Beruflich Dokumente
Kultur Dokumente
Game Theory
For Effective
Resource
Management
p. 30
QUALITY PROGRESS
Just My Type
LEAN AND SIX SIGMA
Organizing employees
by process vs. function
grows efficiency p. 16
Plus:
Bolstering your chances
VOLUME 50/NUMBER 2
PRIMERS
Our Primers are the most widely used texts for certification training.
They can be taken into the exam. QCI offers 16 different Primers.
SOLUTION TEXTS
Detailed solutions to all questions in the corresponding Primer.
CD-ROMS
Interactive software to assist students preparing for ASQ exams.
FEATURES
16 SERVICE QUALITY
Come Together
The pod concept and lean principles streamlined the customer
experience for one insurance organization.
24 CASE STUDY
Subject to Review
A Virginia hospital pinpointed the factors that improved the odds of
success for its Six Sigma projects.
by Todd Creasy
16
30 PROJECT MANAGEMENT
Good Shepherds
The tragedy of the commons theory on resource sharing will give
organizations added insight into managing staff and projects more
efficiently and effectively.
36 CASE STUDY
Small Changes, Big Results
• Added Insight
To detect defects in electronic products, design a highly accelerated
stress screen and understand what adjustments to make to build a Additional figures and tables
better product. to illustrate the sources of
variation in statistical analyses,
by Daniel K. Sarosky the topic of this month’s Back
to Basics (“Uncovering the Truth,”
p. 64).
30
• Volviendo a los
Fundamentos
Back to Basics translated
into Spanish.
• Questions Answered
Check out QP’s archive of the
Expert Answers department.
• Daily Fix
Read Quality News Today for
the latest quality-related headlines.
DEPARTMENTS
10
6
8
LogOn
• The natural force of PDCA.
Expert Answers
• Sample size confusion.
Keeping Current
QP
Mail
Quality Progress/ASQ
600 N. Plankinton Ave.
QUALITY PROGRESS
10
name followed by @asq.org (for example,
vellifson@asq.org).
60 QP Reviews
Article Submissions
Quality Progress is a peer-reviewed publica-
tion with 85% of its feature articles written
by quality professionals. For information
about submitting an article, call Valerie
COLUMNS
Ellifson at 800-248-1946 x2139, or email
manuscripts@asq.org.
Author Guidelines
5 Up Front 44 Statistics Roundtable To learn more about the manuscript review
Keep it simple. Using sensitivity analyses to design process, helpful hints before submitting a
better studies. manuscript and QP’s 2017 editorial planner,
click on “Author Guidelines” at www.
48
qualityprogress.com under “Tools and
48 Standards Outlook Resources.“
Getting a handle on changes to the
automotive standard ISO TS 16949. Photocopying Authorization
Authorization to photocopy items for
internal or personal use or the internal or
personal use of specific clients is granted by
63 One Good Idea Quality Progress provided the fee of $1 per
Measuring effective-quality copy is paid to ASQ or the Copyright Clear-
to improve processes. ance Center, 222 Rosewood Drive, Danvers,
MA 01923, 978-750-8400. Copying for other
purposes requires the express permission
64 BONUS of Quality Progress. For permission, write
Back to Basics Quality Progress, PO Box 3005, Milwaukee,
Addressing variation in statistical WI 53201-3005, call 414-272-8575 x7406, fax
analyses. 414-272-1734 or email reprints@asq.org.
Photocopies, Reprints
And Microform
Article photocopies are available from ASQ
at 800-248-1946. To purchase bulk reprints
(more than 100), contact Barbara Mitrovic at
ASQ, 800-248-1946. For microform, contact
SPECIAL SECTION ProQuest Information and Learning, 300 N.
Zeeb Road, Ann Arbor, MI 48106, 800-521-
ASQ’S 2017 SIX SIGMA 0600 x2888, international 734-761-4700,
RESOURCE GUIDE p. 54 www.il.proquest.com.
4 QP • www.qualityprogress.com
UPFRONT
QP QUALITY PROGRESS
DECLUTTER. DESTRESS. REMOVE complication. It’s something to strive for, but ASSOCIATE EDITOR
Mark Edmund
often easier said than done. Many of us are overscheduled, overburdened and burning the
candle at both ends most of the time. And because it’s so commonplace nowadays, there ASSISTANT EDITOR
Tyler Gaskill
is a whole movement that has sprung up to help people rid their lives of physical things,
activities and even people that create “static.” To commit to being less “busy.” MANUSCRIPT COORDINATOR
Valerie Ellifson
Essentially, that is the goal of lean and Six Sigma initiatives as they apply to manufac-
COPY EDITOR
turing, service, healthcare or other industries. This month, we highlight some of those Susan E. Daniels
methods and tools to help simplify processes and streamline productivity, removing the
ART DIRECTOR
noise from the everyday. Mary Uttech
In “Come Together,” p. 16, learn about how the pod concept and lean principles help GRAPHIC DESIGNER
organize employees by process rather than function, allowing organizations (service, Sandy Wyss
in the example provided) to focus more wholly on the customer. The configuration can PRODUCTION
provide more visibility to processes end-to-end, and provide staff with direct customer Cathy Milquet
feedback. The article explains how to implement pods and their potential pitfalls, then ADVERTISING PRODUCTION
Barbara Mitrovic
provides a brief case study explaining how pods worked for one service organization.
Healthcare settings are popular venues for implementing Six Sigma for improvement DIGITAL PRODUCTION SPECIALISTS
Julie Stroik
projects. “Subject to Review,” p. 24, provides a case study of a 200-bed Virginia hospital Julie Wagner
that conducted several Six Sigma projects over a four-year period. Often, projects are MEDIA SALES
considered either successes or failures, and everyone moves on. This hospital instead Naylor LLC
Lou Brandow
focused on rigorous analysis to gauge the relative successes or failures of its projects, Krys D’Antonio
helping it to determine what was needed to justify future projects. Nicholas Manis
Norbert Musial
“Good Shepherds,” p. 30, explains the tragedy of the commons theory and how it Erin Pande
can be applied to resource management in scheduling staff and projects. Again, if done EDITORIAL OFFICES
effectively, workers receive clearer direction in assignments, allowing them to com- Phone: 414-272-8575
Fax: 414-272-1734
plete tasks more effectively without becoming overworked. More is not always better.
ADVERTISING OFFICES
In this month’s Standards Outlook, you’ll find part two of the series on the automotive Phone: 866-277-5666
standard, IATF 16949:2016. Author R. Dan Reid outlines what quality professionals must
ASQ ADMINISTRATION
know about the changes, and how they should be engaging their organizations’ leadership. CEO
William J. Troy
If you’re going to lose a good team member, it sure lessens the sting if it’s to a good Senior Leadership
Andrew Baines
cause. This month, we say goodbye to Assistant Editor Tyler Gaskill, who leaves us for Kalleen Bruch
a position in the healthcare industry where he’ll get to build and manage his own quality Dick Palmersheim
Shontra Powell
program. I’m sure all he’s learned during his time with ASQ and QP will give him a great
To promote discussion of issues in the field of quality and
jump-start. Best of luck, Tyler! QP ensure coverage of all responsible points of view, Quality
Progress publishes articles representing conflicting and minor-
ity views. Opinions expressed are those of the authors and not
necessarily of ASQ or Quality Progress. Use of the ASQ logo in
advertisements does not necessarily constitute endorsement of
that particular product or service by ASQ.
Seiche Sanders
Editor
February 2017 • QP 5
LOGON
Seen&Heard Tune In
Quality commitment ment (QM) must only report noncon-
“Building Assurance” (January 2017, pp. formities and should not intervene
37-43) was a good example of apply- in shipment decisions. Ideally, this
ing quality management principles to a might work. But departments make
construction project up front and ahead mistakes on their decisions. In some
of schedule to deliver improved quality circumstances, QM may need more
outputs and cost savings on a project. strength and even have to report to
It’s vital for higher levels of the organization to
organizations to stop a bad decision. It all depends on ASQ TV’s latest episode offers
do it right the the situation. some highlights from QP’s 2016
first time with I’m speaking from my experience salary survey report, including the
pride in their in the food industry. I know it sounds average salary for
commitment and silly, but after complaints come in, U.S. quality profes-
workmanship. upper management many times sionals and how
Martin Searson blames QM for not doing anything. holding certifica-
Dublin, Ireland Of course, it also is silly to rely too tions can influence
much on inspection and rejection. QM salaries.
should be more proactive in helping Visit http://videos.asq.org
Everyday quality (and if possible ensuring) departments to access the full video library.
The law of gravity example in “Back to design and execute good processes—
Basics: The Gravity of PDCA” (January beginning with raw materials. That is
2017, p. 64) is so appropriate. Everyone a more preventive approach.
applies the plan-do-check-act cycle in Francis Watson StayConnected
their daily lives, regardless of whether Ica, Peru Find the latest news, quips and targeted
they know its name, scholastic theory content from QP staff.
or steps. Playing games
Director of Knowledge Products
After reading this article, it’s possible In response to “Gaming the System” Seiche Sanders:
@ASQ_Seiche
to easily convince people to use quality (December 2016, pp. 44-48):
management’s many tools and approach- From a quality standpoint, the
es in their everyday tasks without label- great game of business could be Associate Editor Mark Edmund:
@ASQ_Mark
ing it appropriately. played on cost of quality alone. For
Catalina Tutulea example, your employees might
editor@asq.org
Targoviste, Romania receive a profit sharing bonus. The
quality costs could have a bottom-
Resisting ‘Resistance’ line impact on their bonuses that www.facebook.com/
groups/43461176682
With regard to “Expert Answers: Com- they can see and feel empowered to
bating Quality Resistance” (January affect.
2017, p. 8), I cannot agree wholly with Edward Oshaben www.linkedin.com/groups/quality-
progress-magazine-asq-1878386
the author’s claim that quality manage- Euclid, OH
6 QP • www.qualityprogress.com
QP ONLINE
EXTRAS@
www.qualityprogress.com
QP QUALITY PROGRESS
PAST CHAIR
Looking for Career Advice or a New Job? Patricia La Londe (retired – CareFusion)
Subscribe to ASQ’s career-focused monthly enewsletter, Career Connection, to receive
articles, videos and other resources to help grow your professional quality career CHAIR
and improve your workplace. Career Connection also delivers the latest quality job Eric Hayler, BMW Manufacturing
postings and helps your organization find top talent. Subscribe for free at http://asq.org/
newsletters.
CHAIR-ELECT
Elmer Corbin, IBM
What fundamental quality tool or concept must all quality professionals master? DIRECTORS
Donald Brecken, Ferris State University
Jim Creiman, Northrop Grumman Corp.
6.1% Root cause analysis. Ha C. Dao, Emerson Climate Technologies Inc.
