Tutorial
Tutorial
Tutorial
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
2012 Annual RELIABILITY and MAINTAINABILITY Symposium<br />
Closed-Loop Corrective Action Basics,<br />
Best Practices and Application<br />
Brad Cline<br />
PTC<br />
Services Manager<br />
Windchill Quality Solutions<br />
41 W. Otterman Street<br />
Suite 310<br />
Greensburg, PA 15601 USA<br />
Internet (e-mail): bcline@ptc.com<br />
Brad Cline & Ken Stillwell<br />
<strong>Tutorial</strong> Notes © 2012 AR&MS<br />
Ken Stillwell<br />
PTC<br />
Vice President, Strategy & Operations<br />
Windchill Quality Solutions<br />
41 W. Otterman Street<br />
Suite 310<br />
Greensburg, PA 15601 USA<br />
Internet (e-mail): kstillwell@ptc.com
SUMMARY & PURPOSE<br />
This tutorial introduces basic information about closed loop corrective action processes. There are many names for the<br />
closed loop corrective action process, including FRACAS (Failure Reporting, Analysis, and Corrective Action System), CAPA<br />
(Corrective and Preventive Action), PRACA (Problem Reporting, Analysis, and Corrective Action), and others. These processes<br />
bring together a wide range of information, from test results to field data to repair records. Outputs can include both qualitative<br />
and quantitative results that are inherently customizable to the specific needs of the organization. This tutorial also addresses<br />
several of the key obstacles to deriving a successful closed loop corrective action process as well as best practice suggestions<br />
and a proven methodology for realizing the full benefits of a lucrative closed loop corrective action process, including<br />
improvements in quality, reliability, and productivity along with reduced costs. To conclude the tutorial, several case studies are<br />
presented that highlight success stories from a wide range of groups who have implemented effective closed loop corrective<br />
action processes.<br />
Brad Cline<br />
Brad Cline (Services Manager - PTC’s Windchill Quality Solutions) manages enterprise-level implementations, from<br />
requirements definition and design to system installation, with over 50 large-scale FRACAS implementations globally. Brad<br />
also manages consulting services, which functions as a full-service reliability engineering support center. With a BS - Applied<br />
Mathematics - Carnegie Mellon University and MBA - University of Pittsburgh, Brad has experience managing data<br />
warehousing implementations, data analysis activities, and web-based enterprise reporting tools.<br />
Ken Stillwell<br />
Ken Stillwell (Vice President – Strategy and Operations) is responsible for driving efficiency in operations and overall<br />
growth strategies for PTC’s Windchill Quality Solutions, and leads customer support, global education, and global services. Ken<br />
oversees enterprise-level FRACAS implementations, with more than 150 FRACAS deployments completed by the<br />
implementation team. With a BS - Business/Economics - University of Pittsburgh and MBA - University of South Carolina, Ken<br />
has held executive positions in software/IT, biotech and aerospace.<br />
Table of Contents<br />
1. Introduction .......................................................................................................................................................................... 1<br />
2. FRACAS Fundmentals ......................................................................................................................................................... 1<br />
3. FRACAS Best Practices ....................................................................................................................................................... 2<br />
4. Challenges to Effective FRACAS ........................................................................................................................................ 3<br />
5. Steps t Successful FRACAS Implementation ...................................................................................................................... 4<br />
6. Practical Application ............................................................................................................................................................ 6<br />
7. Conclusions .......................................................................................................................................................................... 8<br />
8. References ............................................................................................................................................................................ 8<br />
9. <strong>Tutorial</strong> Visuals…………………………………………………………………………………….. ................................... 9<br />
ii – Cline & Stillwell 2012 AR&MS <strong>Tutorial</strong> Notes
1. INTRODUCTION<br />
To produce high-quality products in an ever more<br />
competitive marketplace, the ability to efficiently discover,<br />
track, and correct incidents or failures found during product<br />
development, testing, and operation is crucial. This need spans<br />
all industries and classifications, including hardware,<br />
software, process management, and services in both<br />
government and commercial sectors. Notably, a survey<br />
conducted by the former Reliability Analysis Center (RAC),<br />
now the Reliability Information Analysis Center (RIAC),<br />
indicated that companies view this subject, most commonly<br />
referred to as FRACAS (Failure Reporting, Analysis and<br />
Corrective Action System), to be among their top two most<br />
important reliability tasks (1). Additionally, as of mid-March<br />
2011, as per Directive-Type Memorandum (DTM) 11-003 –<br />
Reliability Analysis, Planning, Tracking, and Reporting, as<br />
part of a comprehensive reliability and maintainability (R&M)<br />
program, a failure reporting and corrective action system is<br />
required to be maintained through design, development,<br />
production, and sustainment.<br />
Manufacturers and service providers have employed a<br />
variety of methods for tracking product failures and managing<br />
corrective actions. These systems include paper-based<br />
approaches, electronic spreadsheets and personal databases, or<br />
even large enterprise systems that affect hundreds or<br />
thousands of customers, suppliers, and engineers worldwide.<br />
Conceptually, the benefits of a failure tracking and corrective<br />
action process are clear. However, implementing such a<br />
system that yields more reliable products and improved<br />
