An Improved Method of Failure Mode Analysis for Design Changes

1 2
R. Laurenti , H. Rozenfeld
1,2
Nucleus of Advanced Manufacturing, Engineering School of São Carlos,
Department of Production Engineering, University of São Paulo.
Av. Trabalhador Saocarlense, 400 13566-590 São Carlos-SP, Brazil
1
rlau@[Link]; laurentirafael@[Link]
2
roz@[Link]

Abstract
Customers and market changes behaviour are a large part of the product creation and modification.
Though, design changes introduce new potential failures into the products. In this paper, it is presented an
integrated method which turns attention to the analysis of design changes. The method is based on FMEA
and DRBFM methods, and on remarks from four focused interviews. The interviews showed the necessity of
a structured process of managing engineering changes, multidisciplinary work, empowerment of
responsibilities, committed personnel and understanding of the modifications. Yet, further work must be
undertaken to assess and validate the novel method.

Keywords:
Failure Analysis, Design Changes, FMEA, DRBFM

1 INTRODUCTION process. Nevertheless, as previously discussed, many

In the global market, product development has shown to defects are being discovered by the final consumer.
be one of the most important business processes for Moreover, the FMEA method has several shortcomings,
companies in the achievement of competitive advantage for instance, it does not take into account potential
[1]. Over the last decade, new products have been failures due to changes.
appearing at an ever increasing pace. Also, product Therefore, the goal of this paper is to advance an
modifications have increased significantly to meet existing integrated method to proactively find potential failures
needs, emerging wants, and latent expectations of introduced by design changes. The method was named
consumers [2]. Most new products in engineering are Failure Mode and Effect Analysis of Modifications
designed by modification from existing products, namely, (FMEAM). In this paper, it is adopted that the terms
product development involves the steady evolution of an change and modification have exactly the same meaning.
initial design [3] [4]. Also, the review emphasis is placed on failure analysis
However, changes always create an increased potential just for product design; not for manufacturing process.
failure in the design [5] [6]. The failures can affect the A combination of literature review, regarding the methods
reliability and availability of a product and can cause profit FMEA and Design Review Based on Failure Mode
loss to both manufacturer and user. This is particularly (DRBFM), and findings from four focused interviews were
true in the automobile industry. Many researches have the sources of evidence to define the proposed method.
shown that besides financial harm, disclosure of product The interviews not only helped to understand practical
defects (such as recalls) can result in negative abnormal analyses to avoid failures when a product is modified, but
results on the automakers reputation [7] [8], with also they provided substantial ideas of how the FMEAM
consequent losses in stock market valuation [8] [9] and could be integrated into the NPD process.
product sales [8] [10]. In a typical month, several recall
campaigns of motor vehicles are carried out by Design changes can be problematic because designers
automobile manufacturer to correct defective vehicles [9] are not always aware of the connectivity between the
and its incidence are increasing over the time [8]. different parts of a product and can inadvertently ignore
Likewise, the same evidences of profit losses were found the incidental effects of change [3]. Inspected failures due
for non-automotive recalls [11]. to change can be avoided through high redundancy in the
product or, more economically, through an intelligent
In this the scenario, companies have the challenge to anticipation of later failures due to the change [15]. Thus,
proactively prevent failures during the early stages of the the FMEAM should bring superior results to the NPD
new product development (NPD) process, since the later process.
a failure is detected into the product life cycle, the bigger
becomes its financial consequences [12]. Several The rest of this paper is organized as follows: section 2
methods of design failure analysis currently exist and are describes the research methodology; in section 3 a brief
used in industry, but by far the most widely used is the literature review is given, encompassing the methods
Failure Mode and Effect Analysis (FMEA) method [13]. FMEA and DRBFM; in section 4 the interviews are
described; section 5, which constitutes the major part of
The FMEA helps designers to understand and know the this paper, is where the FMEAM is presented and the
potential modes of a failure, to assess the risk of each procedure of carrying it out is explained; and section 6
known potential failure mode, and to identify concludes the paper by discussing the approach taken by
countermeasures to avoid the failure to occur [14]. It has FMEAM and pointing out further work.
been intensively applied over the years in the NPD