71.5% Value stream mapping. David B. Levy, Boyce Technologies Inc.
10.7%
Control chart. Austin S. Lin, Google
11.5% Fishbone diagram. Raul Molteni, Molteni Consulting
Luis G. Morales, Daimler Trucks North America
Mark Moyer, CAMLS
Daniella A. Picciotti, Veridiam
Visit www.qualityprogress.com for the latest question: Barrie Simpson, Genentech Access Solutions
Joann Sternke, Pewaukee School District
What’s the most critical factor in overcoming resistance to Six Sigma programs? Sunil Thawani, Quality Indeed Consulting FZE
• Involving front-line employees. John Vandenbemden, Q-Met-Tech
Allen Wong, Abbott
• Accurately communicating the program’s objectives.
• Top management support. QP EDITORIAL REVIEW BOARD
• Demonstrating success early. Randy Brull, chair
Administrative Committee
Brady Boggs, Randy Brull, Jane Campanizzi,
Larry Haugh, Jim Jaquess, Gary MacLean,
QualityNewsTODAY
R. Dan Reid, Richard Stump
Technical Reviewers
Naveen Agarwal, Ashraf Ali, Suresh Anaganti,
Recent headlines from ASQ’s global news service M. Onur Artan, Andy Barnett, Matthew Barsalou,
David Bonyuet, David Burger, Brooks Carder, Bernie
Medical Device Makers Urged to Fight Hacking Threats
Carpenter, L.N. Prabhu Chandrasekaran, Ken Co-
Preventing cyberattacks on medical devices requires more than great designs. In today’s gan, Linda Cubalchini-Travis, Ahmad Elshennawy,
connected world, device makers and hospitals must continually look for problems in Mark Gavoor, Kunita Gear, Daniel Gold,
existing devices, and communicate early and clearly when potential risks emerge, ac- T. Gourishankar, Roberto Guzman, Ellen Hardy,
Lynne Hare, Victoria Jones, Trevor Jordan, Ray
cording to federal regulators. (http://tinyurl.com/qnt-hacking)
Klotz, Tom Kubiak, William LaFollette, Scott Laman,
Patients Get Bar Codes to Avoid Hospital Blunders Pradip Mehta, N.S. Narahari, Arind Parthasarathy,
Larry Picciano, Gene Placzkowski, Tony Polito,
Bar codes are being used in routine hospital operations for the first time to help reduce Peter Pylipow, Imran Ahmad Rana, John Rich-
human error. The technology is being tested by regulators in England to “revolutionize ards, James Rooney, Ayman Sakr, Mahboubeh
patient safety.” Bar codes enable parts used in a major surgery to be easily traced if faults Samghabadi, Brian Scullin, Abhijit Sengupta, Ami-
are found years later. (http://tinyurl.com/qnt-bar-codes) tava Sengupta, Mohit Sharma, A.V. Srinivas, Adrian
Tan, Joe Tunner, Balan Vaithiyanathan, Manu Vora,
Keith Wagoner, Jack Westfall, Doron Zilbershtein
WANT THE LATEST QUALITY-RELATED NEWS AND ANALYSIS? The QNT Weekly
enewsletter, available exclusively to ASQ members, delivers it every Friday.
Subscribe now at http://asq.org/newsletters.
February 2017 • QP 7
EXPERTANSWE
Sample size confusion
Q: Our customer expects specific levels
sional samples must not change based on
the result of the visual inspection.
to the entire batch, and the best way to
ensure samples are random is to use a
of inspection, such as pulling 157 bottles Sometimes, visual and dimensional random-number generator and perform
for visual testing, but they also want 20 characteristics are correlated. For collections based on the production
pulled for dimensional testing. Can the example, process upsets in injection sequence.
20 additional bottles be a subset of the molding can simultaneously cause visual This may be difficult to do in practice
original testing sample? (cosmetic) defects and dimensional con- because the times of collection will be
When calculating the lot, should I pull cerns. This is why care should be taken to unpredictable. It is much easier to collect
the samples before or after my calcula- ensure that an inspector does not remove stratified samples such as one-third of the
tions? Are the samples included in the cosmetic defects before selecting the total samples required at each of three
produced quantity? For example, if a subset for the dimensional samples. intervals—the beginning, middle and end
customer orders 10,000 bottles and the As a general rule, I advise my clients to of a production run. Studies show that
level-two inspection pulls 200 bottles, this collect a single sample. In many regulated stratified sampling is more appropriate if
drops the total shipped to the customer to industries, one critical defect is suffi- the process quality drifts over time.1
9,800. If 10,200 bottles are produced, the cient to reject a batch. Suppose you are With regard to whether samples
inspection level increases to 315 bottles conducting the visual inspection, and you should be included in the production to-
that must be pulled for testing. What is discover a critical defect. You must reject tal: In most cases, visual and dimensional
the correct sample size and production the batch even though the purpose of the checks are nondestructive. So if there
number? inspection was to check for cosmetic de- is no contractual requirement to retain
fects. If you collect separate samples, you the samples, there is no reason why you
A: You asked some great, relevant are actually increasing the total number can’t ship the acceptable samples to the
questions. For your first question about of units that must be inspected. This will customer. This means you should include
whether you can combine visual and increase the probability of rejecting the the samples in the production total.
dimensional checks in a single sample, the batch. Some facilities conduct large produc-
simple answer is yes. But there is a catch. Some people recommend using ran- tion runs and sample once to approve the
The sampling scheme should be defined dom sampling to select samples. Random entire lot. Later, as orders come in, they
prior to collection. The subset of dimen- samples must be collected with respect break the lot into smaller sublots for ship-
ping. In this case, I recommend counting
the entire production run when determin-
ing the sample size.
Based on the sample sizes you men-
tioned in your question, it sounds like you
are using American National Standards
Institute (ANSI) and ASQ quality control
standards Z1.4 and Z1.9.2 These include
index tables that increase the sample size
as a function of the production total. If
you comply with these standards, there
will be cases in which they direct you
to use nonzero acceptance criteria. This
is inappropriate for regulated industries
such as pharmaceuticals—in which
a single critical defect is sufficient to
reject the batch. For this reason, all of
the double sampling plans and multiple
8 QP • www.qualityprogress.com
RS
sampling plans in the ANSI standard are
unacceptable for critical defects. This is
corresponding to the AQL.
3. Go down the column until you reach
sample—so there is no need to index the
sample size in proportion to the lot size.
because all of these plans have nonzero the A = 0, R = 1 plan. If your organization is not bound by
acceptance standards. 4. Look to the left to obtain the sample regulatory or contractual constraints, you
I advise my clients to use only “ac- size. can consider using nonzero acceptance
cept = zero” (A = 0) sampling plans. This This approach to attribute sampling standards. Table 2 shows that the per-
greatly simplifies the process of selecting does not use the ANSI sampling plan formance of the sampling plan improves
a sampling plan. The process can be sum- “code letters.” Because there is only one somewhat as the sample size increases.
marized as follows: sampling plan with A = 0 for any given Like Table 1, this table shows the prob-
1. Determine the acceptance quality limit AQL, the sample size is independent of ability of passing the sampling plan if the
(AQL) per the quality manual or policy the lot size. Keep in mind that the confi- population quality is equal to the AQL.
directive. dence associated with a sampling plan Table 2 shows that the sampling plan’s
2. Enter the ANSI table in the column derives from the size of the sample, not performance improves somewhat as
the size of the lot. It, therefore, doesn’t the sample size increases. For example,
32 A = 0, R = 1 100,000 0.8796 plan that is appropriate if there is a possi- Some organizations use the ANSI
bility of detecting a critical defect during attribute-sampling tables for attribute
32 A = 0, R = 1 10,000 0.8795
the inspection. and dimensional checks. While they are
32 A = 0, R = 1 5,000 0.8793
The ANSI attribute-sampling tables are free to use any statistically valid sampling
32 A = 0, R = 1 2,500 0.8789
based on binomial probability, but the plan, the attribute sampling plans are
32 A = 0, R = 1 1,000 0.8778
binomial approximation is not accurate inefficient for inspection by variables,
32 A = 0, R = 1 500 0.8760 if a lot size is small or the sample size including dimensional checks. These
AQL = acceptance quality limit is large relative to the lot size. Table 1, organizations may be able to reduce the
therefore, shows the exact hypergeomet- sample size by at least 50% by switching
increases / TABLE 2 ability of passing the sampling plan if the NSF Health Sciences Pharma Biotech
population quality is equal to the AQL. Houston
AQL = 0.4 Acceptance Probability For this table, the sampling plan is based
sample criteria of on AQL = 0.4. It is important to note that REFERENCE AND NOTE
size acceptance 1. Paula Lagares Barreiro and Justo Puerto Albandoz,
similar performance would be obtained “Population and Sample: Sampling Techniques,” Manage-
n = 32 A = 0, R = 1 0.8796 ment Mathematics for European Schools, http://tinyurl.
for other AQL values. com/sampling-teq-qp.
n = 125 A = 1, R = 2 0.9101 2. For more information on sampling, read the American
Table 1 shows the probability of ac-
National Standards Institute (ANSI) and ASQ standard,
n = 200 A = 2, R = 3 0.9529 ceptance is nearly constant regardless of ANSI/ASQ Z1.4 Sampling Procedures and Tables for
n = 315 A = 3, R = 4 0.9611 the lot size. As stated earlier, the power
Inspection by Attributes (2008) and ANSI/ASQ Z1.9–2003
(R2013): Sampling Procedures and Tables for Inspection
n = 500 A = 5, R = 6 0.9837 of sampling is due to the size of the by Variables for Percent Nonconforming (2003).