designs in an efficient and effective manner is complex.<br />
Different groups have attempted to define structured<br />
approaches and guidelines for the implementation of failure<br />
reporting and corrective action systems. Organizations such as<br />
the RIAC, in accordance with the U.S. Department of Defense<br />
(DoD), have studied the requirements of both large and small<br />
companies to develop general guidelines for FRACAS. In<br />
addition to the RIAC guidelines, other similar efforts have<br />
been published for public use, including SEMATECH’s<br />
“Failure Reporting, Analysis and Corrective Action System”<br />
Planning and Implementation Guide (2) and NASA’s<br />
“Preferred Reliability Practices, Problem Reporting and<br />
Corrective Action System” documents (3). While most<br />
government-related initiatives commonly refer to corrective<br />
action systems to as FRACAS, other names, including<br />
PRACA, Quality System, Corrective Action System, and<br />
others abound. Regardless of what the systems are called,<br />
these published guidelines present process-based approaches<br />
to solving the fundamental problem of product improvement<br />
through failure recording, corrective action, and lessons<br />
learned.<br />
Even with existing guidelines and advancements in<br />
computer and communications technologies, the<br />
implementation of a cost-effective FRACAS that produces<br />
more reliable products remains a challenging task. Having<br />
examined the deployment of FRACAS in a variety of<br />
industries, the authors have determined that the publicly-<br />
available guidelines fail to address many of the practical<br />
complications that exist. These intricacies are neither processrelated<br />
nor technical in nature but instead stem from each<br />
company’s unique organizational structure, goals, and<br />
expectations. With this tutorial, the authors seek to provide<br />
additional practical guidance for implementation of a<br />
FRACAS, considering best practices and a step-by-step<br />
process for a successful FRACAS.<br />
2. FRACAS FUNDAMENTALS<br />
2.1 What is FRACAS?<br />
As defined earlier, FRACAS stands for Failure Reporting,<br />
Analysis, and Corrective Action System. The FRACAS<br />
process allows efficient tracking, analysis, and correction of<br />
problems found during product development, testing, and<br />
operation. Though most commonly used with fielded<br />
products, FRACAS can be applied to products, services,<br />
processes, and software applications thanks to its flexibility. A<br />
FRACAS is ultimately a closed loop system that improves the<br />
reliability of the product, service, process, or software<br />
application by focusing on failure reporting and resolution.<br />
Naming conventions vary widely. Variations include CAPA<br />
(Corrective and Preventive Action) System, Corrective Action<br />
System, Quality System, and others where the first letter of<br />
the FRACAS acronym is replaced by a “P” for Problem<br />
(PRACAS), “I” for Incident (IRACAS), and “D” for Data or<br />
Defect (DRACAS).<br />
2.2 Why is FRACAS important?<br />
FRACAS implementations track, analyze, and correct<br />
problems found during development, testing, and operation,<br />
with the goal of improving the reliability and quality of its<br />
target product. In many instances, FRACAS also identifies<br />
trends related to failures and prioritizes the top issues to be<br />
addressed. In any specific implementation, the reason for<br />
employing the FRACAS process should be identified clearly<br />
as the process is defined. In some cases, specific customers<br />
may require using FRACAS explicitly.<br />
2.3 Who uses FRACAS?<br />
The need for FRACAS spans all industries and<br />
classifications, including hardware, software, process<br />
management, and services, in both government and<br />
commercial sectors. Many industries use it extensively,<br />
including:<br />
• Aerospace<br />
• Automotive<br />
• Defense<br />
• Electronics<br />
2.4 What are the results of a FRACAS?<br />
• Manufacturing<br />
• Telecommunications<br />
• Medical Devices<br />
• Others<br />
Depending on the data collected in a FRACAS process,<br />
users can often generate quantitative results such as failure<br />
rate, mean time between failures (MTBF), mean time between<br />
critical failures (MTBCF), reliability, availability, and other<br />
results. FRACAS outputs also regularly include graphs and<br />
2012 Annual RELIABILITY and MAINTAINABILITY Symposium Cline & Stillwell – 1
eports to identify important summary details visually based<br />
on the data collected.<br />
2.5 How is FRACAS performed?<br />
While specifics will vary, FRACAS processes generally<br />
include these steps:<br />
1. Record the failures or incidents. Critical data associated<br />
with each failure or incident is recorded and stored under<br />
defined procedures, often in a database management<br />
system.<br />
2. Analyze the reported failures or incidents. The root cause<br />
of the failure or incident is identified and stored in the<br />
database management system alongside the original data.<br />
3. Identify necessary corrective action. A corrective action<br />
plan for mitigating the failure or incident is developed,<br />
implemented, and stored in the database management<br />
system.<br />
4. Verify the corrective action. Finally, the effectiveness of<br />
the corrective action is reviewed and recorded in the<br />
database management system and the incident or problem<br />
is closed out per established procedures.<br />
2.6 What are the limitations of FRACAS?<br />
As much as FRACAS can substantially improve products,<br />
services, processes, or computer applications, the difficulty of<br />
efficient application is its inherent limitation. Smart planning<br />
and execution during implementation is paramount, otherwise<br />
the system may fail to manage data efficiently, effectively<br />
identify root causes of problems, or close the FRACAS<br />
process loop correctly.<br />
2.7 Where can I learn more about FRACAS?<br />
As noted in the introduction, many FRACAS resources<br />
are available, including some that define structured<br />
approaches and guidelines for implementation. You can find<br />
several key FRACAS resources listed in the REFERENCES<br />
section.<br />
3. FRACAS BEST PRACTICES<br />
While the general closed loop corrective action process<br />
seems to follow a common sense approach, many factors can<br />
impede its application. Because of the potential challenges, the<br />