CIRP Design Conference 2009

 company. Furthermore, since it supports the early
2 METHODOLOGY detection of weaknesses in a design, a reduction of
development costs and fewer changes during series
Research approaches can be divided into the categories: production are expected [20].
quantitative, qualitative and mixed [16]. Also, they can be
classified as exploration, descriptive, predictive, and However, there are numerous shortcomings within the
explanatory research [17] [18] [19]. Exploration research failure analysis of FMEA, its implementation and utility.
involves [17] an attempt to determine whether or not a Some of these shortcomings include a lack of well-
phenomenon exists; i.e. Does X happen? Descriptive defined terms [21], problems with the terminology [22],
research involves [17] examining a phenomenon to more problems with identifying key failures [23] and it is treated
fully define it or to differentiate it from other phenomenon; as a stand-alone technique [20], which is neither
i.e. What is it? How is it different? Predictive research integrated with the design process, nor with other
involves [17] identifying relationships that enable us to methods of quality management. Other common
speculate about one thing by knowing about some other complaints of the FMEA method is that it is tedious and
thing; i.e. What is it related to? Explanatory research time consuming [13] [24] [25], its analyses are subjective
involves [17] examining a cause-effect relationship (based on the user’s experience) [26], it is considered by
between two or more phenomena; i.e. What causes it? the engineers to be ‘‘laborious’’ [20], the analysis is often
done to check rather than to predict [22].
The research approach and its source of data collection
have to be chosen according to the purpose of the When concerned with product design, it is important that
research. In this work, the research the approach taken failure analysis is carried out early in the design process
was qualitative and it can be classified as a descriptive in order to reduce the necessary amount of redesigns. It
research. is important to perform failure analysis in conceptual
design, but it has been reported that FMEA is commonly
Descriptive research requires that the investigator begin performed too late in the NPD cycle and has very little
with a descriptive theory. Accordingly, a review of effect on the overall product design [27].
literature about the methods FMEA and DRBFM was
undertaken. Then, four focused interviews were To overcome the FMEA shortcomings, many adaptations
conducted with product design stakeholders from different and improvements have been made to its process,
companies. Focused interview is used in a situation where application and target. Previous papers [25] [26] have
the respondent is interviewed for a short period of time, described concepts for automated FMEA employing
usually answering set questions [18]. qualitative simulators and reasoning process to produce a
report that is more timely, complete and consistent in the
The Interviews were carried out aiming at understanding design cycle. Other authors [28] take automated FMEA a
the practical analyses performed to avoid failures when a step further, developing a concept for analysis of the
product is modified. In all interviews the first quarter was effects of significant multiple failures as well as single
conversational to comprehend the scenario of the failures. A software that uses quantitative simulator has
company which the interviewee works for. Specific been developed [29], to produce results that are not only
questions dealt with how the analyses are done, their more accurate for designers, but are also more useful to
objectives, whether they are formal or informal, which are test and diagnostics engineers. Bayes belief networks
the employees involved, and whether the lessons learnt has been employed [21] to provide a language for design
are registered and further reused. Contextual questions teams to articulate, with greater precision and
probed which are the benefits and difficulties of analyses, consistency and less ambiguity, a physical system failure
the resources and efforts necessary to accomplish the cause-effect relationship, and the uncertainty about their
analyses. impact on customers. It has been shown [30] that
“function to structure mapping” can be used in the early
3 LITERATURE RESEARCH stages of design to assess diagnosability; i.e., a measure
of the ease of isolating the cause of a malfunction.
It is appropriate to present a brief of the literature review
about the methods FMEA and DRBFM. 3.2 DRBFM
3.1 FMEA, shortcomings and adaptations DRBFM is a method of discovering problems and
developing countermeasures by taking notice of and
Failure Mode and Effect Analysis (FMEA) is a quality
discussion intentional changes (design modifications) and
method that identifies, prioritizes, and mitigates potential
incidental changes (changes in part environment) [31]. It
problems in a given product. FMEA begins with
is carried out, throughout the NPD process, to guide the
identification of functions or requirements of a system,
design engineer during the engineering change process,
subsystem or component, ways that they can fail and its
to integrate design, production, quality and supplier
potential causes of failure. A small, but representative,
personnel into this process, and to achieve a robust
group with members of the design team and other
design [32].
disciplines familiar with the product life cycle performs the
analysis in one or more meetings. For each failure mode DRBFM was developed by Tatsuhiko Yoshimura, who
and cause, the team identifies the probability that they can has worked at Toyota Motor Corporation for 32 years. In
occur and scores them on a scale from 0 to 10. After the Japanese automobile manufacturer, Yoshimura was
identifying the effects, the analysis scores the severity of one of the responsible engineers to assure the quality
each end effect on a similar scale. The team documents and the reliability of the products, dedicating his
which actions have already been taken, and which actions professional life to avoid problems before they occur.
have still to be performed in order to avoid or to detect the However, the other employees acting as
failure mode. Finally, the detection rating scored refers to “troubleshooters”, namely, solving problems just when
the likeliness of catching the failure modes before they they appeared, were apparently the heroes of the
happen. The product of these terms is the risk priority company [33]. It has been reported [33] that the summary
number (RPN) which gives a relative magnitude for each for Yoshimura of this experience is similar than the
failure mode. findings of the study conducted by MIT: “Nobody Ever
Gets Credit for Fixing Problems that Never Happened”
If an FMEA is done properly, the resulting documents
[34].
contain a lot of knowledge about the product design.
Thus, it is a valuable source of know-how for the
 Good Design
robust design,
to reduce/notice changes To nip the problems in
the bud
Interface