February 2017 • QP 9
KEEPINGCURRE
TECHNOLOGY
T
Augmented reality could reinvent manufacturing processes, employee training
The technology behind last summer’s Pokémon Go craze is get- mechanics wire airplanes’ complex electrical systems. Instead of
ting serious. Augmented reality (AR)—overlaying computer-gen- trying to follow paper or laptop instructions, mechanics’ Google
erated graphics onto reality—is transitioning from a kids’ game Glass headsets show them wire placement diagrams and instruc-
to military and industrial applications, such as helping Lockheed tions. So far, the pilot program has reduced production time 25%
Martin engineers assemble F-35 aircraft or allowing factory-floor and errors have been reduced to zero.5
workers to see 3-D blueprints AR’s benefits also extend
overlaid as holograms on ma- to data analysis. Holograms
chines they’re repairing. are useful for visualizing data
As this new technology because they engage the spatial
becomes more affordable and awareness part of the brain that
mainstream, it has the potential allows humans to understand
to create factory-floor efficiencies complex concepts more quickly
and employee training programs and promotes greater reten-
that were previously considered tion, according to Brian Mullins,
the stuff of science fiction. co-founder and CEO of Daqri,
a maker of augmented reality
From fiction to reality smart helmets.6
According to Forrester Research “In cases where you have
Inc., by 2025 nearly 15 million U.S. to react with very complex
workers will use smart glasses, such as Google Glass and Micro- amounts of information, it can tell a much better story than just a
soft’s HoloLens, which is an increase of 400,000 users from 2016. bar graph,” Mullins said.7
Forrester also estimates large organizations will spend $3.6 billion
on smart glasses in 2025, up from $6 million in 2016.1 IoT meets AR
AR’s ability to provide visual representations of data or sche- Incorporating the Internet of Things (IoT) with AR could add yet
matics on the factory floor could be a game changer. A mainte- another level of possibilities to how manufacturers could use the
nance worker who is examining a pressure valve, for example, technology. Many factories are using IoT machines and devices to
could be wearing lenses that display the valve’s temperature and create a connected environment that sends large amounts of data,
pressure as text that hovers above the part, said Mike Campbell, such as power consumption, operating status or dates the ma-
executive vice president of industrial software firm PTC’s Vuforia chines or devices were last serviced. Campbell said using IoT alone,
augmented reality segment.2 however, lacks context and that AR can contextualize IoT data.8
This concept just scratches the surface of AR’s potential “Getting that data is not as hard these days as making sense
applications, Campbell said. The lenses could overlay graphics out of it for the people that need it to make decisions,” said Peggy
that also show how a part’s pieces fit together, instructions for Gulick, director of business-process improvement at AGCO Corp.9
disassembling it and how it could connect to different machines.3 For example, an employee wearing AR gear could view any
Several organizations already are using AR in this way. IoT-enabled machine and receive visual cues about whether
AGCO Corp., a Minnesota agricultural equipment manufacturer, it’s experiencing any problems and where an issue is occurring
has equipped some of its employees with Google Glass, which without even opening the machine for repairs. Experts believe
shows diagrams and instructions to help them perform quality this would not only save organizations time and labor, but service
checks on tractors. This method has made their quality checks technicians could use AR to determine whether a faulty part is
20% faster.4 salvageable or should be replaced.10
Boeing is trying a pilot program in which Google Glass helps “One of the big things that IoT is changing is that people will
10 QP • www.qualityprogress.com
E NT
Q
not need to interact only with other people, but also with machine
Who’s Who in
and devices,” said Luigi De Bernardini, CEO of manufacturing software
provider Autoware. “The more intelligent machines are, the more users
will need to interact with them, and the more they will expect to do it
in a natural way.”11 NAME: Lucita Kahn.
February 2017 • QP 11
KEEPINGCURRENT
MANUFACTURING
SHORTRUNS SURVEY: COMPANIES STAYING
THE BALDRIGE PERFORMANCE Excellence Program re- WITH THEIR SUPPLIERS
While 83% of manufacturers say they have been adversely
leased the 2017-2018 Baldrige Excellence Framework (Busi-
affected in the past by their suppliers’ inabilities to meet
ness/Nonprofit, Healthcare and Education) booklets. All three
their needs, only one-third anticipate a shortage of parts or
versions include the Baldrige criteria, core values and concepts,
services this year, according to a recent ASQ survey.
and guidelines for evaluating an organization’s processes and
ASQ’s 2017 Manufacturing Outlook Survey also shows
results. Baldrige Excellence Framework printed booklets costs that 66% of manufacturers expecting a problem with sup-
$30. For more information, visit http://asq.org/2017baldrige. pliers are working closely with providers to resolve issues,
while 35% are working with their suppliers’ competition.
THE AUTOMOTIVE INDUSTRY Action Group’s southern au- Some manufacturers are stockpiling parts, while others are
tomotive quality summit will be held March 9 in Huntsville, AL. expanding operations to create the necessary parts them-
The one-day event will be the official roll-out of the new Inter- selves.
national Automotive Task Force 16949 standard to the supply “Supply chain plays a critical role in manufacturing, and
base in the southeast United States. For updates on this event, companies simply can’t risk being without the material they
need to be successful,” said Pat La Londe, past chair of ASQ’s
visit http://bit.ly/SouthernAutoQuality2017 (case sensitive).
board of directors. “Companies need to carefully consider
multiple options when faced with a shortage of materials or
A NEW RISK AND regulatory services innovation center has
with suppliers that can’t meet their needs.”
been established by PricewaterhouseCoopers and Carnegie
More than 1,125 manufacturing professionals from around
Mellon University to provide training to organizations and the world responded to ASQ’s 2017 Manufacturing Out-
perform research on risk and compliance-related issues. The look Survey, which was conducted online in November and
center also will address issues related to audit innovation, data December. Respondents to the survey represented several
analytics, information privacy and security. For more informa- industries such as aerospace, automotive, food and medical
tion, visit http://tinyurl.com/new-innovate-center. devices.
For more information from the survey, visit http://tinyurl.
com/asq-outlook-survey.
Ensuring long-lasting battery life for AR measurable savings ranging from 25 to 9. Castellanos, “Augmented Reality, Hologram-Like Images
Enter the Workplace,” see reference 1.
devices is another hurdle tech companies more than 90%,” he said. “[This is because] 10. K.R. Sanjiv, “How Augmented Reality Can Revolutionize
are struggling to clear.17 when you augment people with AR, you’re Manufacturing,” IndustryWeek, Sept. 29, 2016, http://
tinyurl.com/iw-qp-reality.
Still, many experts agree that it’s a mat- not buying equipment: You’re freeing up
11. Luigi De Bernardini, “Augmented Reality Is Key for Smart
ter of when—not if—AR will significantly cash flow because you’re speeding up Manufacturing,” AutomationWorld, June 20, 2016, http://
tinyurl.com/qp-luigi.
change industries and people’s lives. processes.”19 —compiled by Tyler Gaskill,
12. Jeff Kerns, “Will Virtual Reality Replace the Desktop?”
“I believe that every walk of life and assistant editor Machine Design, Jan. 10, 2017, http://tinyurl.com/qp-kerns.
every human activity, which today is medi- 13. Ibid.
REFERENCES 14. Sanjiv, “How Augmented Reality Can Revolutionize Manu-
ated by computers, will be transformed 1. Sara Castellanos, “Augmented Reality, Hologram-Like facturing,” see reference 10.
by AR,” said Microsoft CEO Satya Nadella. Images Enter the Workplace,” Wall Street Journal, Dec. 12, 15. Ted Ranosa, “Doctors Can Now Practice Surgical Pro-
2016, http://tinyurl.com/wsj-qp-augmented. cedures Through Augmented Reality,” Tech Times, Jan. 7,
“This is not happening tomorrow. It’s not 2. Matt Weinberger, “Augmented Reality Is Going From 2017, http://tinyurl.com/ar-surgery.
happening next year. But that is the con- ‘Pokémon Go’ to the Factory Floor,” Business Insider, Oct. 16. Ibid.
25, 2016, http://tinyurl.com/bi-qp-augmented. 17. Tom Mainelli, “The Challenge and Opportunity of Aug-
tinuum we are on.”18 3. Ibid. mented Reality,” Recode, May 8, 2016, http://tinyurl.com/
Dan Arczynksi of Index AR Solutions, 4. Castellanos, “Augmented Reality, Hologram-Like Images qp-recode.
Enter the Workplace,” see reference 1. 18. Gerard Baker, “Microsoft CEO Envisions a Whole New
a developer of AR applications, said AR 5. Ibid. Reality,” Wall Street Journal, Oct. 30, 2016, http://tinyurl.
isn’t a technology that will just provide 6. Ibid. com/wsj-qp-new-reality.
7. Ibid. 19. David Greenfield, “Augmented Reality for Industry,”
incremental, 2 to 5% percent savings. 8. Weinberger, “Augmented Reality Is Going From ‘Pokémon AutomationWorld, June 8, 2016, http://tinyurl.com/qp-
“It’s a game-changing application with Go’ to the Factory Floor,” see reference 2. automation-kc.
12 QP • www.qualityprogress.com
ASQ
ASQ FINALIZES 2017 BOARD
DATE IN QUALITY HISTORY ASQ has named its 2017 board of directors. They are:
QP looks back on a person or event that made a difference in the • Eric Hayler, chair, BMW Manufacturing.
history of quality. • Patricia La Londe, past chair, CareFusion (retired).
ASQ CONFERENCE
February 2017 • QP 13
KEEPINGCURRENT
HEALTHCARE
14 QP • www.qualityprogress.com
STANDARDS
ISO AUDITING STANDARD UNDERGOING REVISIONS
The International Organization for Standardization’s (ISO) popular systems, there has been an increase in the number of sector-
standard for auditing management systems is under revision specific standards to respond to the mandate,” said Denise
and has just reached the first voting stage. Robitaille, chair of ISO project committee 302, the ISO project
ISO 19011—Guidelines for auditing management systems is committee responsible for the revision. “There are now MSSs
currently being revised to reflect the growing number of man- that cover areas such as health and medical, environment,
agement system standards (MSS) and the recent revisions of services, information technology and more. In addition, the two
some of the most widely used ones, such as ISO 9001 for quality most popular MSSs—ISO 9001 and ISO 14001—have recently
and ISO 14001 for the environment. It has just reached commit- been updated, so the auditing of these systems needs to reflect
tee draft stage, meaning those countries involved in its revision the variety and number of standards being developed.”
have an opportunity to make comments on the draft. For more information and updates, visit the ISO website at
“As organizations see the benefit and need for management http://tinyurl.com/iso-audit-stand.
July October
Engagement and Standards and Auditing
Culture
November
August Customer Experience
A Stronger Supply
Chain December
Salary Survey
February 2017 • QP 15
Come
Together Using pods to
streamline the
customer experience
In 50 Words
Or Less
• The pod concept
organizes employees
in a single location by
process rather than
function.
• By relying on lean
principles, pods help
employees see a pro-
cess end-to-end, receive
direct feedback from
the customer and gain
autonomy.
• This can increase
employee engagement,
eliminate handoffs and
improve customer satis-
faction.
February 2017 • QP 17
The pod concept aims to organize employees by ployee engagement. For example, a case manager at
process instead of function, allowing for efficient and an insurance organization we worked with said, “Pods
effective coordination across process steps. The pod definitely helped me build better relationships with
design is based on Henry Ford’s assembly line, which other functions in the organization because I now have
minimized wait time by moving the product along a more face-to-face contact. Better relationships and
line to be worked on by technicians at each step. This knowing your co-workers make for a more productive
same benefit can be achieved in a service organization and rewarding work day.”
by having team members—who represent each func-
tion and understand the work of other functions—or- Implementing pods
ganized into one pod. In essence, a customer is the There is a 10-step approach for implementing pods
product that’s moving down the assembly line with (see Figure 1), and you should start by considering its
minimal wait times. scope.