following best practices are strongly recommended for use<br />
during FRACAS implementation.<br />
3.1 Set expectations and goals<br />
Prior to deploying the FRACAS corrective action system,<br />
well-defined expectations and goals are essential. These<br />
include the roles and responsibilities of all stakeholders in the<br />
process as well as the objectives of the FRACAS system itself.<br />
All key stakeholders in the process must agree upon the<br />
explicitly defined set of clear expectations and goals. With<br />
these in place, the FRACAS implementation can progress with<br />
clarity of purpose and the system can deploy efficiently and<br />
effectively in line with the defined objectives.<br />
3.2 Involve the stakeholders<br />
The support and involvement of all FRACAS<br />
stakeholders is critical. Many of these stakeholders will<br />
originate within the organization, but customers and/or<br />
suppliers may be involved as well. Involving all appropriate<br />
parties leads to support for gathering sufficient failure data<br />
and unification of a common process across the whole<br />
organization.<br />
3.3 Gain active management involvement<br />
Management involvement and support has a strong<br />
impact on the success of the FRACAS. Active management<br />
participation often results in obtaining and maintaining<br />
necessary funding and resources, and may also provide the<br />
leadership needed to implement and maintain a successful<br />
FRACAS.<br />
3.4 Keep the process simple<br />
The most successful FRACAS solutions are easy to use,<br />
employ user-friendly software tools for automation, and<br />
overall require modest investments in resources and training.<br />
By keeping things simple, active participation from those<br />
outside the quality/reliability is more likely. Ultimately, the<br />
FRACAS must be simple enough for both the expert and<br />
novice to use. Of course, the process design must also allow<br />
for a thorough and effective FRACAS.<br />
3.5 Leverage software tools<br />
Software, either a custom in-house tool or customizable<br />
off-the-shelf solution, is one key way to help to automate the<br />
FRACAS and make it easier to use. Software tools help to<br />
automate data entry, analysis, and output, and provide a<br />
central storage area for FRACAS data and results.<br />
3.6 Provide for efficient data entry and analysis<br />
Entering and analyzing data can be two of the most time<br />
consuming tasks for FRACAS users. Simple web-based forms<br />
can provide for efficient data entry, while automated<br />
calculations, graphs, and reports, along with the ability to<br />
filter the data, can increase the efficiency and effectiveness of<br />
data analysis.<br />
3.7 Supply training<br />
Even when the simplest FRACAS process is<br />
implemented, early training can alleviate the stakeholders’<br />
concerns and foster active participation. As users generate<br />
feedback and the FRACAS evolves, providing additional<br />
training is beneficial for the same reasons.<br />
3.8 Encourage and supply feedback<br />
There are two types of desirable feedback. Feedback from<br />
the users of the FRACAS can help those in leadership roles<br />
focus and streamline the system, while feedback from those in<br />
leadership roles to all participants showcases the results of<br />
their hard work and provides encouragement.<br />
2 – Cline & Stillwell 2012 AR&MS <strong>Tutorial</strong> Notes
4. CHALLENGES TO EFFECTIVE FRACAS<br />
4.1 Challenge 1: Complex organization interaction<br />
A proper FRACAS process can span many different<br />
functional groups within an organization. Data is contributed<br />
by and/or needs to be made available to the following groups:<br />
• Manufacturing • Engineering<br />
• Operations<br />
• Quality Assurance<br />
• Reliability<br />
• Customer Service<br />
• Sales and Marketing • Field Services<br />
• Testing<br />
• Suppliers<br />
• Failure Review • Maintenance<br />
Board (FRB)<br />
• Others<br />
With all these different departments involved in the<br />
process, their interaction can become complex. Some groups<br />
may want to be included at multiple points. This increases and<br />
complicates the steps required to close an incident.<br />
Consider the following scenario of how this may occur. In<br />
this simple process, a field service representative reports or<br />
logs an incident. Quality Assurance then reviews this incident<br />
and assigns it to the Reliability group to perform a root cause<br />
analysis. The Reliability group recommends a corrective<br />
action that the Engineering group implements. Finally the<br />
Failure Review Board approves the corrective action and<br />
closes out the incident.<br />
Over time, additional groups request to be involved. For<br />
example, Customer Service believes that they too need to be<br />
part of the corrective action step. Because they directly<br />
interface with the customer, they want to understand and<br />
possibly shape the corrective action response that may occur.<br />
In addition, Sales and Marketing may also be needed to<br />
properly manage a customer’s account, and a Supplier may<br />
desire to be connected with the analysis and corrective actions<br />
that have occurred with its parts. In this situation, the number<br />
of entities involved increases rapidly and the process quickly<br />
becomes convoluted.<br />
This results in a long cycle time between the opening of<br />
an incident and its eventual closing. If the process is too<br />
complex, incidents can even become “forgotten” and never<br />
worked through to completion. Ultimately, the timely<br />
reporting of trends is compromised and, in the worst case, the<br />
entire process breaks down. A typical warning sign of a<br />
breakdown in the process is the phrase, “We just need to<br />
communicate better.” Although this may be true, the process<br />
may simply be too tangled to allow effective communication.<br />
While all parties have the best intentions, the results in such<br />
situations can be disappointing.<br />
4.2 Solution 1: How to overcome the challenge of complex<br />
organization interaction<br />
While working with the stakeholders in the planning<br />
phases of the FRACAS, identify the simplest process or<br />
workflow possible while still keeping key stakeholders<br />
involved at various steps in the process. Whenever possible,<br />
automate the communication. This way, all groups that need<br />
to be aware of an incident’s status can be notified easily.<br />
Assigning a key contact person for each functional group<br />
