room to organize and store the order. Lead times for raw materials vary, with readily available chemicals arriving daily with a 1-2 week lead time. This is in stark contrast to custom printed plastic tubes and packaging materials which arrive once every 4 months due to their long lead times of 8-12 weeks. Long lead times for materials, coupled with time spent in quarantine results in XYZ Pharma holding a large amount of raw materials inventory in order to meet variable customer demand.

Current State for Manufacturing of Product X
This study will evaluate the manufacturing of a poloxamer based topical gel, referred to as product X, sold in a one ounce tube with an average yearly production of 120,000 tubes per year, produced in 30,000 tube batch sizes as requested by the customer. This product is similar in formulation and manufacturing processes too many other products manufactured at XYZ Pharma and is representative of the system. After all QC assays, reconciliation, and necessary documentation are complete, a shipping release form is completed and the product is ready to be shipped to the customer. A schematic overview of the process and the material flow through the buildings in which they take place can be seen in figure 10.

Formulation of Product X
Prior to formulation, all equipment, parts, and the manufacturing area must be inspected and documented for cleanliness by the QC department. ., Visual and smell tests have been validated with chemical analysis support to identify residual materials from previous batches. Upon QC's approval, equipment setup begins and the manufacturing specialist assembles the machinery for use depending on the product to be produced. The manufacturing specialist is responsible for all fonnulation equipment and other materials such as pumps, tubes, etc. There is one equipment washer who works during the second shift and is responsible for cleaning the machine parts. When he is not working, or if the cleaner is overwhelmed, the manufacturing specialist must clean the parts needed for setup. This may result in used kettles left over night, or weekends until the manufacturer is available to clean the area.
Product X is comprised of six ingredients which are brought from the warehouse, while deionized water is hard piped and validated for purity to meet necessary standards into the manufacturing area. Chemicals are stored outside of the manufacturing area, if space is available, or held in the production warehouse area.
Each ingredient must be weighed, checked, and then documented to be accurate by both the QC inspector and the manufacturing specialist. The ingredients are then added in a four phase approach to ensure a pharmaceutically elegant and accurate product. Empty drums are taken to the cleaning area by the manufacturing specialist during production to keep the manufacturing area clear. Excess raw chemical ingredients are placed outside of the manufacturing area for return to the warehouse, or future use. Product X is formulated in a 1000 L capacity Lee Kettle® equipped with a sweep mixer and a built in Toledo® scale. A smaller 300 L Hubert® mixing vat is used for high shear mixing and solubilization of material prior to transfer with an air pump into the Lee Kettle®. The product undergoes a 23 step formulation process, which is described specifically in the Master Batch Record and that can be categorized as weighing and addition of ingredients, mixing, heating, milling, material transfers and de-aeration (settling). During the production period, the QC inspector periodically checks on the manufacturing specialist dependent on which step he is performing to verify his work and sign the master batch record. The product is completed when it has been entirely transferred into a stainless steel storage vat through a milling process. Table 2 gives the data collected through time studies. The value added time has been identified and serves as the basis for the calculations of the formulation kettle's uptime.