In addition to increasing customer focus, pods 1. Scope the project—Selecting the right process
also increase flexibility and scalability. As demand to tackle is imperative to success. You should consider
increases, decreases or changes, the amount and fo- selection criteria such as:
cus of pods can be easily adjusted. They also increase • Clear benefits to the customer.
employee engagement and empowerment because the • Process stability.
pods “own” the customer. • Probability of success.
Lean principles are at the core of the pod concept. • Cost to implement.
Lean seeks to understand value from the viewpoint of • Team availability.
the customer and eliminate activities that don’t add Starting small can reduce risks and increase the
value—waste. Pods apply lean principles such as: likelihood of success. Start with the customer perspec-
• Creating physical co-location, which minimizes tive and select an end-to-end process with clear out-
communication delays. puts.
• Eliminating waste, such as handoffs. 2. Manage the change from the start—Weigh
• Minimizing customer wait time. the tradeoffs of implementing a radical change versus
Pods can help achieve drastic improvements in per- one that’s gradual. Consider the amount of resistance
formance. Table 1 shows how adopting a pod structure you will encounter from those who are affected. Cre-
improved the customer intake process for a healthcare ate a communication plan. Messaging should be clear,
services business. Table 1’s data show substantial and concise, fact-based and sent through several channels.
immediate improvements. In our experience, pods Train frontline managers on how to best handle ques-
have the potential to reduce cycle time by 50% and tions and concerns from employees.
costs by 20%. 3. Develop a project plan—Create a plan with
Adopting a pod structure also can increase em- high-level activities needed and the appropriate tim-
ings for completion. Start with the end state in mind
18 QP • www.qualityprogress.com
SERVICE QUALITY
7. Redesign jobs
tems, equipment and templates are in use and update expertise of the HR department as you redesign jobs to
them as necessary. ensure there is correct salary grading, and don’t forget
5. Develop metrics and establish a baseline— how managers will be affected. Because pods require
Determine how you will measure success. Ask whether less coordination, the number of managers and their
you have a reliable baseline from which to improve. If roles often change dramatically.
you don’t, put together a data collection plan to create 8. Address training needs—As you map the new
one. When selecting measurements of success, ensure process, develop new tools and redesign jobs, think
the process-level metrics align with your organization- about the competencies required at each process step
wide measurements of success. Do not move forward and develop a training plan that addresses what con-
with a pilot until you have the right metrics and mea- tent is required, who is to be trained and how the par-
surement system in place. Senior leadership will want ticipants will be trained—such as virtually or face to
to see results from the initiative, and you must be pre- face. Pod members must have a detailed understand-
pared to show them. ing of how the process works from end to end.
6. Assess staffing needs—A pod will be cross- 9. Prepare the workspace—Prepare for how a
functional, so you will be drawing on resources from pod structure will physically change your office. Con-
several areas. Create a staffing model to allow for fact- sider how many members each pod will have and what
based decisions about how many employees it will the layout will look like. Determine pods’ locations,
take to complete the work. Develop different staffing and mobilize resources to support the move and setup.
scenarios using the staffing model, and let the key de- Don’t assume everything will be ready on day No. 1.
cision makers come to an agreement on what makes Give pod members time to get settled and ensure their
sense. phones and computers work. The last thing you want
7. Redesign jobs—As you decide on the staffing is a pod in place without the necessary tools to service
needs, consider how roles could be consolidated or customers.
expanded. What are the competencies (knowledge, 10. Pilot—The purpose of the pilot is to test the
skills and abilities) needed for each role? Will the process and design assumptions prior to launch Have
current staff be able to perform the redesigned jobs? the project team available to help answer any ques-
What responsibilities may need to shift? Leverage the tions. Use a whiteboard to visualize the major steps in
February 2017 • QP 19
New customer-centric process / FIGURE 2
Centralized services
Sales coordinator
• External
Sold case
the pod process, and collect data on performance and mistake is to jump right into designing the future state.
the status of customer deliverables to identify break- Spend time mapping and understanding the current
downs and opportunities for improvement. Refine the state and pain points that must be addressed in the fu-
design before scaling up. ture state.
Measuring everything because you can—There
What to avoid are many things you can measure, such as turnaround
When adopting the pod concept, there are many com- time, rework rates, production rates, queue size, staff-
mon pitfalls that must be avoided: ing levels or service levels. Measuring everything can
Underestimating the amount of organizational lead to confusion, however, so decide early on how
change—Any process redesign inevitably creates a to best track progress and invest in building a robust
need to change. Understand the amount of change re- baseline.
quired and its implications. Determine what effect co- Expecting immediate results—After the first
locating employees will have on reporting lines. week, you will likely find turnaround times worsen
Delaying communication—Waiting to communi- and there is some failure to complete customer orders.
cate about pods until all pieces of the project are in Be prepared to deal with underperformance and set
place might sound like a good idea, but that can create realistic expectations with regard to a timeline for im-
anxiety and fear. proved performance. Remember, any process change
The more knowledge and understanding people will inevitably decrease productivity in the short run.
have, the more rapidly your organization can adopt Relying too much on the staffing model—The
the pod process. Use various opportunities to explore tendency is to take the initial model at face value, but
the reservations, concerns and complaints about the remember it’s a model and only as good as the accu-
change. Set up a display pod for prospective team racy of the data. The real test is to set up a pilot pod
members to see and experience the changes ahead. and start testing assumptions.
Overplanning—Project plans are important. Underestimating manager’s resistance—Man-
Spend time and effort creating a plan while keeping in agers are often protective of their turf. There is going
mind that it is only a plan, so remain flexible enough to to be inevitable resistance to giving up resources for
adjust course as needed. the pods. Anticipate this and manage it effectively.
Skipping current-state analysis—A common Underinvesting in team building—Teamwork is
20 QP • www.qualityprogress.com
SERVICE QUALITY
the right work, and driving service Implementation specialist—Southeast 0% 40% 60%
levels and turnaround times to in- Implementation specialist—West 0% 40% 60%
dustry standards. Implementation specialist—Great Lakes 0% 40% 60%
Pod kickoff—Leadership creat- Implementation specialist—Other 0% 40% 60%
ed an initiative framework divided SWAT = specialized work action team
into three tracks that would reduce
customer touch points, shared services and functional customer and organized the customer delivery by pro-
initiatives. The team focused on reducing customer cess rather than function.
touch points. To do this, it supported and scoped an To understand the intricacies of the process and
operational process redesign that created value for the create a foundation for how it would continue in the
February 2017 • QP 21
SERVICE QUALITY
22 QP • www.qualityprogress.com
CHECK OUT THE NEW
BOOKS AVAILABLE FROM
ASQ QUALITY PRESS!
THIS MONTH’S PICK! The Certified Six Sigma Black Belt
Handbook, Third Edition
The best Six Sigma Black Belt handbook has
The Certified Supplier been fully revised, updated, and expanded!
Additions include more exercises, new tables
Quality Professional and figures, new content in the DMAIC parts
Handbook of the book, and more.
The importance of a safe, reliable, Item: H1494
cost-effective, high-quality, and
integral supply chain cannot be
overstated. ASQ’s new Certified We Move Our Own Cheese!:
Supplier Quality Professional A Business Fable About
(CSQP) certification provides
Championing Change
valuable credentials to quality
professionals in the growing field This book is about creating change. The
of supplier quality engineering. cheese is a metaphor for what we have in life
and what we believe we want more of. In a
Item: H1520 business context, it represents the business we
are in and what it gives us.
Item: H1523
HEALTHCARE DELIVERY
has changed dramatically over the years: from simple
house calls to health maintenance organizations, point
of service plans, preferred-provider organizations and Affordable
Care Act exchanges. Not surprisingly, the common tread in the
most recent delivery systems is cost containment.
Controlling costs, in fact, has become the priority for health-
care executives. “The challenge for industry executives is to
continue to control spending even in the face of countervailing
winds such as expensive new innovations, improved consumer
confidence, and an aging society that requires more medical care
and services,” according to the 2015 PricewaterhouseCoopers
(PWC) Health Research Institute report.1
February 2017 • QP 25
Authors of the report also suggest the need for which were then reduced to the top eight rea-
“systemness.” When referring to the many services sons.7
offered by a typical hospital, this new buzzword em- • Multiple studies report that up to 70% of projects
phasizes the need for a well-functioning whole rather fail. For total quality management projects, the
than well-functioning disparate parts. Care teams failure rate ranges from 60 to 80%.8
should strive to achieve more by working together, Looking to reduce waste, standardize practices,
eliminating redundancies and reinforcing common improve quality of care, and advance clinical or staff
goals through clinical and administrative standard- productivity, many hospital CEOs have turned to
ization, which should lead to lower operating costs, process improvement or re-engineering to achieve
PWC suggests. Healthcare organizations and hos- operational goals. Many of these efforts have paid
pitals are no strangers to the need to standardize, dividends, but others have not.
economize and seek to do more with less.2-4 For sev- Considering the results of the PWC report and
eral years, healthcare projects have been initiated to its call for systemness, this article seeks to identify
reduce costs, increase quality of care and improve two dynamics while showing why some healthcare
patient, clinical staff and employee satisfaction. projects fail and others succeed. It is hoped that les-
Despite these attempts to improve the effective- sons learned can help reduce failure percentages and
ness and efficiency of healthcare delivery systems, boosting healthcare project success rates.
however, many projects fail to deliver on the oppor-
tunity. Several reports support how project opportu- Case and method
nities often fall short of expectations: A study was conducted at a 200-bed rural Virginia
• A McKinsey Quarterly report in 1994 examined hospital that had engaged in Six Sigma projects
100 organizations that had engaged in significant for about four years. The hospital’s investment in-
projects and how the organizations’ executives cluded lean Six Sigma (LSS) White Belt training for
believed too much had been invested with too few all senior managerial staff, Green Belt (GB) training
bottom-line results.5 for about 25 personnel and Black Belt training for
• A 1994 report cited in ComputerWorld said nearly one senior manager.
85% of top executives were dissatisfied with the During the four-year span, Six Sigma projects
results of process re-engineering.6 were classified as either:
• A Minitab survey of 200 Six Sigma professionals • An unqualified success—a project such as im-
looked to pinpoint the main reasons projects fail. proving a med-cart exchange between nursing
The results were narrowed into 41 failure themes, and pharmacy, the time to get obstetrics supplies,
26 QP • www.qualityprogress.com
CASE STUDY
project would be focused also were named. To ensure 1. Size of the project (too large, not manageable).
valid responses by meeting attendees, each participant 2. Low senior management support and prioritization.
was given an anonymous survey at the beginning of 3. Little to no involvement of process owner.
each project’s dialogue. 4. Project leader (Green Belt) skills and interest lacking.
