Reprinted from

PHARMACEUTICAL ENGINEERING

information systems

THE OFFICIAL TECHNICAL MAGAZINE OF ISPE

JANUARY/FEBRUARY 2015, VOL 35, NO 1

Manufacturing Execution System Technology

Copyright ISPE 2015

www.PharmaceuticalEngineering.org

Implementation of QbD
using MES
by Michael Choi, Mehron Mirian, Pamela Bruen Docherty, and
Gregory Ruklic
This article explores the use of Manufacturing Execution System (MES)
software as an enabling tool for the Product Lifecycle Management (PLM)
aspect of Quality by Design (QbD).

s life sciences industries endeavor

to bring the concept of Quality by
Design (QbD) from opportunity to
implementation, companies can
either develop new tools in software
and data management or apply
existing tools to facilitate a more cost
effective path. One such opportunity
is to utilize Manufacturing Execution System (MES) software and related systems to enable
the knowledge and risk management aspects of QbD. A
major obstacle to this approach is mainly perceptual, in that
MES technology was originally developed and is viewed as
a manufacturing tool set only, even though the technology
can be used in various portions of the product development
phases in the product lifecycle. Product Lifecycle Management (PLM), improvement of process performance and
product quality as in ICH Q10, Section 3 typically utilizes a
series of gate processes to require examination of the state
of lifecycle requirements at each phase, allowing a decision
to pause/correct issues and requirements or release to the
next phase. MES capabilities, such as recipe creation, as well
as data presentation can create knowledge while providing a
consistent tool set and user interface throughout the lifecycle
across departmental functions.
There are a number of activities within the QbD paradigm
to which MES tools can provide an engine for implementation:
Configuring MES as a knowledge repository. MES

technology presents industry with a great opportunity to

utilize existing and proven technology to create and manage a knowledge repository.
Utilizing MES as a continual improvement tool for manufacturing. MES software can be used to interpret the flow
of process data into practical and useful information for
continual improvement.
Using MES as a tool to allow knowledge gained from
manufacturing to be transported back into R&D for assimilation as prior knowledge. New knowledge learned
in manufacturing is systematically collected in MES and
made available through the system to R&D to improve
process design and development of a control strategy.
This, in turn, results in systematic incremental improvement to the robustness of the manufacturing process and
reduction in time and effort for new product launches,
completing the circle of knowledge movement within the
lifecycle.
This article will introduce these concepts and explore the
opportunities in application to QbD.

QbD Concept
The concept of QbD incorporates the paradigm that quality
can be designed instead of tested, as defined in ICH Q8(R2)
and outlined in the FDA guideline Pharmaceutical Quality
for the 21st Century: A Risk-Based Approach. The focus
of QbD is building quality into a product, which involves
a thorough scientific understanding of the product and
processes by which it is developed and manufactured, and

PHARMACEUTICAL ENGINEERING

JANUARY/FEBRUARY 2015

information systems
Manufacturing Execution System Technology

can be found in the concept paper by the

ISPE PAT COP.1

MES Concept
MES is considered to be the complete
interactive system of human, electronic,
and mechanized functionality to execute
manufacturing operations. Manufacturing execution systems are an integral
part of a complete automation solution
especially in processes that interface with
other applications. The MES domain as
shown in Figure 2 includes recipe management and weigh and dispense with
integration to Laboratory Information
Figure 1. QbD approach showing overarching principles and some enabling tools (source:
Management System (LIMS), control
ISPE PQLI Guide Series: Part 1 Product Realization using Quality by Design (QbD):
systems, Enterprise Resource Planning
Concepts and Principles, including Overview, Criticality, Design Space, and Control Strategy).
(ERP), document management and other
systems or applications. Common MES
knowledge of the manufacturing risks at product realization
functionality includes process management, data collection
and throughout the lifecycle of the product. The conceptual
and acquisition, product tracking, and parts traceability
application of QbD through a products lifecycle is shown in
(genealogy of raw materials, product, equipment calibraFigure 1. A product lifecycle approach makes use of product
tion, etc.) Note that this is a functional model and does not
formulation and process characterization in the forms of starestrict a function to any particular software application.
tistical, mechanistic, and first-principle model information
to be in a format that facilitates Knowledge Management
MES as Lifecycle Knowledge Repository
(KM). Similarly, the risk-based approach to QbD requires
MES is largely a data management system. Data is verified
the same disciplined approach to identify, analyze, control,
and transported from MES into downstream system to faciliand communicate quality risks throughout the life of the
tate human interaction and automated execution. Newly
product.
created data from these activities are transported back to
MES for distribution, storage and further activities. To
realize MES as a lifecycle knowledge repository, data must

Manufacturing execution
systems are an integral part
of a complete automation
solution especially in processes
that interface with other
applications.