Filling of Product X
A mechanic is responsible for the setup of the filling equipment and must have a QC inspector check all of the parts for cleanliness prior to setup of the machine. The mechanic performs the machine setup, and must prime the machine with product to make adjustments and ensure accurate filling within specifications.
Product X is filled into the hopper from the storage vessel using a transfer pump.  Regardless of the filling machine used, a QC inspector must monitor the fill weights of five random tubes every half hour. A calibrated balance is kept within six feet of the filling machine and is used to weigh the tubes. If one of the tubes does not meet the weight specification, there must be an adjustment and five more tubes are tested. If tubes are not able to be brought into specifications, the machine is stopped for further adjustments and testing. Tubes manufactured during this time are weighed individually to determine if they are within specification, if they fall outside of a +/-10% range, they must be rejected. This in-process testing aims to minimize the number of out of specification products that are produced. Although this system is relatively crude and does not adhere to statistical process control ideas, it does give QC an indication of gross errors.
Operators are responsible for loading empty tubes into the chamber to ensure a continuous supply of tubes for filling. The hopper is fed with formulated material by operators periodically adding the bulk drug through the transfer pump from the storage vat. The hopper has been overfilled by accident, especially when the machine is run with fewer operators. This error requires operator cleanup, machine shutdown and wasted bulk product.
After tubes are filled and crimp sealed they are ejected from the machine and inspected by operators. Each tube must be squeeze tested and visually inspected to ensure a visible lot number and expiration date. If a tube fails either of these criteria it is thrown into a rejection tote which generally consists of <l % of total production.
If the tube passes the inspection criteria, it is put into a plastic storage tote on a pallet.
Each pallet holds up to 20 plastic totes, and the company owns 52 totes.
When the totes are full, boxes are assembled (boxes are later reused for other products) by operators and filled in place of totes if the next process, cartoning, is not yet running. This occurs when the slower machine is used because the cartoner runs at a significantly higher rate and they want to prevent machine starvation. Therefore, the batch is completed to build up inventory and stored near the cartoning machine until it is ready to be cartoned. In contrast, they have overlapped the IWKA ®and the cartoning machine with a buildup of approximately 6,000-8,000 tubes of WIP and were able to reuse totes and avoid using cardboard boxes for WIP storage.

Cartoning of Product X
A mechanic must setup and adjust the Jones® cartoner to accommodate the size of the carton, the feed rate of cartons into the machine and the tape machine (a subsystem of the Jones® cartoner) to fit the different cardboard box sizes.
Additionally, the mechanic must setup the machine to produce an accurate and visible lot number and expiration date. A QC inspector must check and document that the lot number and expiration date is correct and visible prior to beginning cartoning. Before production begins, three to four operators assemble large quantities of preprinted cardboard boxes which are readily available to be filled with cartoned product.
Cartoning occurs at a load rate of 78-82 cartons per minute. The machine can operate through all three shifts and run with between six and eight operators.

Current State Value Stream Map
Data for the current state map of the system for product X at XYZ Pharma was collected through time studies and is presented in Tables 1 through 4.
Additional data was retrieved from the master batch records of four recent production runs of Product X, as shown as in table 5. Uptime percentage was calculated by dividing the value added time (of the batch) by total production time, both value and non value added. Non value added time was comprised of breaks, lunch, machine breakdowns and setup/change over times. Table 5 compiles data from four batches of product X. The batches of the first two columns were filled with the slower filling machine and waiting times designated as * *, were not included because they are not accurate representations of the system when the newer machines are used. Additionally, one of data points for vat storage designated as *, was not included because it was unusually long. The current state value map, shown in figure

Development of the Pro Model® Simulation of Current State
The ProModel® simulation utilizes collected data presented m tables 1 through 4 and compiled master batch records, table 5, to create a model which depicts the current state of the system. All processing times, waiting times, and personnel assignments have been created through observation and verified through company documentation. A full text version of the model can be found in Appendix III. Figure 12 shows 67 6. Raw ingredients are taken directly to the manufacturing area, excess is first stored in the formulation holding area and the remainder is kept in the production warehouse area. 7. Each batch of formulated bulk requires 2 pallets of Raw_Ingredients_Drum and 2 pallets of Raw_ Boxed_ Ingredients to be combined. 8. After a batch of formulated bulk drug has been removed, the manufacturer immediately begins the next batch without a setup, until all batches have been formulated. 9. Mechanic setups of the filling machine begin after the stored bulk drug vessel has been brought to the filling area. 10. Filling using the IWKA ® machine is completed by the "operator", which consists of three operators specially trained on the machine who only work during the second shift (7:30am -4:00pm). 11. The filling machine fills 84 WIP totes, which contain 375 tubes per tote to account for a small batch overage.
12. Mechanic setups of the cartoning machine begins after the entire batch of tubes has been filled into WIP totes and moved to the Jones® WIP feed area.