Element Element
(component) (sub-system) Good Design

GD 3

Good Good Design

Discussion Review

Area of potential problems due design or

environmental changes

Systematic search of problems

Figure 1: GD3 philosophy.

Despite that, Toyota has successfully implemented the However, the company’s major difficulty is the short time
DRBFM method. Nonetheless, the method has not yet to test the new machine before sending it to the client.
deeply investigated by the academic researchers nor For this reason, sometimes, after a couple of months,
broadly disseminated throughout other companies. when they have already sold a few machines of the new
A recent study [8] showed that Toyota has the highest model, a design error can be found and they need to
product variety of all other car manufacturers and has low make a recall. It was said by the interviewee that this is a
volume of recalled vehicles as a percentage of better situation than not sending the machine to the client
registrations. This suggests an ability on Toyota’s part to and losing the sale.
simultaneously offer relatively high variety whilst retaining Lessons learnt are registered in a validation report for
tight control of product development and manufacturing further sharing, but this has not shown to be efficient. So,
processes [8]. most of the lessons learnt are verbally shared.
3
Toyota has a philosophy called GD or Mizen Boushi 4.2 Second interview
(roughly translated as “countermeasures”). GD3 stands for
Good Design, Good Discussion and Good Design Review The second interviewee is a project manager from a
[35]. The principles of Good Design are to use, as much medium-sized manufacturer of medical-ophthalmic
as possible, robust components and avoid design equipments. There, the “engineering change
changes to reduce the complexity of the error prevention. management” process is systematically structured and
Besides, this principle tries to nip the problems in the bud. controlled. All the analyses of design changes are
Good Discussion and Good Design Review are the coordinated by an improvement group. First, they
processes of thoroughly discussing design plans to document who requested the change, the type of change
discover previously undetected problems and they are (structural, material, treatment, etc.), the reason (cost
used to formulate the best countermeasures to these reduction, upgrade, improvement of technology, etc.),
3
problems [31] [5]. The GD philosophy is represented in product documents (drawings, procedures, lists, etc.) that
figure 1. In Toyota, the DRBFM method is applied in this may need to be modified, and the “in charge” of the
contest. modification. Then, the group judges the risk of the
change. If there is a risk, a change process is started;
otherwise, no analysis is needed. The improvement group
4 THE INTERVIEWS empowers a multidisciplinary team (mechanical and
Results from the interviews and respective description of electrical specialists) to assess the potential failures
companies are summarised in the following subsections introduced by the modification. For that, the team uses
the methods FMEA and FMECA (Failure Mode, Effects
4.1 First interview and Criticality Analysis) complemented with CAD
The first interview was performed with the design (Computer Aided Design) and FEM (Finite Element
engineering manager from a medium firm, which Method) analyses, followed by several tests executed in
manufactures machines and equipments for plastic functional prototypes and discussions within the members
transformation. They manufacture two segments of of the multidisciplinary team. After the tests, the new
machines: for bag making - automatic machines for small design is approved or declined by the project manager
and large plastic bags, T-shirt bags, rounded-bottom (technical responsible).
bags, sleeve labels pre-perforated or not; and for It was said that recalls do not occur due to modifications,
thermoforming and vacuum forming - machines for one- but rather to upgrades and technology improvements.
way packaging, i.e., pots, lids and trays. Also, it was stated that before implementation of the
When a family of machine is required to be modified, a design changes management process, requests for
design engineering group first creates a virtual model and modifications were constants and they diminished just
performs virtual tests. Then, a complementary analysis of after the implementation. Nonetheless, it was added that
failure is done based upon knowledge of the group; the beginning was difficult for the employees to aware
therefore, there is the need of competent employees. The themselves that it was necessary to work in group.
analyses are done to verify the novel design and to
prevent failures from happening in field.
4.3 Third interview manufacturing process and less due to design
modifications.
A third interview was done with the research and
development manager from a medium-sized company. 4.5 Findings
The enterprise develops equipments for automation to a Results from the interviews not only confirm that design
broad range of segments, such as, ethanol, pulp and modifications introduce potential failures into the product,
paper, chemical, metallurgical and agricultural. but also suggest the necessity of the following:
They have a structured and formal process for • Multidisciplinary work;
engineering change. The process can be initiated by
claims from costumers or when internal personnel or the • Structured process of managing engineering changes;
field technical support give suggestions of improvements. • Empowerment of responsibilities;
Subsequently, the manager and the head of the research
and development department carry out a critical analysis • Committed personnel;
identifying impacts (in the company) of the modifications, • Understanding of the modifications.
risks involved, necessary resources and time, and level of
difficulty for implementing it. Then, if the modification is
approved, a plan is developed which includes necessary 5 DESIGN FAILURE MODE AND EFFECT ANALYSIS
personnel, schedule of alterations, and needed tests. OF MODIFICATIONS (FMEAM) PROPOSAL
Tests are performed by groups of software and hardware FMEAM was developed based on literature about FMEA
specialists in different modules. The groups can consult a and DRBFM methods, and remarks from the focused
database of previous problems. After the tests, the interviews. FMEAM attempts to find any potential failures
engineering development manager gathers the tests introduced into the product due to modifications.
results and approves the modification or not. FMEAM aims to encourage creative discussion of even
It was said that the analysis is carried out to clearly more issues than FMEA does, by stimulating each other’s
identify what is needed to be done, and to become known brains to make one another notice things, thus preventing
the possible failures in the functionality of the equipment problems. FMEAM links design and evaluation in order to
due to the change. The benefits related were: higher promote an integrated prevention of failures.
quality and reliability of products, more efficient work It should be seen as a live document integrated with the
(focus on necessary work), superior customer satisfaction NPD process, rather than just a task to be done after the
and decline of reworks. design is completed. Namely, its documentation shall be
A good and fast product development depends on the constantly updated with data from the results of
company’s labour force. Therefore, the interviewee evaluation and field.
pointed out that the main difficulty is to have skilled labour 5.1 Conducting FMEAM in a integrated manner
to correctly perform the analyses.
Most current FMEA procedures are a one-time task done
Besides, the interviewee affirmed that a product in the phase of Testing and Validation, which may serve
modification causes higher probability of failure, and as an important design check, but otherwise, it
added that their recalls are associated with a design contributes little to the design. The proposal here is a
modification. Although a pressure to quickly launch the series of activities to be done throughout the NPD
novel product into the market exists, it is better to have a process.
robust product, rather than a weak product with a client.
During the Conceptual Design phase, data, such as field
Owing that, they have spent larger time with the tests.
reports, checklists and other guidelines based on lessons
4.4 Fourth interview learned, technology advances, and the history or analysis
The fourth interview was carried out in a local site of a of similar designs, which has proven to be successful,
large global manufacturer of hermetic compressors for air should be collected and block diagrams developed - to
conditioning and refrigeration products and centrifugal illustrate the physical and functional relationships
pumps. They produce hermetic compressors for domestic between items and interfaces within the system.
and commercial refrigeration and air conditioners. In the Detailed Design phase a FMEAM meeting shall be
An interview was conducted with the manager of product conducted just prior of the release of final drawings. The
engineering group. When asked how failure analyses of following tolls, regarding the concerned system of the
“new” and “variant” products were done there, it was analysis, must be provided beforehand of the FMEAM
answered straight away “through the FMEA method”. He meeting:
added saying “a design error discovered by the client can • Block diagram: to show interdependencies of
cause us large financial harm”, consequently he functional entities and interfaces.
recognizes the necessity of applying FMEA. • Fault tree analysis (FTA): to show logical relationship
In the company, the FMEA is used to check the design between a failure and its causes, and to provide a
using nonfunctional prototypes to help visualization. This logical framework for expressing combinations of
is done during a meeting promoted by the product component failures that can lead to system failure.
designer (leader), who summons leaders from the quality • Change point list: a list of what has been changed or
departments, manufacturing process, supply chain (if intended to be changed in the components of the
supplier is involved) and research and development (if it is system. This list aims to clearly identify and organize
a new product). However, it is common for the leaders not the changes. Examples of change point could be