Quality Target Product Profile (QTPP) and Critical Quality Attributes (CQAs) focus on establishing the relationship
between quality attributes of the product and their impact
on safety and efficacy. The subsequent parts in Figure 1
greatly benefit by application of the enabling tools and data
management. For further information, refer to the International Society for Pharmaceutical Engineering (ISPE) Product Quality Lifecycle Implementation (PQLI) Guide Series.
Also, more detailed examples of knowledge needs for QbD

JANUARY/FEBRUARY 2015

PHARMACEUTICAL ENGINEERING

Figure 2. MES domain (source: GAMP Good Practice Guide:

Manufacturing Execution Systems A Strategic and Program
Management Approach).

information systems
Manufacturing Execution System Technology

in KM allows estimation of the probability and detectability of potential hazards

that are used in risk assessment, such as
in Failure Mode Effects Analysis (FMEA).
If the risk is high, changes are implemented in the process to reduce this risk.
The changed process is then monitored
and communicated via real-time alerts,
alarms, and reports of Out-of-Trend
(OOT) and Out-of-Specification (OOS)
conditions. Changed process monitoring and communications are available in
current MES technology to accommodate
the risk knowledge (via automated and
manual observation and information
capture, and risk-based controls and
communication.)
In addition to the factual and risk
information captured as knowledge,
the scope of KM in MES should also
encompass description and experiences.
Risk-based justifications are often made
in consideration of internal quality poliFigure 3. Example illustration of data to knowledge (SCFM = Standard Cubic Feet per
cies and procedures, agency guidances
Minute; SPC = Statistical Process Control).
(e.g., SUPAC), and agency responses
be transformed into knowledge (Figure 3). Data manageto regulatory compliance concerns such as documented
ment systems collect values of qualitative and quantitative
483 inadequacies). KM at the minimum should capture
(factual) information. Typically, data by itself represents
these descriptions and experiences, which also would allow
pieces of information and not necessarily whole informaimproved estimation of the probability or detectability of
tion, e.g., reporting process parameter data without units
potential hazards.
of measure or other contextual information has no physical
MES, as primary or as an integrated system, may be a
meaning until these data are combined. Similarly, a set of
repository of :
information is combined into knowledge through establishing information relationships, performing analysis and mak API characterization
ing judgments,e.g., whole information about one parameter
Finished-product characterization
is not considered useful knowledge until its relationship
Process characterization
to process and product performance is established. There Manufacturing capability
fore, to address the knowledge management needs of QbD,
MES capabilities would include means of organizing data
The contents may include:
into information, and further organizing information into
knowledge.
Data from clinical, development, validation to manufac Another component of QbD is risk management. Qualturing
ity Risk Management (QRM) as in ICH Q9, consists of risk
Correlations between parameters and quality attributes
assessment (identification, analysis and evaluation), risk
Risk rankings justifications for and means of correcontrol, risk review and communication of risks. Within each
lating values of uncertainty, severity, probability, and
part of QRM, risk information is captured and tracked on
detectability in risk management
top of the factual information. For example, the relationship
Justifications (e.g., risk priority/threshold value) for critibetween a CQA and a process parameter is established and
cal parameters and internal limits
captured as a relational knowledge in KM. If this relationship
Quality and regulatory limits, policies, standards, and
is such that the variability of this parameter impacts the CQA,
procedures
this parameter is identified as a Critical Process Parameter
(CPP) and a potential risk is identified. Analysis of developThe interfaces to the repository should include all relevant
ment information and manufacturing capability information
data sources, e.g., LIMS, ERP, change-management system,

PHARMACEUTICAL ENGINEERING

JANUARY/FEBRUARY 2015

information systems
Manufacturing Execution System Technology

and Process Analytical Technology (PAT) and historian data.