Chapter Review
This chapter introduced XYZ Pharma as a manufacturer and described the system required to produce Product X. Data collected through time studies and master batch records were presented to give the reader an understanding of how the product is currently manufactured. The current stream value map helped in the identification of problematic areas throughout the system. The results of the current and altered future state ProModel® simulations are presented in the next chapter.

Results
This chapter will present possible areas for improvement to the current system based on the ProModel® simulation and current stream value map. Three of these areas for improvement are modeled in a future state. The primary study, metric, that was chosen, was the minimization of throughput time to improve meeting deadlines for contract customers.
The results of these future state simulations were analyzed for statistically significant factors and interactions.

Results of the Current State ProModel® Simulation
The data tables presented are separated into two groups: inventory and transfer holding locations and equipment locations. The equipment locations are areas that processing of WIP occurs, while the inventory and transfer holding locations are all of the supporting locations for the manufacturing systems. Table 6 shows the results of the current state simulation to gain an understanding of the steady state of the system. Unlike the value stream map, the simulation runs overnight and on the weekends, to describe a more realistic picture.  The average time per entry (min), is the average time an entry spends at a location. The utilization percentage is defined as the percentage of capacity occupied. The utilization percentages appear lower than the uptime calculated by hand for the current stream map because physical capacities were not always fixed or clearly defined. The simulation calculates utilization over the course of the entire run and is different from the value stream map which calculated the uptime for each machine to process the first unit of the product. For the production warehouse area an infinite capacity was assigned to prevent the blockage of incoming or outgoing material; therefore, there is no calculation of the utilization %. Table 6 shows that the two highest utilization percentages of equipment or inventory transfer and holding are the vat holding area and the IWKA ® filling machine, respectively. These areas appear to be bottlenecks in the current system and should be viewed as areas for improvement for the future state. While the production warehouse area does not have a calculated utilization percent, the average time per entry is the second highest and should also be considered as a potential area for improvement.

Areas for Improvement
All of the areas for improvement suggestions considered in this paper are from an efficiency approach. cGMP's must be followed for each product and manufacturer considering changes. Experimentation and validation may be required to implement the proposed changes. The FDA and appropriate regulatory documents should always be consulted if there is uncertainty in a regulatory area.

6.2.1.l Testing
Testing bulk at the same time as finished goods is double end product testing.
This is a common practice where management is confident that its process will result in an accurate product because it has been rigorously validated, but it does leave the company vulnerable if the assay fails because many tubes have already been filled by this point. A bulk drug product can be "reworked" to meet specifications if a deviation report is filed, but once the product is filled and crimped into tubes, it is far too labor intensive and cost prohibitive to empty each one for rework.

JIT QC
The quality control department is responsible for reconciling (counting) all materials that have been used, defects produced and finished goods for release. The majority of finished goods may not be released from "quarantine" until the product has successfully passed shipping release testing for quality and QC reconciliation to meet FDA requirements. Finished goods can be shipped under "quarantine" if the reconciliation documents have been completed for the materials shipped and a sample has been taken for testing. Upon successful testing of the product, the customer is faxed confirmation of the test results and the product is removed from "quarantine." This practice would shorten lead times and could be considered for low risk products to ensure that a batch is not defective. Often the QC inspector will wait until the entire batch is completed before counting the finished goods inventory and compiling the final batch release documentation. Due to space limitations in the production facility warehouse, quarantined finished goods may be moved to the larger warehouse building until they pass shipping release parameters.
Unfortunately, the QC inspector is often unaware if a pallet of quarantined finished goods has been moved to the other building and can cause delays and errors during reconciliation. Conversely, if the inspector does count a portion of the finished goods and they are available to be moved to the main warehouse, it will not be moved by the material handlers unless the warehouse supervisor instructs them. By having the QC inspector keep an ongoing count of finished goods pallets as they are produced and placing a marking such as a neon sticker onto the pallets, the material handlers will know to put the pallets on the milk run truck. This will free up space in the production warehouse area and decrease the time required for reconciliation after production is complete.