to attend FRACAS planning meetings and communicate<br />
important information to the group further streamlines<br />
communication.<br />
Regular FRACAS process review meetings involving all<br />
key stakeholders from the various functional groups can be<br />
helpful as well. These meetings can decrease in frequency<br />
over time.<br />
Lastly, make sure that all stakeholders sign off on the<br />
proposed communication plans before the final FRACAS is<br />
deployed to ensure organizational complexity has been<br />
adequately addressed.<br />
4.3 Challenge 2: Lack of prioritized goals<br />
Based on a recent survey conducted by PTC, the top three<br />
reasons for implementing a closed-loop analysis and<br />
corrective action system are:<br />
1. Compliance with customer requirements<br />
2. Gaining insights into the reliability of products<br />
3. Improving the next generation of product design<br />
Depending on the functional groups involved, the relative<br />
importance of these reasons may vary. Consider an example<br />
organization without clearly prioritized goals. A customer<br />
contract specifies that a FRACAS be utilized for a particular<br />
project. The project manager focuses on implementing what<br />
the customer has specifically required, but financial<br />
constraints mean the resources needed to complete a fullfeatured<br />
FRACAS implementation are unavailable.<br />
Project/program managers implement a temporary spreadsheet<br />
or database solution with full intentions of replacing this<br />
solution in the future with a more substantial system, but a<br />
full-featured FRACAS never materializes. Consequently, a<br />
minimal FRACAS becomes the standard resulting in poor<br />
efficiency and a lack of cohesiveness.<br />
Here’s another example: a company believes that a<br />
FRACAS is important to the success of a product, but they see<br />
it as something they will get to in the future. As the product<br />
lifecycle progresses, the company starts to appreciate the<br />
necessity of a FRACAS. By this time, the company is far past<br />
the point of the budgeting and planning processes. Therefore,<br />
a proper implementation does not occur and teams are left<br />
scrambling late in the project to find resources to implement a<br />
FRACAS. Once again, a minimal FRACAS develops.<br />
In a third example, a program manager struggles to<br />
provide a FRACAS implementation while her executive<br />
management has much higher expectations for this system.<br />
They envision that the FRACAS will save the company<br />
money on warranty costs. In addition, the reliability group is<br />
expecting to gather significant data to explore trends and<br />
improve their product’s design over time. Consequently, these<br />
groups expect a full-featured implementation with significant<br />
data gathering, analysis, and reporting capabilities.<br />
Unfortunately, because the program manager has minimal<br />
financial resources, she can only implement a basic FRACAS.<br />
The deployed system is not set up early enough to sufficiently<br />
track information, which limits the data available to executive<br />
management and the reliability group. The system needs more<br />
2012 Annual RELIABILITY and MAINTAINABILITY Symposium Cline & Stillwell – 3
time to gather enough data to perform meaningful analysis.<br />
The resulting implementation falls short of everyone’s<br />
expectations.<br />
What caused the problems in these examples? First, the<br />
groups involved did not discuss and prioritize the list of goals<br />
and expectations. Second, there was not objective FRACAS<br />
leadership across the departments. Finally, effective executive<br />
sponsorship verifying and tracking the goals did not exist.<br />
4.4 Solution 2: How to identify and prioritize goals<br />
To overcome the problems caused by a lack of prioritized<br />
goals, several points must be considered while planning the<br />
FRACAS. First, make sure all stakeholders, including<br />
management, discuss and agree on the goals and expectations<br />
of the FRACAS. Objective FRACAS leadership ensures that<br />
all stakeholders’ goals are considered in the planning. Lastly,<br />
determine which goals can be thoroughly accomplished in the<br />
planned FRACAS, and make sure all stakeholders sign off on<br />
the goal prioritization.<br />
4.5 Challenge 3: Ineffective and inefficient data tracking<br />
As mentioned previously, the key to an effective<br />
FRACAS solution is the gathering and reporting of<br />
meaningful data. This seems to suggest that the more data<br />
gathered on an incident or failure, the better the FRACAS will<br />
be. However, this is not necessarily the case. Too much data<br />
may even prevent users from discovering meaningful trends<br />
because they can’t “see the forest through the trees”.<br />
For example, many legacy FRACAS solutions collect 80<br />
or more fields of information when recording a failure.<br />
Although much of this data is valuable, the ramifications of<br />
gathering that much data can result in several problems.<br />
If it takes customer support personnel recording an<br />
incident five seconds to enter data in each field, they will<br />
spend 400 seconds or 6.7 minutes filling out the form for all<br />
80 fields of data, not including any research time. Soon, many<br />
begin to feel that fully logging a problem is just too timeconsuming.<br />
They start routinely skipping fields and not even<br />
reporting some issues at all.<br />
To prevent this, designers develop input screens that will<br />
easily accept incident information. However, they often use<br />
free-flowing text fields to allow customer support personnel to<br />
type any information that they think is useful. The support<br />
users, though, become confused and do not always know what<br />
to enter. Consequently, they begin to include extraneous or<br />
irrelevant data, sometimes just leaving the field blank. The<br />
resulting “data” becomes unable to support any meaningful<br />
analysis and manually scanning incidents for trends becomes a<br />
tedious chore.<br />
4.6 Solution 3: How to effectively manage data<br />
To avoid the problems of inefficient and ineffective data<br />
tracking, organizations will want to establish functional<br />
procedures before data collection begins. Use of simple,<br />
streamlined forms that store data in a central database is often<br />
the best approach. Taking the time to train the FRACAS users<br />