to attend the meeting and also to lack knowledge about change in structure, material, surface, thermal
what was modified in the product. Thus, as the meetings treatment, manufacturing process and stress/load.
are inefficient, many of them are needed to accomplish
the necessary work. • Relational matrix between functions and components:
a matrix which correlates the components and its
In the end, it was answered that the lessons learnt are
functions. It intends to clarify failure modes by
documented, yet are not retrieved in further analyses. The
identifying which function is affected.
interviewee concluded saying that their recalls are not
frequents, mainly occurring due to failures from the • Previous drawings and prototypes (if available): to
help visualization.
• History of field failures: to prevent a known problem Engineers from design, evaluation, production
from occurring again. engineering and manufacturing, inspection and material
The tools should be used throughout the FMEAM departments shall be selected for the meeting. Yet,
meeting. They are meant to identify the intentional experienced engineers with the intention of getting
changes made and also the changes resulting from them. actively involved in the discussion by putting themselves
In addition, they should help to visualize the system in the place of the design leader should be selected.
structures and the functions of components during the In the FMEAM meeting, initially, the design engineer
FMEAM meeting. (leader) should explain the mechanism and the functions
Results from the discussion of the FMEAM meeting shall of the concerned part, a general idea of its design, and
be reflected in the Testing and Validation phase. In any specific factors that have been given special
addition, redesigns may occur in the NPD process. consideration. The participants then should ask questions
Therefore, those results should be also reflected in and/or make remarks about any matter of concern in
redesigns in a timely manner. relation to the given explanation, and discuss with the
design engineer based on the raised questions and
Finally, during the Product Use and Support phase, indications. It is a way of ensuring thorough mutual
FMEAM should be used as a guide to collect field data for understanding among FMEAM team in relation to the
assessing analysis accuracy, and for developing concerned part.
maintenance troubleshooting procedures.
Afterwards, the discussion is conducted by filling out the
Conducting the FMEAM in this manner it shall enforce a FMEAM worksheet sections. The following elucidates all
disciplined review of the baseline design and may allow the headings of the worksheet.
timely feedback to the design process.
1. Component name / change: enter the name of the
5.2 Adaptations made in the traditional FMEA component subject to analysis. Enter the modification
worksheet made or planned to be made in it and the details of
The headings of the FMEA table were modified. The the modification. In the case of a newly adopted part,
scores for severity, occurrence, detection, and the risk it is preferred to have a comparable part on hand, if
priority number (RPN) were replaced by “Adverse effects possible, for the purpose of comparative evaluation.
on customer” with a scale of three levels: A, B, C (A being 2. Function: enter the function intended for the subject of
the most important). The "customer", as referred to for the analysis by as concisely as possible. If there are more
purpose of FMEAM, includes not only the "end users" but than two intended functions, enter all such functions
also the design staff and teams of the subsequent separately.
processes and the engineers operating in such fields as 3. Potential failure mode due to change: enter as how
the production, assembly, service, etc. the component will fail as a result of the modification.
The most relevant section inserted was “Discussion If there are more than two intended functions, make
Results”, since the value added by FMEAM activity separate entries for each of the functions. Also make
depends on the extent that new and specific items are entries regarding the factors that cause the loss of
identified and entered into this section. The section commercial value, such as abnormal noise and poor
encompasses “Items to reflect in design work” and “Items appearance quality, in a concrete expression phrased
to reflect in evaluation work”, each of them followed by a from the customer's point of view.
column “Responsibility and deadline”, where the person 4. Root cause / dominant cause: the causes of the
responsible to follow-up the action and the deadline for its failure of function and commercial value are, in a
implementation is entered. sense, the weak-points of the current design. Indicate,
Table 1 shows the FMEAM headings. therefore, the root causes of the failure and loss of
commercial value as concretely as possible to
5.3 Performing FMEAM facilitate the future implementation of the measures
FMEAM has to be performed in a meeting by a formulated from design perspective and to allow
multidisciplinary team to take advantage of sharing their participants to clearly visualize the concern.
specific knowledge. The design engineer responsible for
the system (part), which will be analysed, should lead the
meeting.