2. 2 Deliveries
"Milk-Run" Truck Efficiency Improvement Currently the "milk run" truck runs between buildings approximately every one and a half to two hours. The truck is not always full, and leaves the production facility empty on its last run at 4:00pm. This is surprising considering the production building warehouse is normally very congested and items are usually blocking each other. Cardboard generated during manufacturing can be found throughout the facility and is only removed occasionally. Excess raw ingredients and components often stay in the production warehouse area for a few days after they could have been removed. By utilizing the proposed QC improvements above, the truck will have 76 excess material to load throughout the day and can fill the truck more often. Storing the truck at the production facility overnight will allow late shift employees to place materials directly into the truck and avoid the production warehouse area completely.

.l Decreasing Storage Vessel Batch Sizes
The steps involved in the formulation of bulk drug can not easily be changed, as the process has been given by the customer and validated by XYZ Pharma. A significant change in the formulation would require revalidating the process for the FDA. This product is relatively complex and requires many processing steps, many of which have been developed with current batching equipment in mind. Until development of continuous formulation methods and equipment improve, the product is not currently a strong candidate to make the transition from batching.
However, after the bulk drug has been formulated, there are opportunities to decrease the batch sizes of the WIP in the system. Currently, all 877 Kg of the bulk drug is milled directly from the Lee Kettle® to a large stainless steel transfer vat. This process takes an average of two hours and the transfer vat is then available for the mechanic to prime the next process, the filling machine and complete the setup vat, which are often in short supply.

Decreasing WIP Buffers
Currently tubes are filled with product X and then put into plastic totes. follow SOP's (standardize), which leads many in the pharmaceutical industry to feel that only the remaining 3s applies. Storage areas are less strictly monitored and XYZ Pharma is in dire need of a 5s plan for its production warehouse area.
Although the wall does contain signs to indicate where incoming and outgoing material belongs, it is rarely followed due to the lack of free space. There are pallets with materials that do not belong there which should be returned to long term storage or other locations. There are no markings on the floor to alert material handlers where materials can safely be stored without obstructing walkways. Material handlers waste movements when they must repeatedly move materials to access 79 blocked pallets. Usually materials that are blocked tend to be left in the storage area long after they have been used or needed, which may be considered an inappropriate long term storage area by the FDA.
Sort: By evaluating the contents of the current production warehouse area and removing unnecessary materials, the flow of the product will be greatly improved.
Stabilize: Creation of designated "homes" for raw materials, finished goods, pallet trucks, recyclables, etc. will facilitate the rapid identification of materials that are out of place or in need of attention. Shine: All of the cardboard boxes and empty material drums should be removed and the loading dock and production warehouse area should be power washed. Standardize: Create standard operating procedures, which are based on visual information, such as simple diagrams and cartoons to indicate the proper method to store and remove material. Create a regular schedule, every three to four days, for empty drum removal. Sustain: Create a reward system for maintaining the area and following the standardized work. Incorporate cleaning and organization into employee's jobs, and utilize down times to ensure the work place is maintained.

Visual controls to decrease setup times
Incorporate visual control techniques into the equipment and manufacturing areas. Use shadowboards for mechanics tools and for parts used in equipment setups (tubes, valves). Some setups have over 50 individual components that are placed on a moving cart where QC inspects them, but QC is unable to identify whether parts 80 are missing or in excess. If parts are missing, the setup process is disrupted and results in the mechanic having to track down the proper tool or part. Additionally, manufacturers can easily assemble kettle components, such as rubber blades of the sweep mixer incorrectly, which can have severe impact on the product. While the manufacturers and mechanics are highly trained, these suggestions would benefit new employees being trained and help the current staff avoid errors. These measures would facilitate cross training of employees and give more employees a broader system's perspective.