responsible for data entry helps to ensure that all important<br />
data is entered correctly and efficiently. Reminding those<br />
responsible for the failure data entry of the importance of<br />
capturing the data at the time of failure and the long term<br />
benefits to the organization that result from that timely data<br />
capture is critical. Whenever possible, data capture should be<br />
automated as well.<br />
5. STEPS TO SUCCESSFUL FRACAS IMPLEMENTATION<br />
While this tutorial has discussed three of the most<br />
common challenges during FRACAS implementations, the<br />
issues presented are not all-encompassing. Instead, they<br />
outline typical problems that prevent a company, regardless of<br />
industry, from realizing the dramatic results that can be<br />
achieved with a FRACAS.<br />
Few documented tools and techniques exist to aid the<br />
successful implementation of a closed-loop analysis and<br />
corrective action system. This set of steps for successful<br />
FRACAS implementation is intended to help companies<br />
overcome the obstacles outlined above. The eight step<br />
approach is meant as a framework that can be modified as<br />
needed to fit a specific situation.<br />
5.1 Step 1: Define the goals and success factors<br />
Defining the goals of all intended users and stakeholders<br />
is the foundation for a successful FRACAS implementation.<br />
Every step in the process of establishing a FRACAS will be<br />
based upon this definition. A mistake or misunderstanding at<br />
this step can have negative consequences later. Therefore,<br />
implementations require a commitment to thorough research<br />
and documentation is required as otherwise issues may not<br />
come to light until months later.<br />
To begin this process, hold a series of short team<br />
meetings with each of the groups (as identified in Issue 1) and<br />
the representative stakeholders within the FRACAS process.<br />
Using general facilitation techniques, map out specific goals<br />
or expectations of the FRACAS. Typical goals include<br />
lowering maintenance costs, improving overall reliability, and<br />
improving next generation product design. Be careful of the<br />
common pitfall of moving off of goal establishment and into<br />
detailed requirements. The purpose of this exercise is for each<br />
group to reach a consensus on the priority of its main goals.<br />
Once each group has set its primary goals, call a crossfunctional<br />
meeting. One representative from each group and<br />
an executive and/or management representative should attend<br />
to review, consolidate, and prioritize the goals. During this<br />
same meeting, attach what success realistically means for each<br />
of these goals in concrete terms. For example, if a goal is to<br />
lower maintenance costs, a quantifiable success factor may be<br />
to reduce these costs by ten percent over the next 12 months.<br />
At the conclusion of this meeting, each attendee signs a<br />
document that lists the consolidated and prioritized goals<br />
along with their success factors. This single technique will<br />
immediately highlight the level of agreement between the<br />
groups.<br />
Finally, gain executive approval and support and have one<br />
executive take overall ownership and support of the<br />
implementation. Encourage that person to communicate the<br />
4 – Cline & Stillwell 2012 AR&MS <strong>Tutorial</strong> Notes
goals and success factors back to the teams.<br />
5.2 Step 2: Define the output<br />
With goals and success factors in place, define the results<br />
or output that each group expects from the FRACAS next.<br />
Based on the goals, each group will decide the outputs they<br />
need from the FRACAS to achieve the previously established<br />
success factors. Typically, the results are stated in terms of<br />
calculations, charts, graphs, and reports. For example, the<br />
Reliability group may need a Pareto chart indicating the<br />
number of failures by part number. The Field Service group<br />
may require a report indicating the cost of warranty repairs by<br />
part number. The Quality group may require a reliability<br />
growth chart.<br />
A common pitfall is that each group will want as much<br />
output as possible, resulting in a bloated “wish list” that is<br />
difficult to reasonably implement. The purpose of this<br />
exercise, though, is to focus only on the basic necessities for<br />
success based on the goals determined earlier. To make this<br />
clear, map each output requirement back to a goal and success<br />
factor.<br />
5.3 Step 3: Map the Process/Workflow<br />
Through a series of meetings and interviews with<br />
stakeholders, next discuss what process or workflow they<br />
expect to follow. Most of the groups will focus on their own<br />
internal process but may not understand the overall process to<br />
be followed. Based on each group’s individual input and<br />
through additional investigation, develop a draft of a single<br />
consolidated process diagram. Use this to search out<br />
inefficiencies and actively find ways to simplify the process.<br />
Involve the stakeholders in simplifying and coordinating<br />
the process steps between the groups. Question any step that<br />
does not assist with producing the output requirements<br />
established previously. Also, ensure that the overall process<br />
does not grow too complex. Remember that many steps can<br />
slow the working of incidents or decrease the ability to report<br />
trends in a timely manner. Change to the overall process is<br />
inevitable, but this step establishes a workable starting point.<br />
5.4 Step 4: Map data required and input method<br />
Using the process diagram and the output requirements,<br />
determine the minimum data fields required to support the<br />
workflow process at each step. The investment of time in this<br />
step helps avoid collecting data that has no direct purpose and<br />
focuses efforts in the most important areas.<br />
Once the data fields are established, determine how the<br />
user will view the data. One large form with all data fields or<br />
specific forms with tailored fields to simplify understanding<br />
for each group may be preferable. Involve the stakeholders to<br />
understand what they are expecting and why.<br />
Additionally, specify how the input data will be gathered.<br />
Recall from Issue 3 that efficient collection of valid,<br />
meaningful data is paramount. Popular methods include prepopulating<br />
fields based on previous input, drop-down menu<br />
choices, and even bar code entry.<br />