Table 1 shows the survey instrument for successful 5. Overly complex process.
and moderately successful projects. Table 2 shows a 6. Lack of clearly defined project goals.
similar survey instrument for projects considered dis-
7. Little or no collaboration between departments or within department.
appointments or failures.
8. Continually changing project goals.
Using these surveys before each project’s discus-
9. Low team cohesion.
sion, the meeting attendees were asked to identify the
10. Poor measuring system.
top three reasons that would explain the improvement
projects’ success, moderate success or failure. After 11. Little or no control mechanism to keep process performing well.
February 2017 • QP 27
rectly influenced by the project or departmental lead- support of previous research, the balance between
er—and process owner involvement all exhibit some project success and failure depends heavily on the
form of leadership. A 2002 survey of 126 organizations department leader and the project manager (in this
showed top management sponsorship and commit- case, a GB). As such, it is imperative the process
ment were the most important contributors to any suc- owner—most likely a clinician—and the Six Sigma
cessful change initiative. 9
project leader be held responsible for the improve-
Within a healthcare environment, process owners ment initiative.
are usually clinicians: for example, nurse managers, Interestingly, for moderately successful projects—
lab supervisors, physical therapy department leaders, those that could have attained more but didn’t—notice
registered dietitians, doctors and surgeons, and ER that no forms of leadership were selected to justify
directors. Without the involvement and support of that ranking. This section is classified as “project dy-
these employees, success at the clinical level will be namics,” which could mean that even with lesser forms
difficult. of leadership, if a project team has grounded itself se-
In terms of leadership’s importance related to curely with several fundamental tenets, there is a good
project failure, similar results were found. Project change of some project success:
management, process owner involvement and col- • Properly defined goals was one of those funda-
laboration among departments—which is directly in- mentals and seems to support Smith’s research that
fluenced by that department’s manager—each relate noted “the most important contributors to re-engi-
directly to leadership. The absence of leadership can neering success were a clear mission statement and
have a profound negative effect on projects, accord- understanding of the business re-engineering defini-
ing to recent research: tion.”13
• When examining the top eight reasons projects fail, • A good measuring system was the second rea-
the Minitab survey of 200 Six Sigma professionals son given for a moderately successful project. This
noted the top failure mode, not surprisingly, was the factor is supported by research that noted an ab-
lack of management support.10 sence of data was the No. 4 reason for a project to
• “Little to no support from the leadership … imple- fail.14
menting the initiative” is one of the three leading • The size of the project was ranked high to ensure
reasons for quality initiative failures, wrote authors moderate project success. Selecting a manageable-
Jamison Kovach and Jerry Mairani.11 sized project was supported by Louis Johnson, who
• The lack of executive leadership, support and par- noted large project size as being the No. 5 reason
ticipation is a big influence when projects fail, wrote projects fail,15 and Smith, who wrote that projects
author Martin Smith.12 “kept small and manageable” had a greater chance
Based on this sample’s collective opinion and in of success.16
28 QP • www.qualityprogress.com
CASE STUDY
February 2017 • QP 29
Good
Shepherds
by Kathryn D. McIver
PROJECT MANAGEMENT
February 2017 • QP 31
For this article, the only resource being discussed pasture for grazing their sheep. For the good of the
is employees who are working normal workweeks. community, the farmers have a set number of sheep
Resource management—such as the availability of they are allowed to graze in this pasture. This set
team members—is a crucial factor a lean practitioner number of sheep is the output of a formula the town
or project manager must consider when reviewing a elders used to determine exactly how many sheep
current-state process or planning an upcoming proj- the town can put in that pasture and still allow the
ect. The core principle of resource management is to pasture to remain usable year after year.
maximize outputs while minimizing inputs, which are A farmer decides he wants to add one more sheep
complementary goals to reducing waste in process im- because he thinks “just one” won’t hurt and, after all, he
provement efforts. needs the extra wool for his oldest son’s college fund.
Similar to an improvement project in which each Of course, the problem with this mindset is that every
process element contributes to outcomes, an organi- farmer gets the same idea, and they all add one more
zation’s resources contribute to organizational perfor- sheep. After every farmer adds a sheep, the pasture is
mance and must be assessed as assets and resources quickly overgrazed and cannot support sheep, which
that can be exploited. The element that’s missing from harms the village’s income.
this view of resource management, however, is the ap- This example shows the overexploitation of a re-
plication of lean manufacturing’s pull principles. Pull source—the pasture. The rationale behind the TOTC
is a cornerstone of process improvement and stream- is simple: If a resource is commonly owned, it will be
lined operations in which a need is filled in a just-in- overused. A more tangible and relevant example of the
time manner, rather than at the pace of an upstream TOTC is internet usage.
factor. If you think of employees as resources, pull When internet providers’ pricing structures switched
means they receive tasks when they’re available to from consumption-driven or time-based pricing to a
complete them, rather than having the work pushed flat rate, internet use spiked and caused more users to
or assigned independent of their current workload or download at all times, which slowed everyone’s down-
capacity. load speeds.
The TOTC is most frequently used to model or ex-
plain the effects of consumption and resources on a Applying TOTC in organizations
global scale. It’s considered highly applicable to the Applying this analogy to an organization and project-
study of sustainability and environmental resource level work requires viewing the organizational re-
management. source as a pasture, departments as the farmers, an
organization’s product as the wool and projects as
TOTC and counting sheep sheep that draw on the organization’s resources. These
The name “tragedy of the commons” goes back to resources can include budgets, equipment, human cap-
medieval times when communities had common land ital or any other resource that’s used in the course of
for the villagers to raise their livestock. Sheep farm- producing a deliverable.
ers, therefore, provide a classic example of TOTC: TOTC depends on the number of consumers or proj-
Imagine a village in which all farmers use a common ects (sheep), as well as available resources (pasture)
and production requirements (mak-
ing the wool). Dependencies on
Organizational project roadmap the number of projects linked to a
example / TABLE 1 resource can be seen at a high level
in portfolio management but may
not be clear at the project level. The
Resource Project Timeline Level of effort
number of projects that draw from
Bob Standard work March 2014 to July 2014 10 hours/weekly a specific resource is equivalent to
Bob Streamlining April 2014 to July 2014 10 hours/weekly the number of sheep in that pasture.
Tim Standard work March 2014 to July 2014 5 hours/weekly The overuse of a specific resource
Anne Standard work March 2014 to July 2014 5 hours/weekly is the scenario that must be consid-
32 QP • www.qualityprogress.com
PROJECT MANAGEMENT
February 2017 • QP 33
PROJECT MANAGEMENT
34 QP • www.qualityprogress.com
2017 LEAN AND SIX SIGMA CONFERENCE
February 27 – 28, 2017 | Phoenix, AZ
Join hundreds of other global quality and process improvement professionals like
you for two exciting days as we focus on:
February 2017 • QP 37
The design and development of HASS profiles are pri- ent air temperature of the environmental chamber up
marily based on the results of a highly accelerated life in 10º C increments. Dwell times at each temperature
test (HALT). There are several necessary steps that lead level can vary depending on the product, but a mini-
up to and include HASS tuning: mum of 10 minutes is recommended following ther-
mal stabilization (see Figure 1).
1. HALT 3. Rapid thermal transitions—The minimum thermal
To understand the various aspects of HASS tuning, it is cycle range is within 5º C of the lower and upper ther-
pertinent to review the basic construction of a typical mal operating limits discovered during the cold and
HASS profile, starting with a HALT.3 hot-step stress tests. A minimum of three cycles is
A HALT is a limit test, not a pass-fail test. The main recommended with minimum dwell times of 10 min-
objective of a HALT is to determine the product’s envi- utes at each thermal limit (see Figure 2).
ronmental (temperature and vibration) limits.4 Also, it’s 4. Vibration-step stress—Start at 5 root mean square
used to measure a product’s robustness. acceleration [g(RMS)] (lower vibration level—that
There are five tests to perform during a HALT that are is, 1 g(RMS) for delicate products and increment in 5
required to develop a HASS profile. The temperature and g(RMS) steps. The dwell time should be a minimum of
vibration levels and dwell times are suggested starting five minutes and can vary depending on the product’s
points and can be tailored to the product. The five tests mass. The ambient air in the chamber should be at
are: room temperature. Continue stepping the vibration to
1. Cold-step stress—Begin at 20º C and step the am- the point at which the product becomes out of specifi-
bient air temperature of the environmental chamber cation or reaches its destruction limit (see Figure 3).
down in 10º C decrements. Dwell times at each tem- 5. Combined environment—A minimum of five cycles
perature level can vary depending on the product, but is required unless a failure occurs prior to completing
a minimum of 10 minutes is recommended following the five cycles. The thermal cycle should be within 5º
thermal stabilization (see Figure 1). C of upper and lower thermal limits. The starting vi-
2. Hot-step stress—Begin at 30º C and step the ambi- bration level is determined by dividing the maximum
160
140
120
100
80
Temperature
60
40
Cold-step stress Hot-step stress
20
0
8:00:00
8:10:00
8:20:00
8:30:00
8:40:00
8:50:00
9:00:00
9:10:00
9:20:00
9:30:00
9:40:00
9:50:00
10:00:00
10:10:00
10:20:00
10:30:00
10:40:00
10:50:00
11:00:00
11:10:00
11:20:00
11:30:00
11:40:00
11:50:00
12:00:00
12:10:00
12:20:00
12:30:00
12:40:00
12:50:00
13:00:00
13:10:00
13:20:00
13:30:00
-20
-40
-60
-80
-100
Time
38 QP • www.qualityprogress.com
CASE STUDY
Temperature
2. HASS 40
12:25:00
12:27:30
12:30:00
12:32:30
12:35:00
12:37:30
12:40:00
12:42:30
12:45:00
12:47:30
12:50:00
12:52:30
12:55:00
12:57:30
13:00:00
13:02:30
13:05:00
13:07:30
12:10:00
13:12:30
13:15:00
and maximum temperatures would be 80% of the mini- -20
mum and maximum operating temperatures deter- -40
mined during the cold and hot-step stress tests.5 These -60
temperatures are the levels at which the unit under test -80
was fully functional during the HALT. Time
The maximum vibration level is typically 50% of the
upper destruct limit determined during a vibration step
Vibration-step stress test / FIGURE 3
stress test. Figure 5 (p. 41) illustrates a general HASS
60 60
profile shape. The square wave is the temperature, and Temperature
Vibration
the trapezoidal wave is the vibration. 50 50
The HASS profile in Figure 5 shows two HASS cy-
cles. For screening products, there is no set number of 40 40
Temperature
Vibration
cycles or industry standard to use.6 It’s dependent on
the type of product being manufactured and the vol- 30 30
February 2017 • QP 39
Combined environment stress test / FIGURE 4
140 60
Temperature
Vibration
120
50
100
80
40
60
Temperature
Vibration
40
30
20
0
15:10:10
15:20:10
15:30:10
15:40:10
15:50:10
16:00:10
16:10:10
16:20:10
16:30:10
16:40:10
16:50:10
17:00:10
17:10:10
17:20:10
17:30:10
17:40:10
17:50:10
18:00:10
18:10:10
18:20:10
18:30:10
18:40:10
18:50:10
19:00:10
19:10:10
19:20:10
19:30:10
19:40:10
19:50:10
20:00:10
-20
-20
-40
10
-60
-80 0
Time
profile for a predetermined number of cycles to ensure in stress levels are strategically applied to the base HASS
the profile is not too stressful. profile to increase detection and precipitation probabil-
A good starting point for the number of cycles is 40. If ity for a given failure mode.9 The stress level changes de-
time allows, a product can be exposed to as many cycles pend on the type of defects escaping into the field.