MES as Continual Improvement Tool

Continual improvement tool requires continual flow of
information translated into useful knowledge and applied
to improve the process. As an example, assume a CPP-CQA
correlation model created at the R&D stage from first principles, mechanistically, or statistically through Design of Experiments (DOE). The model coefficients are progressively
improved by the continuous circulation of new information
back to developmental phases for testing, confirmation, and
update of process specifications concurrent with commercial
manufacturing. Of course, these changes are managed by
proper configuration management and change control gate
processes coordinated with lifecycle management gating
processes. The new information may be from online (realtime), offline, and at-line data. As the model improves with
new information, the specified control limits will be adjusted
accordingly. Variability in manufacturing can be used to
correlate the effect of process, operations, and raw material
properties on quality attributes, which results in continuous enhancement of process knowledge and in turn leads to
an improved process. Either the continual improvement or
variability approach can systematically address the need for
continued process verification.2
Additionally, a model in combination with PAT can
optimize the process in real time, and ultimately be used for
real-time release. For example, PAT may include a control
system with an NIR sensor that detects the dryness endpoint
of fluid bed dryer. Combined with a drying model, this system can be made to control multiple variables (e.g., the process air flow rate and inlet air temperature) simultaneously
within the predefined Design Space (DS) to speed up the
drying process while maintaining the same quality endpoint.
The outputs from the new control setting are transmitted to
MES and verified against the acceptable range. The correlation between acceptable range and product performance
can be used as an enabler to real time batch release The new
control settings and process outputs feed back into MES
to further enhance the process knowledge by refining the
model coefficients for future application.
Enhanced features of MES, as the continual improvement
tool, may include:
Real-time monitor of process capability of univariate and
multivariate (assuming the multidimensional QbD design
space is implemented) systems using statistical and multivariate analysis capability.
Continuous tracking and updating of empirical models
describing the CPP-CQA relationship.
Adjustment of recipe parameters/set points based on the
changes to current product knowledge.
Update of risk ranking and tolerance values for alerts and

JANUARY/FEBRUARY 2015

PHARMACEUTICAL ENGINEERING

alarms based on process feedback.

Considering the above, MES technology can be a tool to
improve design space by facilitating analysis of risks in multidimensional combination of CQA and CPP and improve
the criteria for release of batches per improved process
specification limits based on the updated understanding
of CQA-CPP relationship. These enhanced features may
be integrated directly into MES or be a separate software
linked dynamically to MES. Since new type of analysis may
be needed with new knowledge, modular add-on software
design may be used to accommodate new improved models
for the design-space characterization.

MES as Prior Knowledge Provider for R&D

MES as prior knowledge provider is a natural extension of
the knowledge repository and continual improvement tool in
R&D. MES as knowledge repository provides development
history and knowledge of prior studies done during development. If scientific principles were applied in prior studies
such that mathematical and statistical models explain manufacturing process models or drug-delivery mechanisms,
fewer experiments should be needed to confirm the prior
findings and not repeat the same exploratory studies that are
costly and time-consuming. MES as continual improvement
tool provides process knowledge gained from manufacturing back into R&D. This knowledge, in the form of CPP-CQA
relationships, contributes to reducing process development
efforts and time to market while strengthening the robustness of the manufacturing process design for a new product.
This knowledge is directly applicable to setting the control
strategy and understanding the risks associated with the
control strategy.
MES, as the prior-knowledge provider, may require:
First-principle models (e.g., thermodynamics)
Empirical and semi-empirical model coefficient data to be
generated, stored, and recalculated
Multivariate statistical modeling from manufacturing
data
Statistical design of experiment set up and modeling using experimental results
Non-dimensional parameters that are scale independent
Deterministic or Monte-Carlo simulation tools to assess
risks associated with control strategy
For example, to determine the initial control strategy of a
coating process for a new product, process performance
information from coating processes of existing commercial
products from the same coating equipment can be correlated
by analyzing the operating conditions and process responses
(e.g., using a process thermodynamics relationship).3 Process variability information also can be obtained in the same

information systems
Manufacturing Execution System Technology

regression analysis that is used to build the correlation. This

variability information can be used to estimate the risk probability information for an FMEA-type risk assessment. This
process correlation and risk information can be used as prior
knowledge for setting and understanding the risks of the
initial control strategy.

MES enables an integrated

smooth tech transfer from R&D
to commercial manufacturing by
managing process controls.

Note that control strategy is a key element of QbD and

risk-based manufacturing approach, and the main deliverable of Stage 1: Design Qualification in FDA Guidance for
Industry.2 MES plays a key role in managing the continuous
process quality verification of CQAs and controlling CPPs.
This also includes the management of the process models.
The control strategy must be developed and managed along
the product quality lifecycle. MES enables an integrated
smooth tech transfer from R&D to commercial manufacturing by managing process controls. This in total builds the
overall process performance and product quality monitoring
system.