5.5 Step 5: Implement a prototype FRACAS<br />
Now that the goals, success factors, expected output,<br />
workflow process, and entry forms are defined, choose an<br />
implementation approach. A variety of automation tools that<br />
can be used to support a FRACAS exist, though they fall into<br />
three major levels of capability and sophistication.<br />
The first level of software products that support FRACAS<br />
are off-the-shelf, general purpose tools such as spreadsheet<br />
and personal database programs. Microsoft Excel and<br />
Microsoft Access are two popular options in this category.<br />
They are relatively inexpensive and can be used for a variety<br />
of purposes outside the FRACAS realm. In addition, many<br />
companies already have these tools available. However,<br />
general purpose tools have limited capacity to handle large<br />
amounts of data and may not support data sharing between<br />
multiple users. Also, these tools do not intrinsically support<br />
FRACAS calculations and graphing. Furthermore, this<br />
approach may require internal support to create and maintain<br />
the custom application.<br />
The second level of FRACAS tools includes Workgroup<br />
applications that specifically provide FRACAS support for a<br />
small group of users and moderate amounts of data. FRACAS<br />
Workgroup tools can typically support multiple users,<br />
workflow capabilities, and FRACAS calculations. They also<br />
may connect with other Reliability and Maintainability tools<br />
for more detailed analysis.<br />
The third level of FRACAS tools includes Enterprise<br />
applications that provide intrinsic FRACAS support, including<br />
workflow, calculations, graphs, as well as additional<br />
capabilities for handling large amounts of data and many<br />
users. Enterprise FRACAS tools typically support multiple<br />
means of data entry and reporting, including internet browser<br />
support that allows them to support a large, distributed user<br />
base. Enterprise FRACAS tools also often integrate with other<br />
Enterprise systems, including ERP/PDM, Mail, Workflow,<br />
and Directory Services.<br />
For both Workgroup and Enterprise FRACAS tools, a<br />
continuum of product offerings is available. Customdeveloped<br />
solutions, integrated custom tools, complete<br />
commercial off the shelf (COTS) options all exist on the<br />
market. With today’s technologies and FRACAS product<br />
offerings, custom FRACAS solutions cannot easily be<br />
justified. Custom development is usually the most expensive<br />
option because of the time and resources required to design,<br />
implement, support, and grow a single FRACAS<br />
implementation. Given that FRACAS implementations<br />
typically have common functional and data requirements, a<br />
configurable COTS solution, which allows changes without<br />
programming skills, is a more cost effective approach to<br />
deploying a FRACAS.<br />
To select the appropriate level of a FRACAS tool, users<br />
must examine their immediate needs in relation to future<br />
plans. Factors in the selection criteria must include a firm<br />
understanding of the functionality or capabilities required, the<br />
number of potential users, and the data storage necessities<br />
over time. These decisions may seem daunting, but careful<br />
2012 Annual RELIABILITY and MAINTAINABILITY Symposium Cline & Stillwell – 5
selection of the approach is imperative. To validate the<br />
decision, implement a prototype FRACAS based on the<br />
selected approach.<br />
5.6 Step 6: Accept feedback and modify FRACAS<br />
At this stage, reconvene the stakeholders and start testing<br />
the prototype system. Does the previously defined workflow<br />
process work in this environment? Can the data be efficiently<br />
entered into the system? Can the expected outputs be<br />
generated? Is the system easy to use and understand?<br />
Most importantly, revisit the goals and success factors.<br />
Does the system meet these goals? Will the success factors be<br />
met? Particular areas will likely need additional tweaking or<br />
reworking. This is the time to accept constructive feedback<br />
and make the necessary modifications.<br />
Before proceeding, gain signoff and approval from all<br />
stakeholders. With their continued involvement, the chances<br />
of gaining their support is high, and they can be the best<br />
advocates for the project when proceeding to the next step —<br />
general rollout.<br />
5.7 Step 7: Rollout and train<br />
Companies can take several different approaches for<br />
general rollout, and each method has its own set of advantages<br />
and disadvantages.<br />
The “Big Bang” approach allows all users on the system<br />
at once. This approach is typically used where time is short.<br />
When all the users begin to use the system at the same time,<br />
many problems are likely. With so many users involved,<br />
seemingly minor issues can become major obstacles.<br />
Although this approach can work, it requires significant<br />
planning and coordination. It should only be considered when<br />
absolutely necessary.<br />
Alternatively, most companies prefer a phased approach<br />
in which several groups at a time are trained and brought<br />
online. In this case, unforeseen issues can be addressed<br />
without involving the entire user base. Additionally, the<br />
implementation and training teams can adapt to provide better<br />
support for later groups. Typically, the more individualized<br />
attention ultimately gains more support and acceptance from<br />
the general user base.<br />
Training is extremely important, especially when typical<br />
FRACAS processes involve users of many different areas of<br />
interest and levels of expertise. When users understand what is<br />
expected of them and are empowered by training to use the<br />
tools given to them, the system has a much greater chance for<br />
acceptance and success.<br />
5.8 Step 8: Continue to change<br />
Continual change should be expected for the FRACAS<br />
system based on user feedback regarding what works and<br />
what needs improvement. As business objectives and<br />
processes evolve, the FRACAS needs to adapt and support<br />
these developments. As a result, active management needs to<br />
accept and validate changes to the FRACAS in relation to the<br />
original high-level goals. Through this oversight, a highperformance,<br />
functional system will remain viable for many<br />
years.<br />
6. PRACTICAL APPLICATION<br />