as it takes before the first failure occurs. Determining
the amount of useful life that is removed from a prod- Tuning example
uct after being exposed to the HASS profile is a simple, An electrical assembly with an inadequate or insufficient
straightforward calculation and expressed in a percent- solder joint on a resistor fails after only several hours of
age as:8 product operation out in the field (see Figure 7, p. 42). The
Useful life = (# of cycles in HASS profile / # cycles is product originally passed a HASS and all other produc-
SOS) * 100. tion testing without issues before being shipped to the
For example: customer. Examination of the failed board clearly shows
Useful life = (2 / 40) *100 = 5%. a manufacturing defect (missing solder), which is exactly
If the production HASS profile has two cycles and the type of anomaly that HASS is designed to capture.
there were 40 cycles performed during SOS, the amount Addressing this workmanship defect, a stress level
of useful life removed for a product therefore is no more test is performed on the HASS profile to initially consider
than 5%. modifying its vibration. When changing stress levels, it is
recommended that modifying the amplitude should be
5. HASS tuning considered before increasing the dwell time.10
Despite implementing a HALT/HASS program and fol- If just changing the amplitude doesn’t detect the de-
lowing all the standard guidelines, organizations still will fect, consider adjusting the dwell time. Changes in the
experience initial or early failures that escape into the vibration level should be made incrementally—for exam-
field, albeit a notable reduction in defects versus a prod- ple, 5 g(RMS) steps. After modifying the vibration level,
uct program lacking HASS. the defect (the poor solder joint) must be seeded in a
In HASS tuning and timing, slight changes or tweaks product and screened using the modified HASS profile.
40 QP • www.qualityprogress.com
CASE STUDY
80
60
Temperature/Vibration
40
20
0
17:47:51
17:49:31
17:51:11
17:52:51
17:54:31
17:56:11
17:57:51
17:59:31
18:01:11
18:02:51
18:04:31
18:06:11
18:07:51
18:09:31
18:11:11
18:12:51
18:14:31
18:16:11
18:17:51
18:19:31
18:21:11
18:22:51
18:24:31
18:26:11
18:27:51
18:29:31
18:31:11
18:32:51
18:34:31
18:36:11
18:37:51
18:39:31
18:41:11
18:42:51
18:44:31
18:46:11
18:47:51
-20
-40
-60
1 cycle 1 cycle
-80
Time
If the profile detects the defect, the profile was success- at an accelerated ramp rate (for example, 60º C/minutes)
fully tuned. If not, the stress level must be changed and during the HASS profile operation, it would create a low-
the process repeated (see Figure 7). cycle fatigue condition in the highly stressed area.
Increasing the vibration level may not be the solution When modifying the thermal limits, it’s important not
for detecting poor solder joints. This can be attributed to to exceed the upper and lower temperature operating
an insufficient mass of the leads and secondary attach- limits identified during HALT because the product po-
ment methods (for example, spring-force lead forming). tentially can be stressed to destructive-limit levels. Modi-
Low-level vibration excitation—commonly referred to as fying the temperature to near or at the upper operating
tickle vibration—has been proven to detect defects char- limit identified during HALT, for example, may weaken
acterized with intermittent connections (for example, electrical components and cause them to fail or wear out
mating direct-current power connectors).11 prematurely. Again, after the thermal changes are made,
Low-level random vibration is defined between 1 and seeded failure testing would need to be performed.
5 g(RMS). The unit g(RMS) is the square root of the area If changing only the vibration stress level or only the
under the acceleration spectral density curve in the fre- temperature stress level does not improve the profile’s
quency domain. The tickle vibration excitations should detectability, the temperature and vibration stresses may
be implemented in the detection stage of the HASS pro- need to be changed for the profile to detect the defects.
file, but also can be implemented in the precipitation After the tuned HASS profile has demonstrated it can
stage.12 The theory behind tickle vibration is that low vi- repeatedly detect the defect, SOS must be revisited—as
bration levels will ferret out certain defects, which would originally for the HASS profile construction—to ensure
otherwise go undetected at higher vibration levels. the improved profile does not remove useful life from the
If tuning the vibration stress levels does not improve product. SOS is the process of exposing the product to
defect detection, modifying the thermal stresses may ap- repeated HASS cycles.
ply. In the poor solder joint example, another possible
approach to improving detection would be to increase Improving reliability
the upper and lower-temperature limits of the HASS pro- HASS tuning is engineering craftsmanship. There are
file. This increases the temperature delta which, if used no specific industry guidelines documented on how to
February 2017 • QP 41
CASE STUDY
HASS tuning
25.0
X axis - grid lines
Y axis - grid lines Safety of screen
0.0
tune HASS profiles. Success in HASS tuning will come performed to ensure the changes to do not damage the
from experience and strategic test trials. It is the cor- product or remove useful life (more than 3 to 5%) from
rect incremental iteration to address ineffective HASS the product.
profiles. The tuning of HASS profiles is based on the ar- Improving the reliability of electronic products does
chitecture and complexity of the electronic products. not end at the design phase. Environmental screening
Slight changes to the stress levels can increase the de- plays an important role in keeping the product’s quality
tectability of a profile. and reliability at desirable levels. After being correctly
The product field performance is an excellent gauge implemented, HALT/HASS programs will increase
and a telltale sign that HASS profiles are either effec- product reliability. QP
tive or need improvement. An acceptable failure per-
centage may vary based on the organization’s product REFERENCES
1. Gregg K. Hobbs, Accelerated Reliability Engineering: HALT and HASS, John
volume. A desired annualized failure rate for complex Wiley, 2000.
2. Harry W. McLean, HALT, HASS & HASA Explained: Accelerated Reliability
products (greater than 1,000 electrical components) is Techniques, ASQ Quality Press, 2000.
less than 2%. 3. David Rahe, HALT Standard Document Revision 5.0, Professional Testing,
1999.
HASS tuning does have its shortcomings, however, 4. Wikipedia, “Random Vibration,” http://en.wikipedia.org/wiki/random_
vibration.
such as introducing unwanted damage. If the profile’s 5. Ibid.
stress is increased too much, the added stress will 6. McLean, HALT, HASS & HASA Explained: Accelerated Reliability Techniques,
see reference 2.
cause damage or even latent defects that will cause the 7. Ibid.
product to fail prematurely in the field, leading to cus- 8. Rahe, HALT Standard Document Revision 5.0, see reference 3.
9. Hobbs, Accelerated Reliability Engineering: HALT and HASS, see reference 1.
tomer dissatisfaction. 10. McLean, HALT, HASS & HASA Explained: Accelerated Reliability Tech-
niques, see reference 2.
All changes will need to be validated via proof of 11. Hobbs, Accelerated Reliability Engineering: HALT and HASS, see reference 1.
screen, or seeded failure testing, and SOS. If the prod- 12. Ibid.
42 QP • www.qualityprogress.com
8
Online, self-paced
training courses.
44 QP • www.qualityprogress.com
within the data sets are a good study for
heterogeneity. If equations one to three
one suggested thresh-
old). The FEM assumes
4 linear regression
measured a different variable for a given that all the studies have equations with
x1, this demonstrates heterogeneity in the a “true” effect size that Anscombe’s data / TABLE 3
relationships of the variables. Within each is identical and the only
Equation 1: Y1 = 3.00 + 0.500 X1
equation, there are examples of heteroge- variation in a study’s
Predictor Coef SE coef T P
neity. Equations one and two are examples results is sampling error.
Constant 3.000 1.125 2.67 0.026
of heterogeneity in the residuals to the The REM assumes each
X1 0.5001 0.1179 4.24 0.002
regression, known as heteroscedasticity; study provides informa-
S = 1.23660 R-sq = 66.7%
the pattern in the heteroscedasticity of tion about different effect
Equation 2: Y2 = 3.00 + 0.500 X1
equation two also would indicate the poor sizes and controls for this
Predictor Coef SE coef T P
fit for linear regression. between-study variability Constant 3.001 1.125 2.67 0.026
For equations three and four, the outlier in its summary values. X1 0.5000 0.1180 4.24 0.002
points may be interesting or mundane. It When reporting results S = 1.23721 R-sq = 66.6%
could simply (and often) be an entry mistake as an REM, researchers Equation 3: Y3 = 3.15 + 0.486 X1
or calculation error. If it is a trustworthy data can assume they have Predictor Coef SE coef T P
point, however, it is worthwhile to investi- controlled for heteroge- Constant 3.152 1.136 2.77 0.022
gate. These points may hint at underlying neity between studies X1 0.4861 0.1191 4.08 0.003
subpopulations that have a distinctly differ- but have not explained S = 1.24948 R-sq = 64.9%
Equation 4: Y4 = 3.07 + 0.496 X2
ent response (heterogeneity in the sample or why the studies vary. It is
Predictor Coef SE coef T P
independent variables), or the measured out- even possible that when
Constant 3.071 1.082 2.84 0.019
come is distinctly different at that measure- switching from a FEM to
X2 0.4963 0.1134 4.38 0.002
ment (heterogeneity in the function, output a REM, the results will
S = 1.18948 R-sq = 68%
or dependent variable), or both. switch conclusions from
Coef = coefficient SE = standard error
While it is often frustrating to find these favoring one treatment to P = p-value T = test statistic
issues in an analysis, heterogeneity can favoring another.
lead to new questions and conclusions In a meta-analysis comparing the For this meta-analysis, several covariates
that create better models and discoveries. persistence of asthma over and under age were available: year of the study, location of
For multivariate linear models, diagnostic 12, seven studies were identified and a the study, percentage of females in the study,
statistics such as Cook’s D, leverage and risk difference (RD) (> age 12 - <= age 12) lost to follow-up study and year of asthma
residual analyses can point to issues with was synthesized over these studies. The onset. Because of the small number of stud-
the relationship between independent and results (see Figure 2) show that with the ies, each was examined separately. Initially,
dependent variables.2 FEM, the overall effect was significant (p < a jackknife procedure omitting individual
0.001) and was negative (significantly lower studies was performed, and the omission
Controlling heterogeneity: persistence over age 12). For the REM, the of one study (the only study from Iceland)
Using REMs in meta-analyses overall effect was inconclusive and not reduced the heterogeneity to 51%.