For Further Information

For more detail and related information, the following ISPE
resource is available:
MES Special Interest Group (SIG)
The MES-SIG is comprised of volunteer professionals who
identify, analyze, and propose solutions to specific MES
problems faced by life science companies.
This article has been developed by the Manufacturing
Execution Systems Special Interest Group (MES SIG) of the
GAMP Community of Practice (COP), a technical subcommittee of ISPE

Acronyms
API

Active Pharmaceutical Ingredient

CQA

Critical Quality Attribute

CPP

Critical Process Parameter (in this article, CPP, for

practical purposes, is defined as all critical independent controls that impact CQA including raw
material attributes)

DOE

Design of Experiment

DS

Design Space

ERP

Enterprise Resource Planning

FDA

Federal Drug Administration

FMEA Failure Mode Effects Analysis

KM

Knowledge Management

ISPE

International Society for Pharmaceutical

Engineering

LIMS

Lab Information Management System

MA

Material Attribute

MES

Manufacturing Execution System

NIR

Near Infrared

PAT

Process Analytical Technology

PLM

Product Lifecycle Management

PQLI

Product Quality Lifecycle Implementation

QbD

Quality by Design

QTPP

Quality Target Product Profile

R&D

Research and Design

SCFM

Standard Cubic Feet per Minute

SPC

Statistical Process Control

References
1. ISPE Process Analytic Technology (PAT) COP Data
Management Task Team, Implementing Knowledge
Management in Bioprocesses: A QbD Driven Approach
Turning Data into Knowledge in Reference to the CMC
A-Mab Case Study, Concept Paper, March 2012, www.
ispe.org.
2. Section IV.D Stage 3 Continued Process Verification
in Guidance for Industry, Process Validation: General
Principles and Practices (FDA, 2011), www.fda.gov.
3. Choi, M., Applications of Process Thermodynamics in
Pharmaceutical Coating, Tablets and Capsules Magazine, April 2007, www.tabletscapsules.com.

About the Authors

Michael Choi, PhD is General Manager
for Johnson & Johnson HyangNam Pharmaceutical plant in the Janssen Supply
Chain. Previously, he served as Head of
Process Robustness for Americas Technical Operations at Teva, Senior Principal

PHARMACEUTICAL ENGINEERING

JANUARY/FEBRUARY 2015

information systems
Manufacturing Execution System Technology

Scientist for Consumer Healthcare R&D at Schering Plough,

and Senior Process Engineer for Pharmaceutical Technology
and Engineering at Merck & Co., Inc.
Johnson & Johnson HyangNam Pharmaceutical Plant, 45
Jeyakgongdan 2-gil Hyangnam-Eup, Hwasung-si, Gyeonggido, 445-937, Korea.
Mehron Mirian has more than 14 years
of experience as a quality professional in
the medical device, pharmaceutical and
diagnostic industries. Mirian holds a BS in
biology and has worked for several companies, including B. Braun, Bio-Rad, and
Cardinal Health in the United States. He currently manages
a commissioning and qualification of automation project
at B. Braun Medical, Inc. as the Senior Computer System
Validation in Quality Assurance Department.
Pamela Bruen Docherty has more than
15 years experience in batch automation
and spent 12 years focusing on pharmaceutical production. She began her career
at Siemens Energy and Automation where
she worked five years as a batch product
specialist. She then worked at Merck & Co. in West Point,
Pennsylvania as an automation engineer supporting their
biologics pilot plant. After Merck, Bruen Docherty joined
Honeywell POMS where she provided Manufacturing Execution System (MES) technical demonstrations. She recently
returned to Siemens Industry as the Life Sciences Industry
Manager for the United States.
Gregory Ruklic, independent senior
computer systems compliance and technology specialist, formerly Pfizer Global Engineering and Wyeth Central Engineering,
has an extensive background with more
than 25 years in the biopharmaceuticals
industry in automation and integrated MES-related systems
design, validation and quality/compliance oversight. Working for engineering and quality unit organizations, Ruklic
has successfully implemented validated manufacturingrelated systems internationally, and has held leadership
positions in the development and implementation of site,
operating unit and corporate level systems standards and
guidance. He is the USA co-chair of the GAMP MES Special
Interest Group and contributing author to GAMP publications including GAMP 5. He has presented at US and
international forums on topics ranging from MES design
in support of flex-team based manufacturing, to systemic
(built-in) risk management.

JANUARY/FEBRUARY 2015

PHARMACEUTICAL ENGINEERING