Active discussion about the design changes

Importance

Current design to avoid

Component Potential Root cause / Adverse effects
concern point (incl. design
name / Function failure mode dominant on customer
rule, design standard &
change due to change cause (System)
check items

Discussion Results

Action Results
Items to reflect in Resp. & Items to be reflected in Resp. &
design work Dead Line evaluation work Dead Line

Table 1: FMEAM headings.

5. Adverse effects on customers: customers, as Neither isolated nor just for design checking the FMEAM
explained in the previous section, can be assumed to should be applied. Furthermore, its practice should
exist in various levels. In this column, indicate the comprise a continuum application within and between
phenomena, which customers will presumably NPD processes in order to constantly be detected
experience, using expressions phrased from potential failures and to create a historic of potential
customer's point of view. So, the type of adverse failures due to changes.
effects can be clearly understood. Classify the FMEAM should promote identification of potential
adverse effects into three ranks - A, B or C - according problems through active discussion of modifications and
to the severity of impacts and enter the rank in the the causes of such problems. Consequently, it is vital that
"Importance" column. all participants of the discussion have a good
6. Current design to avoid concern point: enter the understanding of the substance of the modifications.
considerations made in the design to prevent the Although FMEAM method was based on stabilised
failure and loss of commercial value. Entries shall be methods, certainly, it is necessary to take further action
made in such a manner so that the information in this research approaches, in order to establish the feasibility,
column can be readily associated with the relevant usability and utility of the new method.
information in the "Root cause / dominant cause"
column. In Summary, the intention of the proposed method is to
meet the current necessity of companies, which are to
7. Items decided in FMEAM to reflect in design work and launch novel products into the market in shorter cycles
to reflect in evaluation work: the participants are to and to effectively assure the quality of its new products.
make attempts to identify potential problems through
discussion, and then summarize the decisions made
into these columns using concise expressions. Enter 7 ACKNOWLEDGMENTS
those items that should be considered in the design The authors are grateful to the companies for
work into the "Items to reflect in design work" column volunteering to take part in this research, to NUMA
and those items to be considered during the test and colleagues and to CNPq for the financial support.
evaluation work, such as in relation to the conditions
and items of evaluation, into the "Items to reflect in
evaluation work" column. Be sure to provide clear and 8 REFERENCES
readily understandable instructions on what needs to [1] Unger, D. W., Eppinger, S. D. 2006, Improving
be done. Enter the name of responsible employee and Product Development Processes to Manage
the deadline of implementation into the "Responsibility Development Risk, MIT Sloan Research.
and deadline" column, for each of the Items decided. [2] Lauglaug, A. S. 1993, Technical Market Research:
8. Action results: For each items listed in the previous Get Customers to Collaborate in Developing
columns (7.), enter the information regarding the date Products, Long Range Planning, 26(2): 78-82.