This section presents several case studies to address the<br />
practical application of FRACAS. In some instances the<br />
FRACAS was implemented as a result of a customer<br />
requirement and in others as an internal decision by the<br />
organization. They represent a range of successful FRACAS<br />
implementations based on systematic steps and the FRACAS<br />
best practices described above.<br />
6.1 Case Study #1: Traditional FRACAS<br />
A manufacturer in the aerospace industry instituted and<br />
continues to use a traditional FRACAS. A desire for reliable<br />
products as well as a customer requirement drove the<br />
implementation. Management established various goals,<br />
especially:<br />
• Capture of incident or failure information in a closed loop<br />
system.<br />
• Provide periodic reports to the customer with the closedloop<br />
resolution for each incident.<br />
The required outputs include metrics like MTBF. Reports<br />
are generated as well including Failure Summary, Failure<br />
Analysis, and Reliability reports.<br />
For workflow, the process follows a series of defined<br />
steps: Incident Entry, Quality Review, Failure Investigation,<br />
Root Cause Analysis, Corrective Action, Failure Review<br />
Board, and Closeout.<br />
Data inputs to the FRACAS include data from the<br />
previously implemented database system as a starting point.<br />
Subsequent data is entered manually using customized tables<br />
and forms for each workflow step.<br />
Two programs initially deployed the prototype FRACAS.<br />
During prototype review, necessary additional reports,<br />
calculations and alerts were identified and the FRACAS was<br />
updated to support the added requirements.<br />
The solution provider supplied training to key users for<br />
the FRACAS on both programs.<br />
Since the initial implementation, the FRACAS has<br />
changed slightly to better suit the organization’s needs.<br />
Modifications are ongoing as new needs arise.<br />
This FRACAS has proven to be a success thanks to the<br />
systematic planning approach used during deployment.<br />
6.2 Case Study #2: RMA System<br />
A supplier to the aerospace and defense industry<br />
implemented and continues to use a FRACAS to manage and<br />
track their return material authorization (RMA) process.<br />
The goals and success factors for the FRACAS are as<br />
follows:<br />
• Efficiently track and manage the RMA process.<br />
• Provide procedures and tools to gather reliability data<br />
from various groups, including customer service, repair<br />
and overhaul, original equipment, and other service<br />
centers.<br />
• Attain reliability data to help with quoting expected<br />
maintenance costs.<br />
6 – Cline & Stillwell 2012 AR&MS <strong>Tutorial</strong> Notes
• Identify reliability and maintainability metrics as possible<br />
inputs to other reliability analyses.<br />
Numerous outputs were identified as required, including a<br />
range of reports. Example reports include:<br />
• Incident Report by Serial Number<br />
• Failure Modes by Part Number<br />
• RMA Status Report<br />
The organization defined its workflow steps for the RMA<br />
process as: Field Service Entry, RMA Creation, Return Item<br />
Review, Preliminary Inspection Evaluation, Test Evaluation,<br />
Disassembly Evaluation, Engineering Evaluation, Final<br />
Inspection, and Closeout.<br />
Legacy data from the existing RMA database was<br />
imported to the new FRACAS with plans to input new data<br />
manually going forward.<br />
Based on the goals, available inputs, and desired outputs,<br />
a prototype FRACAS was created and initially deployed to<br />
power users. During the prototype review, users were satisfied<br />
with the FRACAS as implemented.<br />
Initial key users of the FRACAS received training from<br />
the software vendor in a “train-the-trainer” program. These<br />
users then provided training to all other personnel before<br />
rolling the new FRACAS for RMA across the organization.<br />
A systematic step-by-step process allowed for the creation<br />
of a FRACAS process and system that met all requirements<br />
for the RMA.<br />
6.3 Case Study #3: Business Intelligence System<br />
A global information technology corporation instituted<br />
and continues to use FRACAS as a business intelligence<br />
system to support better data mining and ultimately improved<br />
business decision-making. The FRACAS provides one means<br />
of gathering historical, current, and predictive views of<br />
business operations related to testing and fielded product<br />
performance. Prior to reorganization of the testing and fielded<br />
product data capturing and analysis procedure, users entered<br />
data into a customer support system. Then, for critical cases,<br />
users streamlined the data for consistency and readied it for<br />
analysis by design engineers. These data mining processes<br />
were difficult and time-consuming and the company required<br />
a new solution.<br />
The goals and success factors for the FRACAS are as<br />
follows:<br />
• Integrate previously disjoint data sets across the product<br />
lifecycle to create single source of truth for failure<br />
analysis.<br />
• Use actual reliability and availability data to refine targets<br />
from generation to generation and effectively track<br />
reliability growth.<br />
• Track failures during testing to identify potential failure<br />
modes, easily recognize trends, and determine metrics,<br />
including MTBF.<br />
• Perform fielded product performance analysis to manage<br />
outing information and reporting.<br />
The organization identified many outputs, including<br />
numerous graphs and reports. The graphs and reports provided<br />
a view of data trends over multiple years with comparisons to<br />
previous generations of products in an instant dashboard view.<br />
Disparate requirements for testing and tracking field<br />
failures required multiple workflows. For example, for fielded<br />
product failures, the workflow steps include Problem<br />
Definition, Root Cause Analysis, Corrective Action, and<br />
Closeout.<br />
The new tool imported legacy data from the previous<br />
customer support system to streamline the data mining<br />
processes. Going forward, the customer support team will<br />
input their data directly to the FRACAS via the carefully<br />
designed forms.<br />
Based on the goals, available inputs, and desired outputs,<br />
a prototype FRACAS was created and initially deployed to<br />
power users. After gathering user feedback, necessary changes<br />