The majority of meta-analyses are conduct- significant (p = 0.513). Highly significant After further examination, this study
ed using either fixed effects models (FEM) heterogeneity was seen with the variation (with the largest positive effect size) had an
or random effects models (REM). The deci- in the RD attributable to heterogeneity = imbalance of females and had many more
sion to switch from FEM to REM is often 76% (Cochran’s Q statistic p < 0.0001). The patients lost to follow-up than the other six
based solely on the values of Cochran’s REM was chosen as the appropriate model studies. The final decision of the authors
Q-statistic or the I statistic.
2
and presented by one of the authors. was to include the REM as well as an
Cochran’s Q statistic represents the total Selecting the appropriate model for re- extensive sensitivity analysis to identify the
variance between the studies, while I2 is a porting, however, should not be the final step potential causes of heterogeneity.
measure of how much heterogeneity there in the process. Identifying possible causes of
is between the studies.3 Researchers will heterogeneity and analyzing the results—in- Improving future studies
use the REM if Cochran’s Q is significant or cluding important covariates in the model— These examples illustrate the problem
if the I2 statistic is large (greater than 50% is should be part of the analysis plan. that simply examining results of statistical
February 2017 • QP 45
analyses—even those as simple as sum-
Plots of Anscombe data / FIGURE 1
mary statistics—can hide heterogeneity
in the data set. Several rules of thumb for Equation 1 Equation 2
Scatter plot of Y1 vs. X1 Scatter plot of Y2 vs. X1
statistical analyses can be suggested by 10
11
these examples: 10 9
8 7
data set is large, take a random sample
Y1
Y2
7 6
and examine the distribution. 6 5
Y4
8 8
without outliers, and attempt to identify
7 7
why the data point is an outlier. Include 6 6
Ballardini (2012) 0.03 (0.02, 0.05) 26.18 Ballardini (2012) 0.03 (0.02, 0.05) 15.36
Burr (2013) -0.01 (-0.04, 0.02) 4.32 Burr (2013) -0.01 (-0.04, 0.02) 14.21
Finnbogadóttir (2012) 0.10 (0.06, 0.15) 1.83 Finnbogadóttir (2012) 0.10 (0.06, 0.15) 12.31
Gough (2010) -0.02 (-0.05, 0.02) 3.41 Gough (2010) -0.02 (-0.05, 0.02) 13.65
Nissen (2013) 0.03 (-0.00, 0.06) 2.87 Nissen (2013) 0.03 (-0.00, 0.06) 13.95
Williams (1998) -0.05 (-0.05, -0.04) 50.36 Williams (1998) -0.05 (-0.05, -0.04) 15.51
Ziyab (2010) 0.01 (-0.01, 0.03) 11.03 Ziyab (2010) 0.01 (-0.01, 0.03) 15.01
Overall -0.01 (-0.02, -0.01) 100.00 Overall 0.01 (-0.02, 0.05) 100.00
(I-squared = 96.3%, p = 0.000) (I-squared = 96.3%, p = 0.000)
46 QP • www.qualityprogress.com
2017 WORLD CONFERENCE ON
QUALITY AND IMPROVEMENT
May 1 – 3, 2017 | Charlotte, NC
Addressing risk
Risk is still an immature concept in
International Organization for Stan-
dardization (ISO) documents because
there are different definitions of the
term. Risk was explicitly introduced as
a QMS requirement in ISO 9001:2015 and
defined as “effects of uncertainty.”
The new implicit ISO 9001:2015
requirement is for “risk-based thinking”
rather than “risk management,” which
is addressed in ISO 31000:2009—Risk
management—Principles and guide-
48 QP • www.qualityprogress.com
Having a process owner is a
lines (ISO, 2009). The revision now
requires an organization to identify
February 2017 • QP 49
STANDARDS OUTLOOK
You do.
50 QP • www.qualityprogress.com
THE VALUE OF SIX SIGMA
FORUM MEMBERSHIP
When you become a valued member of the ASQ Six Sigma Forum, you join a
distinguished group of problem solvers and analytical thinkers driven by a desire
to achieve breakthrough improvements. Your membership not only exposes you to
a remarkable network of individuals, but also links you to the Six Sigma Body of
Knowledge and supporting resources and case studies. Lastly, you will stay plugged
in to the Six Sigma community through monthly newsletters and members-only online
communities. So don’t delay, call ASQ Customer Care at 1-800-248-1946 and ask
to have the Six Sigma Forum added to your membership.
GET YOUR
ASQ CERTIFICATION
Looking for a way to stand out in 2017? ASQ certification can help with just
that. Choose from 19 certifications that are sure to not only differentiate you
from your competition, but can help result in a higher salary and help land
a promotion. It’s time to become a recognized expert in your field.
Apply for the April 2017 exams today!
videos.asq.org
Special Advertising Section
To be included in next year’s Six Sigma Resource Guide, contact Lou Brandow, 866-277-5666.
February 2017 • QP 55
Special Advertising Section
• No entrance or comprehensive
DEGREES IN QUALITY SYSTEMS MANAGEMENT exams
• Small, interactive classes
Bachelor of Science Degree Completion* - 10 courses over 12 months
with Juran Global Lean Six Sigma Green Belt Certification - Classes Start March 21 • Accelerated, accredited degrees
• Innovative, project-based learning
Master of Science Degree - 12 courses over 12 months
with Juran Global Lean Six Sigma Black Belt Certification - Classes Start March 22 • Weekly interaction with faculty and
fellow students
Contact info@ngs.edu or (800) 838-2580 ext. 505 to learn more. . • Financial Aid and scholarships
available for qualified students
Transform yourself. Become a leader in your industry.
NGS offers innovative, flexible degrees designed to support your THE NATIONAL GRADUATE SCHOOL
professional success and development. One-on-one interaction with OF QUALITY MANAGEMENT
faculty and support staff ensure that you won’t get lost in the crowd. Success you can measure
Ask about our special reduced tuition rates for ASQ members. (800) 838-2580 x505 | info@ngs.edu | www.ngs.edu/asq
NGS is accredited by The New England Association of Schools and Colleges (NEAS&C). NGS is a not-for-profit institution.
Certified to operate by State Council of Higher Education for Virginia (SCHEV). *Not available in MA. Not available to MA residents.
56 QP • www.qualityprogress.com
2
THE ASQ
GLOBAL
STATE OF
QUALITY
RESEARCH
58 QP • www.qualityprogress.com
for events affecting their business by using and utility applica-
either a mobile device or web browser. EtQ tions in combined
recently announced the addition of Life heat and electricity
Science features in its Traqpath product to generation. To ensure
meet the needs of regulatory compliance. stated accuracy classes, each heat
The Life Science features include: meter undergoes a three-point factory
• Forced authentication: Meet the needs calibration performed with water. coordinate-measuring
of 21 CFR Part 11 for electronic signa- With its direct path configuration, Opti- machines. They also can be used with
tures and records within your Traqpath sonic 3400 is not susceptible to (magnetite) 3-D laser scanning machines. With its
workgroup, on web and mobile devices. scaling-like meters that use reflectors. It of- aluminum-extruded structure, the metrol-
• E-Signature categorization: Offers the fers bi-directional flow measurement and a ogy tables can hold a lot of weight for any
ability to apply a meaning to your elec- fully welded construction without potential size, material or shape of part you want
tronic signatures to further define what leak points. Optisonic 3400 also features to inspect.
you are signing. enhanced diagnostic and status indications Options include either an English or
• Password management: Built-in and measurement of the velocity of sound metric-threaded plate for locating and
password rules help to keep security in to gain more information about medium elevating your parts off the surface for
line with regulatory standards, through and process. full inspection—whether using touch
automatic password aging and complex- • Email: info@krohne.com. probe or laser. R&R also offers a variety
ity rules. • Visit: www.us.krohne.com. of components to be used with the
• Field history and audit trail: Offers the metrology tables and are available in
ability to track the history of all fields— Metrology tables many sizes.
•
with each record having its own history. R&R Fixture metrology tables are designed • Visit: www.rrfixtures.com.
Users can access and download Traq- for fixturing parts and mounting of portable • Call: 616-847-6045.
path on their website or on a mobile device
through the App Store or Google Play.
• Call: 800-354-4476.
• Visit: www.etq.com.
•
Ultrasonic flowmeter
Khrone’s ultrasonic flowmeter Optisonic
3400 is available with heat meter approval
according to MID 2014/32/EU Annex VI
MI-004 in accuracy classes 1, 2 and 3.
Target applications include district heating
networks, commercial and industrial heat
metering applications, as well as energy
February 2017 • QP 59
QPREVIEWS
Semi-Markov Processes: optimization based on discrete state chang- “The amount of time it takes to complete
Applications in System es with concrete applications in industry. any single project is affected by the total
Reliability and Maintenance I. Elaine Allen number of projects that are active at any
Franciszek Grabski, Elsevier, 2015, 270 pp., San Francisco point in time.” For example, a basic rule of
$150 (book). lean: minimize work in process to reduce
This is a com- Innovating Lean Six Sigma: A cycle time.
prehensive and Strategic Guide to Deploying the This is a useful book for managers at
rigorous research World’s Most Effective Business all levels, but particularly C-suite leaders
text focusing on Improvement Process who want to lead positive change in this
continuous time Kimberly Watson-Hemphill and Kristine competitive, fast-moving world.
semi-Markov Nissen Bradley, McGraw-Hill Education, Bill Baker
processes (SMP) 2016, 304 pp., $35 (book). Santa Fe, NM
with a discrete This book is down
set of states. After to earth and The Certified Pharmaceutical
a mathematical practical with GMP Professional Handbook
and statistical many charts and Mark A. Durivage, ASQ Quality Press, 2016,
introduction to discrete Markov processes graphics to hold 516 pp., $89 member, $135 list (second
and semi-Markov processes, the author ex- the reader’s atten- edition, book).
pands and generalizes into important cases tion. The authors This book is a re-
of these models. Advanced methods are draw from their markable standout
presented with clear applications to quality experience, using by itself while
processes—making this text a good refer- multiple industries also representing
ence for statistical quality professionals. and situations a significant en-
The first six chapters cover the concepts with differing expectations for needed hancement of the
and equations of SMPs necessary to intro- changes as examples. first edition. This
duce more advanced models, specifically The content is divided into two parts. second version is
in the area of reliability and maintenance. Part one is for organizations—in the early not just a simple
These chapters are concise and present the stages of their improvement journey—that “refresh,” but an
characteristics of SMPs to enable math- must understand and apply the building overall revision of content and format. It
ematical analyses of the models presented blocks and culture to begin the transforma- represents current quality and regulatory
in later chapters. tion process. thinking about pharmaceutical quality and
The most interesting part of this book Part two is for those already well on expands the overall body of knowledge for
are the processes that have direct implica- their way to transformation and want to the certified pharmaceutical good manufac-
tions for quality-control applications such work on the big problems their organization turing practices professional (CPGP) exam.