of implementation of the measures, details of [3] Clarkson, P. J., Simons, C., Eckert, C. 2004,
measures implemented, and the consequences of the Predicting Change Propagation in Complex Design,
implementation. Journal of Mechanical Design, 126(5): 788-797.
Finally, a FMEAM report shall be written to summarize the [4] Gerst, M., Eckert, C., Clarkson, J., Lindemann, U.
results, reports the decisions and recommended actions 2001, Innovation in the Tension of Change and
to be taken for elimination or reduction of failure. Also, it Reuse, Proceedings of the 13th International
should include the problems which could not be corrected Conference on Engineering Design: Design
by design. Since FMEAM is intended to be a living Research – Theories, Methodologies and Product
document, this report shall be integrated in the NPD Modelling, Professional Engineering Publishing,
process as a deliverable to the design and evaluation 371-378.
work.
[5] Schmitt, R., Krippner, D., Hense, K., Schulz, T.
2007, Keine Angst vor Änderungen! Robustes
6 FINAL REMARKS Design für Innovative Produkte, Qualität und
Failure Mode and Effect Analysis (FMEA) has been Zuverlässigkeit, 52(03): 24-26.
intensely used to ensure the quality and reliability of [6] Chao, L. P., Ishii, K. 2007, Design Process Error
products. Additionally, over the years, several adaptations Proofing: Failure Modes and Effects Analysis of the
and improvements have been made in the method FMEA. Design Process, Journal of Mechanical Design,
Besides the use of the FMEA and the several attempts to 129(5): 491-551.
mitigate its shortcomings, many incidents in the field are [7] Rhee, M., Haunschild, P. R. 2006, The Liability of
still occurring, which is costly to companies. Good Reputation: A Study of Product Recalls in the
This paper has presented a method to analyzing the U.S. Automobile Industry, Organization Science,
effects of changes made in the design. The method was 17(1): 101-117.
conceived from the knowledge that changes carry a [8] Bates, H., Holweg, M., Lewis, M., Oliver, N. 2007,
higher potential failure. It is based on the methods FMEA Motor vehicle recalls: Trends, patterns and
and DRBFM and on findings from focused interviews. The emerging issues, Omega, 35(2): 202-210.
novel method was called Failure Mode and Effect
Analysis of Modifications (FMEAM) and it aims to assure [9] Barber, B. M., Darrough, M. N. 1996, Product
the product quality after design changes. Reliability and Firm Value: The Experience of
American and Japanese Automakers, 1973-1992,
The interviews indicated that when carrying out analyses Journal of Political Economy, 104(5): 1084-1099.
of engineering changes, it should be a structured process.
Additionally, it was accentuated that there is a need of [10] Haunschild, P. R., Rhee, M. 2004, The Role of
multidisciplinary work group, empowerment of Volition in Organizational Learning: The Case of
responsibilities and personnel committed with the work Automotive Product Recalls, Management Science,
and aware of the modifications. 50(11): 1545-1560.
[11] Davidson, I., Worrell, D. L. 1992, Research Notes
and Communications: The Effect of Product Recall
 Announcements on Shareholder Wealth, Strategic [29] Montgomery, T., Marko, K. 1997, Quantitative FMEA
Management Journal, 13(6): 467-473. Automation, Proceedings of the Annual Reliability
[12] Schmitt, R., Krippner, D., Betzold, M. 2006, and Maintainability Symposium, IEEE press,
Geringere Fehlerkosten – höhere Zuverlässigkeit, Philadelphia, USA, 13-16 January, 226-228.
Qualität und Zuverlässigkeit, 51(06): 66-68. [30] Clark, G. E., Paasch, R. K. 1996, Diagnostic