were implemented, including additional graphs and reports to<br />
let both engineers and management quickly acquire up-to-date<br />
information from the FRACAS.<br />
The software vendor team who implemented the<br />
FRACAS trained power users. These power users, in turn,<br />
trained the remaining user base on the role-specific features<br />
and functions. Following training, all customer support<br />
personnel, design engineers, and management began using the<br />
new tool.<br />
A systematic step-by-step process facilitated the creation<br />
of a FRACAS system to improve the efficiency of data mining<br />
and other analysis, which provides critical information driving<br />
advantageous business decisions for the firm.<br />
6.4 Case Study #4: Non-Traditional FRACAS<br />
A manufacturer of highly specialized, leading edge<br />
industrial components implemented and continues to use a<br />
FRACAS to track details related to raw materials inspection.<br />
The primary goal and success factor for the FRACAS is<br />
to provide the process and system to track data pertaining to<br />
incoming product, work-in-progress, and final inspection. The<br />
required system allows inspectors to enter the results of their<br />
material reviews and track progress of lots. The firm also<br />
anticipated that the FRACAS would help to determine metrics<br />
such as the first pass yield and percent scrapped.<br />
The implementation team identified many outputs<br />
including various reports. Example reports include:<br />
• First Pass Yield Report<br />
• Scrap Report<br />
• Action Report<br />
The workflow for the process was identified and includes<br />
three key inspection steps: Incoming, Work-in-Progress, and<br />
Final.<br />
As no legacy data was easily available for the new<br />
FRACAS for raw materials inspection, all data is entered<br />
manually to the FRACAS using custom forms and tables.<br />
With the goals, inputs, and outputs in place, the software<br />
vendor built the prototype FRACAS for initial deployment to<br />
power users. After gathering feedback, the vendor<br />
implemented certain changes, including the addition of alerts<br />
to automate notifications based on other entered data and<br />
better control for dates associated with acceptance or rejection<br />
of the materials inspected.<br />
2012 Annual RELIABILITY and MAINTAINABILITY Symposium Cline & Stillwell – 7
The vendor also trained the initial key users of the<br />
FRACAS, who then provided subsequent training to all<br />
personnel. The new FRACAS for raw materials inspection<br />
was then rolled out corporate wide.<br />
A systematic step-by-step process allowed for the creation<br />
of a FRACAS process and system to meet requirements for<br />
the raw materials inspection system.<br />
6.5 Case Study #5: CAPA<br />
A manufacturer of highly specialized, leading edge<br />
industrial components implemented and continues to use a<br />
FRACAS to accommodate their Corrective Action Request<br />
(CAR)/Preventative Action Request (PAR) process.<br />
Key goals and success factors for the FRACAS are as<br />
follows:<br />
• Efficiently track and manage the CAR/PAR process.<br />
• Track metrics like time to close each CAR/PAR.<br />
• Incorporate alerts to notify of overdue actions.<br />
Various outputs were identified, including reports and<br />
graphs. Example reports and graphs include:<br />
• Average Time to Close CAR/PAR<br />
• Issues by Source<br />
The FRACAS team defined the workflow steps for the<br />
CAR/PAR process as: CAR/PAR Entry, Review, Root Cause,<br />
Action Taken, Effectiveness, and Closeout.<br />
The new tool imported legacy data from the existing<br />
CAR/PAR system, and new data will be input manually going<br />
forward.<br />
Based on the goals, available inputs, and desired outputs,<br />
a prototype FRACAS was created and initially deployed to a<br />
small group of users familiar with the process. Updates to<br />
further streamline the forms for more efficient data entry,<br />
among other changes, were implemented based on user<br />
feedback.<br />
The vendor trained users of the CAR/PAR FRACAS<br />
before the company rolled out the new tool to replace the<br />
previous system.<br />
Again, a systematic step-by-step process was critical for<br />
the creation of a FRACAS to meet requirements for the<br />
CAR/PAR process.<br />
7. CONCLUSIONS<br />
Closed loop corrective action systems like FRACAS<br />
provide an effective process for controlling, tracking, and<br />
analyzing failures. They can lead to improvements in quality,<br />
reliability, and productivity while simultaneously reducing<br />
costs. When working to implement FRACAS, be aware of the<br />
challenges that may arise. Keep best practices in mind and,<br />
like the organizations described in the case studies, follow the<br />
recommended eight step procedure for successful FRACAS<br />
implementations.<br />
8. REFERENCES<br />
1. Failure Reporting, Analysis and Corrective Action<br />
System (FRACAS) Application Guidelines, Product Code<br />
FRACAS, Reliability Analysis Center, 1999 Sep., p 5.<br />
2. M. Villacourt, P. Govil, “Failure Reporting, Analysis and<br />
Corrective Action System (FRACAS)”, Doc ID #:<br />
94042332A-GEN, SEMATECH, 1994 Jun., Available at<br />
http://www.sematech.org/docubase/document/2332agen.p<br />
df<br />
3. NASA, Preferred Reliability Practices “Problem<br />
Reporting and Corrective Action System,” Practice No.<br />
PD-ED-1255, available at<br />
http://klabs.org/DEI/References/design_guidelines/design<br />
_series/1255ksc.pdf<br />
4. MIL-HDBK-2155 Department of Defense Handbook:<br />
Failure Reporting, Analysis and Corrective Actions Taken<br />
5. RAC: Reliability Problem Solving, Failure Reporting and<br />
Corrective Action System (FRACAS) and Reverse<br />
Engineering, http://src.alionscience.com/pdf/fracas.pdf<br />
6. E.J. Hallquist, T. Schick, “Best Practices for a FRACAS<br />
Implementation”, pp 663-667, RAMS 2004.<br />
7. J.S. Magnus, “Standardized FRACAS for nonstandardized<br />
products”, pp 447-451, RAMS 1989.<br />
8. A. Mukherjee, “Integrated FRACAS systems for F117<br />
infrared acquisition designation system (IRADS) support<br />
yield higher MTBMA”, pp 26 - 29, RAMS 2005.<br />
9. MIL-STD-721C: Definitions of Terms for Reliability and<br />
Maintainability, 1981.<br />
10. Directive-Type Memorandum (DTM) 11-003 –<br />
Reliability Analysis, Planning, Tracking, and Reporting,<br />
March 2011, US DoD.<br />
8 – Cline & Stillwell 2012 AR&MS <strong>Tutorial</strong> Notes