as multitask operation and failure rate. In- is facing. Part two was right on target—fo- The book incorporates updated sources
cluded are numeric examples that illustrate cusing on customers, culture change and and regulatory references—domestically
the use of the models and descriptions strategy deployment (Hoshin planning) to and internationally.
of the algorithms inherent in the models. align at all levels and applying lean to prod- While serving as the central resource
The final chapter, on semi-Markov decision uct and process design and redesign. for ASQ's CPGP exam, it also can be used
processes, provides models for the optimal I really like the last part of chapter by industry and research to help address
strategies for maintenance operations and five, which addresses project pipeline theoretical and practical questions regard-
renewal processes showing algorithms for management by referring to Little’s Law: ing pharmaceutical quality, lab operations,
60 QP • www.qualityprogress.com
manufacturing and related activities. planning, process metrics and risk mitiga- Part seven of the book presents
The author lists 34 contributors to this tion, root causes, auditing, and beyond workshops, or practice exercises, on basic
edition and he shapes their efforts into a compliance and sustainability. quality and lean tools. These workshops are
comprehensive examination of what con- The book has 18 chapters with tools and thinking and activity-based and may be used
stitutes a safe and effective drug product. templates, plus a CD-ROM with document as a resource for a single reader or a study
The impact of activities—such as files. There is also team—possibly a team of students in a
development, technology transfer and a chapter offered course or training program. The workshops
manufacturing—are also examined and as a workshop are interesting and provide further under-
considered in relation to the many types of in which tools standing and insight for readers.
drug substances, drug products and related and procedures As a resource for leaders tasked with
packaging. explored in the delivering quality results, Smith’s book is on
The book's format has changed consid- book can be put the recommended list. The book also is a
erably and now includes eight major parts into practice. resource for the practitioner on the produc-
broken into 60 chapters. The book serves With the tion line, as well as the program developer,
as an excellent resource with research threepage glos- manager and the assessor of results.
presented through a detailed list of tables sary and references list, Smith’s book can Two resources in the book may be of
and figures, acronyms and abbreviations, a be shelved with the high-value reference value in the broad field of quality. First
preface, acknowledgements, four appen- collection. are examples of quality processes and
dixes, a glossary, a comprehensive list of As a resource tool, the book is focused resources presented in the list of examples
references and suggested readings and on outcomes and results—based on appearing as a section in the table of con-
then it closes with an index. knowledge and insights of happenings while tents. Second, the book itself is an idea-rich
The detail and logic of this organiza- the processes building results take place. resource for those needing examples of
tion allows the reader to understand many Though failure is referenced, emphasis is how to present a complex topic. It is full of
technicalities in a flawless fashion, while on successes, improvements and systems graphics, reference tables, charts and tools
covering a significant number of details, delivering predictable results. for users. This book is a reliable tool for
suggestions and potential problems. This book is a recommended resource experienced persons in the field, as well as
The book is easy to read and follow. for those in the broad, quality field including others new to the field.
Most everyone will find something valuable government services, education, healthcare Gerald Brong
in the writing because it is a useful and and fields beyond manufacturing. Ellensburg, WA
valuable tool for every level of experience.
Frank Pokrop
San Diego
Advertisers Index
The Art of Integrating Strategic
ADVERTISER PAGE PHONE WEB
Planning, Process Metrics, Risk
CyberMetrics Corp. OBC 800-777-7020 www.cybermetrics.com
Management and Auditing
Dekra Certification Inc. IBC 800-768-5362 www.dekra-certification.us
Janet B. Smith, ASQ Quality Press, 2016,
EtQ Inc. IFC 800-354-4476 www.etq.com
216 pp., $36 member, $60 (book and
Minitab 2 800-448-3555 www.minitab.com
CD-ROM). The National Graduate School 56 800-838-2580 www.ngs.com
Between the covers of this book, there is SigmaXL 54 888-744-6295 www.sigmaxl.com
a course. The course represented by this QI Macros for Excel 56 303-756-9144 www.qimacros.com Quality
book is organized into five parts: strategic Council of Indiana 1 800-660-4215 www.qualitycouncil.com
February 2017 • QP 61
For information on placing an ad,
www.amazonconsultinginc.com
Auditing
Training
Consulting
Q Email to
Sales@inspectionjobs.com
myQualityCloud.com
62 QP • www.qualityprogress.com
ONE GOOD IDEA BY BOAZ RONEN AND ALEX COMAN
Tips on Effective-Quality
Defining quality throughput using constraint management
WE WERE ASKED to measure the ef- do not result in a sale. This leads to cause- The input didn’t change while the EQ rose
fectiveness of a large commercial bank’s effect analysis on garbage time. significantly.5
audit department. The department’s output A study by the pharmaceutical company EQ is particularly effective at reducing
consists of recommendations for action— Merck shows that if 100 sales reps call on waste when bidding for new business.
scored from one to five according to the a physician, 17 depart without seeing the Orion, bidding for heavy-civil marine con-
level of potential risk. The common mea- physician because competitors’ reps are struction projects, submitted bids to proj-
surement is the number of recommenda- waiting for the physician and 27 drop off ects worth $670 million, and it won projects
tions that the department makes during the samples at the receptionist’s desk because worth a mere $76 million6—an EQ of 0.068.
year. A better measurement limits the count the physician is unavailable. Less than 24 Years later, Orion focused its efforts and
to severe recommendations. But these mea- conduct meaningful two-way discussions bid for projects worth only $429 million.
surements have two weaknesses: with the physician and are remembered Less bidding resulted in higher-quality bids.
1. They encourage auditors to make more because the average call lasts 4.6 minutes. 1
Orion’s wins rose to $117 million,7 and the
recommendations—in other words, re- An EQ of 0.24 is unacceptable, and Merck EQ rose to 0.27.
ducing the department’s signal-to-noise. urges “new ways of engaging physicians on A common mistake senior executives
2. The measurements ignore the auditors’ their terms … with greater productivity.”2 make is to measure output rather than
actual impact—in other words, the ef- We want to assess the throughput—the throughput (effective output). EQ drives the
fect they have on the organization. effective output—and measure its ratio to organization toward process improvement. A
It is essential to measure effective-qual- the system’s output. We noticed that many better process reduces waste—which in turn
ity (EQ), which equals throughput divided organizations do not even measure the improves the quality and value delivered. QP
by output. Throughput is the effective output of a process, but assume that the
REFERENCES AND NOTE
output of the system. system’s input (budget and person-months,3 1. John Mack, “Merck Rejiggers Its Marketing Mix,” Pharma
Marketing News, Vol. 6, No. 7, 2007, VirSci Corp., www.
Back to the bank’s audit department, the for example) accurately reflects its output. news.pharma-mkting.com/pmn67-article01.pdf.
EQ is measured as the number of recom- For example, many IT departments in 2. Ibid.
3. Businessdictionary.com defines “man month” as a unit of
mendations that were implemented by financial organizations are measured by work representing the productive effort of one person dur-
ing a four-week period. It is also called “labor month.” See
the organization divided by the number of person-months invested rather than by www.businessdictionary.com/definition/man-month.html.
recommendations actually made. The EQ— their effective throughput (for example, the 4. W. Chan Kim and Renee Mauborgne, “Tipping Point Leader-
ship,” Harvard Business Review, April, 2003.
represented in decimal points—can only number of features delivered, value added 5. Ibid.
be refined by counting the most important to the organization and bugs fixed). 6. Orion Marine Group, “Orion Marine Group Inc. Reports
Third Quarter 2011 Results,” press release, Nov. 3, 2011,
recommendations (scored four or five). Measuring the effectiveness of the New http://tinyurl.com/orion-pr-2011.
7. Orion Marine Group, “Orion Marine Group Inc. Reports Sec-
To evaluate such a department, we use York Police Department (NYPD) is another ond Quarter 2014 Results; Reports Record Backlog,” press
additional measurements, especially the example. Before Bill Bratton’s tenure as the release, July 31, 2014, http://tinyurl.com/orion-pr-2014.
absolute throughput of the department. To city’s police commissioner, NYPD seemed BOAZ RONEN is a professor (emeritus)
of technology management and value
assess continuous improvement, however, waste prone: widespread patrols of the creation at Coller School of Manage-
we use the EQ measurement. ment, Tel Aviv University in Israel.
subway system, massive paperwork and
He has his doctorate in business
attempted arrests during daytime hours. administration from Tel Aviv University
and co-authored Focused Operations
Making a sale When measured in terms of input, this seems Management: Achieving More With
Salespeople are measured by the amount reasonable.4 Existing Resources (Wiley, 2008).
of sales made. To focus on improvement ef- The waste is only salient when effective ALEX COMAN is a professor of
project management and value
forts, you must identify where waste hurts actions are measured. Bratton virtually creation at the Academic College of
Tel Aviv-Yaffo in Israel. Coman has
an organization. The EQ measurement eliminated wasteful tasks by replacing them his doctorate in IT from Claremont
highlights areas where a high number of with preventive quality-of-life crimes and Graduate University in California.
salesperson meetings with potential clients performing arrests during sleeping hours.
February 2017 • QP 63
BACK TO BASICS BY ARVED HARDING
N = sample size
0.05
Std. dev. = initial comparison of standard ARVED HARDING is a senior statisti-
0.04 standard deviation
cal associate with Eastman Chemical
deviations may at first appear
Co. in Kingsport, TN, and an adjunct
0.03
to be statistically correct, it instructor at Northeast State Com-
0.02 munity College in Blountville, TN. He
could very well not be the cor- earned a master’s degree in statis-
0.01 tics from Virginia Tech in Blacksburg.
rect conclusions from the data
0.00 Harding is an ASQ senior member
70 80 90 100 110 and problem at hand. and an ASQ-certified quality engineer.
Y
64 QP • www.qualityprogress.com
Your One Stop For All
Certification and Training
ISO 9001 | ISO 14001 | AS9100 | AS 9120 | ISO 13485 | ISO 50001
IATF 16949 | TL 9000 | ISO 27001 | ISO 20000-1 | ISO 22301 | FSSC 22000
Need training?
Established in 1989, AQS Solutions provides training in support of organizational improvement and
implementation of international management system standards. Offering a full range of innovative
training services to industry, government, non-profit, and service in areas of Quality, Environment,
Energy, Food Safety, and Occupational Health and Safety.