[13] Stone, R., Tumer, I., Stock, M. 2005, Linking Product Modelling and Diagnosability Evaluation of
Functionality to Historic Failures to Improve Failure Mechanical Systems, Journal of Mechanical Design,
Analysis in Design, Research in Engineering Design, 118(3): 425-431.
16(1): 96-108. [31] Shimizu, H., Imagawa, T., Noguchi, H. 2003,
[14] Stamatis, D. H. 1995, Failure Mode and Effect Reliability Problem Prevention Method for
Analysis: FMEA from Theory to Execution, ASQC Automotive Components - Development of GD3
Quality Press. Activity and DRBFM (Design Review Based Failure
Mode), Proceedings of the International Body
[15] Eckert, C., Zanker, W., Clarkson, P. J. 2001, Engineering Conference, SAE International, Chiba,
Aspects of a Better Understanding of Changes, Japan, 371-376.
International Conference on Engineering Design,
Glasgow, UK, 21-23 August. [32] Schorn, M., Kapust, A. 2005a, Im Fluss: Wie Toyota
von DRBFM Profitiert, Qualität und Zuverlässigkeit,
[16] Creswell, J. 2003, Research design: Qualitative, 50(04): 56-58.
Quantitative, and Mixed Methods Approaches,
Thousand Oaks, USA, Sage. [33] Schorn, M. 2005, Entwicklung mit System: Wie
Toyota von DRBFM Profitiert, Management und
[17] Dane, F. C. 1990, Research Methods, Pacific Grove, Qualität, 12, 8-11.
USA, Brooks/Cole.
[34] Repenning, N. P., Sterman, J. D. 2001, Nobody
[18] Yin, R. K. 1994, Case Study Research: Design and Ever Gets Credit for Fixing Problems that Never
Methods, Thousand Oaks, USA, Sage. Happened: Creating and Sustaining Process
[19] Marshall, C., Rossman, G. B. 1995, Designing Improvement, California Management Review,
Qualitative Research, Thousand Oaks, USA, Sage. 43(4): 64-88.
[20] Wirth, R., Berthold, B., Krämer, A., Peter, G. 1996, [35] Schorn, M., Kapust, A. 2005b, DRBFM - die Toyota
Knowledge-based Support of System Analysis for Methode, Vdi Z Integrierte Produktion, 147(7/8), 67-
the Analysis of Failure Modes and Effects, 69.
Engineering Applications of Artificial Intelligence,
9(3): 219-229.
[21] Lee, B. 2001, Using Bayes Belief Networks in
Industrial FMEA Modelling and Analysis,
Proceedings of the Annual Reliability and
Maintainability Symposium, IEEE press,
Philadelphia, USA, 7-15.
[22] Kara-Zaitri, C., Keller, A., Barody, I., Fleming, P.
1991, An Improved FMEA Methodology,
Proceedings of the Annual Reliability and
Maintainability Symposium, IEEE press, Orlando,
USA, 248-252.
[23] Bednarz, S., Marriott, D. 1988, Efficient analysis for
FMEA, Proceedings of the Annual Reliability and
Maintainability Symposium, IEEE press, Los
Angeles, USA, 26-28 January, 416-421.
[24] Hunt, J. E., Pugh, D. R., Price, C. J. 1995, Failure
Mode Effect Analysis: A Practical Application of
Functional Modelling, Applied Artificial Intelligence,
9(1): 33-44.
[25] Price, C., Pugh, D., Wilson, M., Snooke, N. 1995,
The Flame System: Automating Electrical Failure
Mode and Effects Analysis (FMEA), Proceedings of
the Annual Reliability and Maintainability
Symposium, IEEE press, Washington, USA, 16-19
January, 90-95.
[26] Bell, D., Cox, L., Jackson, S., Schaefer, P. 1992,
Using Causal Reasoning for Automated Failure
Modes and Effects Analysis (FMEA), Proceedings of
the Annual Reliability and Maintainability
Symposium, IEEE press, Las Vegas, USA, 21-23
January, 343-353.
[27] McKinney, B. 1991, FMECA, The Right Way,
Proceedings of the Annual Reliability and
Maintainability Symposium, IEEE press, Orlando,
USA, 29-31 January, 253-259.
[28] Price, C. J., Taylor, N. S. 2002, Automated Multiple
Failure FMEA, Reliability Engineering and System
Safety, 76(1): 1-10.