Autonomation
The incorporation of autonomation to a number of locations at XYZ Pharma would result in improved product quality and fewer operators needed on machines.
Currently, the beverage industry commonly uses online sensors to monitor filling rates and accuracy during production, thereby removing the need for end product content uniformity testing which is currently a time consuming and labor intensive FDA requirement (Jeffries, 2003). The addition of online sensors to the filling machines at XYZ Pharma would allow operators more freedom to attend to other tasks and decrease the total number of operators required. These operators could be reassigned to other areas or serve to cover break periods.
Specifically, the IWKA ® is pre-wired but not equipped with a sensor to monitor the loading of empty tubes. The sensor would automatically shutdown the machine in the event of a misfeed or a tube jam. This is a common problem which currently requires an operator to monitor tube feeding. The hopper is also susceptible to bulk product overflow, where currently a mirror is the only equipment used to monitor the fill height. Ultrasonic sensors have been employed for a number of years by biotechnology, oil, food and other industries for processing monitoring and control, automotive techniques and chemical analysis. (Hauptmann, 1998) These sensors have evolved into low cost tools which could monitor hopper fill levels, as well as monitor the bulk material viscosity and homogeneity for enhanced online monitoring.

Cross Training Personnel
The current training of operators is performed on the job. Most operators are trained on two to three pieces of equipment and rotate when needed. While job rotation is performed, it could be expanded. The new IWKA ® filling machine is operated by a select team of operators and is limited to one shift. By increasing the number of operators that can run the machinery, the overall processing time could be increased. As there is a surplus of operators, and often slow periods are used to build boxes for later use, training could be accomplished during these points. Also, by creating a rotating team of cross trained employees to cover break and lunch periods, machines would remain operational and keep pace of production at a level pace. An additional mechanic for the third shift is required to setup and maintain the filling machinery, which they are currently in the process of training.
Additional areas for cross training could include cleaning responsibilities. To alleviate both the mechanics and the manufacturers, operators could rotate into the 82 cleaning area to sanitize equipment parts during changeovers. This would decrease waiting times and give mechanics and manufacturers greater time to use their skills for specialized tasks. There are only two manufacturing specialists currently operating 5 days a week. Cross training other employees in basic areas of the manufacturing specialist's responsibilities such as prep work, setup and changeover would also give the manufacturing specialist more time. If a manufacturing specialist is sick, their manufacturing area is not used which results in greater product lead times. Training additional personnel to formulate would alleviate problems arising from unexpected work absences, as well as increase the capacity to produce during the weekend.

Off shift personnel
In addition to cross training employees, off shift personnel could decrease a significant portion of non value added time that is currently experienced. A small team of personnel could perform a number of tasks during non operational hours (10:30pm -5:30am) to improve flow during operational hours. Cleaning would be the primary responsibility of the team. Areas would include cleaning formulation kettles, parts for equipment setups, empty raw material drums and the facility itself.
Material movement would be the other focus, such as cardboard for recycling, components for return to the production warehouse area and raw materials for production the next day. The truck for inter building deliveries could be stored at the production facility loading dock and loaded throughout the night in order to remove unneeded materials. 83 6.3 Development of the Future State ProModel® Simulation  Mechanic_ 2 has been trained for the third shift, while the fust mechanic has been moved from the second to the first shift. In the case of a cross training scenario without the addition of the second filling machine, the 3 dedicated operators from the IWKA ® will be cross trained and distributed. Operator_ I and operator _2 will each gain one operator and have 8 each; the odd operator will be removed and considered reassigned.
Factor C changes the batching rules during production. First, bulk drug substance is no longer transported in one large transfer vessel.        Decreasing the amount of WIP in the system is an important goal of lean manufacturing. WIP crowds the plant's floor space and should be considered a liability. Since XYZ Pharma is a contract manufacturer they are not paid to manufacture WIP, but are paid for their finished goods. This lean thinking challenges the manufacturer to decrease the lead times in order to make more money.
This methodology is in stark contrast to the traditional approach of a large WIP inventory, which may appear to be giving the manufacturer an economy of scale.
Additional inventory in the system will hide problems such as defects, inefficiencies, long setup times and long lead times, while incurring the costs of storage and delayed lead times (Liker, 2004).  pallet is currently unavoidable. In an ideal setting, palletizing would be an unnecessary step and boxes would be shipped as soon as they had been completed and reconciled. Unfortunately, due to QC testing and customer requirements decreasing shipments to single pallets instead of individual or groups of batches, 6 pallets each is the best approach. Scenario 4 spends approximately 50% of the time as the current state, scenario 1, while scenario 8 only reduces the average time per entry (min) by 15% of the current state. Additionally, scenarios 3 and 7 reduce the average time per entry by roughly 35% each. The common factor among these four scenarios is the cross training of employees. Since the IWKA ® filling machine is operated for two shifts in these scenarios, the probability that the Jones® cartoner will be waiting for material is reduced and time spent waiting to accumulate enough boxes to fill the pallets will be decreased. Cross training will result in more available material handlers, as they are assisted by the driver and operators in other areas of production.  Table 10 shows the utilization(%) of locations in the model, which is useful to identify possible new bottlenecks created in the system after changes have been made. An increase in utilization can not automatically be described as a bottleneck.
"A bottleneck is any resource whose capacity is equal to or less than the demand placed upon it. And a non-bottleneck is any resource whose capacity is greater than the demand placed upon it" (Goldratt, 2004). For example, a utilization of 99% does not constitute a bottleneck if it is able to meet demand. Fluctuations and variability in the system must be accounted for with excess capacity to prevent the formation of bottlenecks throughout the system. Therefore, the utilization can be used to identify locations for further investigation. Many manufacturers strive to reach high utilization rates for expensive machinery and research has shown that this can result in a large buildup of WIP in front of the machine to avoid starvation (Li, 2007). This buildup of WIP has negative effects on the system and can be seen from the batching used at XYZ Pharma.   improvements in scenario 4 improve upon the current system by spreading out the arrivals of the raw materials, but inventory is still held in the formulation holding area. If this JIT system was instated, the current formulation holding area could be converted into a processing area, or other purposes.

4.1 Statistical Analysis of Results
To determine whether the three factors and their interactions are significant a full factorial design was analyzed. The adjusted R 2 value is defined as, "a statistic that is adjusted for the 'size' of the model, that is, the number of factors" (Montgomery, 2005a). The adjusted R 2 value is correlated with the amount of explained variability. Adjusted R 2 was calculated to be .89, which indicates that model is expected to explain 89% of the variability in the data. This value is given in Table 11 , along with the unadjusted R 2 , number of observations and the root mean square error. The summary of fit and the lack of fit tests are statistical methods to determine whether there is an unacceptably high level of variability in the system due to factors which do not belong in the model.   Table 13 is the analysis of variance (ANOVA), "tells us whether there is a difference among means. It does not tell us which means differ" (Montgomery, 2005b ). The P value is given as zero, which indicates that the difference in means is statistically significant. In order to determine which factors differ and if it is significant the effects test was performed in table 14. The interaction between each of these factors is also examined by this statistical   Even though all of the factors and interactions except one, parallel manufacturing and cross training, are statistically significant, that does not imply that they all have the same effect on the outcome. Figure 22 shows parallel manufacturing having virtually no effect on the throughput time, while cross training and batching changes have greater impacts. Figure 23, the Pareto chart, quantifies and ranks the effects of each of the terms and interactions.

Future State Value Stream Map
The future state value stream map was created based upon results found from scenario 4. This scenario yields the lowest simulation throughput time which is the primary goal for a contract manufacturer. In contract manufacturing overproduction is not a concern because company XYZ Pharma only produces what their customers have ordered. This scenario utilizes cross training and change in batching rules to decrease lead times. Differences from the current state with regard to cross training are the increase in shifts that are available for the IWKA ® filling machine to operate, due to a greater number of operators able to run the machine. The most significant factor that was found through the statistical analysis was the batching rule changes.   While orders are made to meet customer orders, the internal flow of product is pushed through the system. This is due to batch processing which results in materials being pushed onto down stream machinery. This method then results in manufacturing of as much of the order as possible in their shift, which is then batched to move to the next process. There is a lack of production leveling which results in uneven production of materials.

1. 2 Quality
The focus on product quality is extremely high in the pharmaceutical industry to avoid potentially fatal and costly defects. While PAT is an emerging resource, many tools are used to ensure that the product meets the rigorous specifications required to be safe and efficacious. The pharmaceutical industry is in the process of moving away from end product testing and toward in process testing which has been used for many years by other industries.
Product quality for pharmaceuticals comes at a price. The pharmaceutical industry carries a large amount of inventory "just in case" and face problems other manufacturing industries may not. Large quantities are held in inventory for two reasons, 1. The average lead time for custom components is 4-6 weeks and for materials is 1-2 weeks and a last minute increase in the order size would result in a long lead time. Additionally, components from the customer are not purchased and may incur the cost of holding which is then build to the customer if held for long periods of time and 2. All batches of incoming materials and components are quarantined and require identity conformation. Regardless of the size of an incoming order, only one test is required and is regarded as more convenient by quality control. Component manufacturers are not willing to change to a more JIT approach and experience many similar problems due to their batch mentality.
Quarantine is not an issue for most other manufacturers and requires an additional inventory buffer to ensure that proper quantities of raw materials are readily available.
With only a small number of capable and reliable specialized pharmaceutical component manufacturers, they are able to dictate long lead times without losing business.

Equipment
cGMP compliance requires many elements that other industries do not face.
Validated methods for cleaning of equipment and manufacturing areas are important to decrease the chance of unsafe levels of contamination from the environment and personnel. The cleaning required between products is rigorous in order to decrease the chance for cross contamination between products. This results in longer time required for setups and changeovers. If setup times are addressed as a problem and visual controls are instituted, any significant changes to the physical equipment may require validation from the FDA. If the process has undergone validation, the use of two parallel machines is acceptable, but must be evaluated carefully due to long setup times and cleaning requirements.

1.4 Facility
Some production facilities have a "Job Shop" layout, with all formulation equipment grouped near each other and products transported in large vats for filling 112 and palletized totes of WIP. A redesigned layout would place all the equipment in proximity to each other in order of flow. During formulation, the general area is kept clean for manufacturing and is designed for ease of cleanability. While filling operations, it tends to generate far more scrap materials. Due to cGMP regulations, certain layouts may not be feasible for fear of contamination to the product. An improvement to the flow of material within the guidance can still be investigated.
Another facility problem is the storage of components and chemicals.
Storage areas for raw material and finished goods can become cluttered due to the quarantine process and the large quantities required for batch processing. This is in contrast to a JIT system which would remove finished goods upon completion and store needed components close to the manufacturing line.

Study Limitations
Every effort was made to accurately capture the current state of manufacturing for product X. Data for the model was based on four batches with two batches using the older filling machine and two batches using the newer filling machine. The future state model data was based on assumptions derived from observations and were not data collected after implementation of the suggested improvements. This case study does not accurately represent the entire pharmaceutical industry, especially some of the larger companies which have greater financial resources to investigate lean and alternative techniques.

Future Work
Further studies using ProModel® simulation software should be conducted to improve the process understanding and diagnosis of problems. Future areas of study addressing DfM, failure modes effects analysis (FMEA) and design of experiments should be addressed. Experimenting with continuous manufacturing equipment during the formulation process and process development should be considered, especially for large scale and volume products. The methods and findings from this study could be tested to see their applicability to product dedicated equipment and facilities.