603999-DEC 2015 - Selected-Pages PDF

Industrial Engineer
Engineering and management systems at work
HELPING
THE HELPERS
Adaptive trai i g combats silo behavior a d
comm icatio reakdow s that bedevil
hum itari supply chai s
How simulation leads

to organizational insight
DECEMBER 2015
VOLUME 47 : NUMBER 12: $16.50
Make better decisions

in batch processing
Books an enduring source
of continuous improvement
Taking lean Six Sigma tools
into retirement
Expand Your Abilities

with IIE Training
Spend your professional development dollars wisely! Apply new knowledge to your work
immediately with hands-on practice, tools and ongoing support from IIE Training. Expert
instructors will help you understand new concepts and specific applications for your job.
Process Improvement Courses
Management and Specialty Courses
Six Sigma Green Belt

January 20-22, 2016 | Norcross, GA
Introduction to Engineering Analytics

March 21-24, 2016 | Orlando, FL
Engineering Decision Making and Risk

Management
April 26-28, 2016 | Norcross, GA

Six Sigma Black Belt
January 25-29, 2016 (Session 1)
February 22-26, 2016 (Session 2)
Los Angeles
March 14-18, 2016 (Session 3)
Staffing Models

Norcross, GA
Work Measurement
February 16-17, 2016 | Norcross, GA
April 4-8, 2016 (Session 3)

Lean Six Sigma Green Belt
February 29-March 4, 2016 | Norcross, GA
Healthcare and Ergonomics Courses

Six Sigma Green Belt for Healthcare
April 26-28, 2016 | Norcross, GA
Six Sigma Black Belt for Healthcare
Norcross, GA
April 4-8, 2016 (Session 3)

Lean Six Sigma Green Belt for Healthcare
February 29-March 4, 2016 | Norcross, GA
Principles of Occupational Ergonomics
March 29-31, 2016 | Norcross, GA
PE Exam Review for Industrial Engineers

Design of Experiments
March 17-18, 2016 | Norcross, GA
ING
AIN1948 CE
LE AR
ER
NT
Facilities and Workplace Design

IIE T
R
Statistics
N. IMPROVE. PROFIT.
Register at
www.IIETrainingCenter.org.
Table of Contents
December 2015 | Volume 47 | Number 12 | www.iienet.org/IEmagazine
Cover Story
Features
24
24 | Improving humanitarian logistics

Taking supply chain training to the front line taught
the teachers as well as the students
By Bublu Thakur-Weigold, Jonas Stumpf and Stephan Wagner
30 | Simulating improvement
Tips from decades of modeling can beneft your
manufacturing operation
By Edward J. Williams
36 | Baking in better batch processing

Standardized capacity analysis can produce results
throughout the organization
By John Preston
41 | These pages transfer knowledge

Your favorite improvement type person could
use a good book for the holidays
30
Compiled by Michael Hughes
the Front line
perspectives
16 | Performance
Nine magic numbers
36
18 | Management
How are you doing?
20 | Health Systems
The prediction predicament
22 | Innovation
41
DMDII tackles digital industrial innovation
12 | All I want for Christmas? A drone delivery

13 | Take fewer knocks to the head
13 | Sustainability, island style
14 | 3-Ds nothing; lets go 4-D
14 | Meeting demand
15 | How to make crazy talk not
15 | Teach entrepreneurship early
the institute
52 | Electioneering and hope
in every issue
12
44
4
Industrial Engineer | www.iienet.org/IEmagazine
6 | Editors Desk
10 | Trending at IIE
15 | Dilbert
44 | Case Study
46 | Research
50 | Tools & Technologies
56 | Careers
66 | Final Five
CO N F E R E N C E 2 0 1 6
Lead Efforts to Improve Healthcare Quality Discuss

Ideas, Applications and Research at the Healthcare
Systems Process Improvement Conference
Optimize your healthcare organization and abilities with new tools and
methods learned from your peers. HSPI offers innovative thinking in quality
improvement, case management, simulation, human factors and safety.
Connect with healthcare professionals tasked with quality, efficiency and
productivity in a collaborative environment that prioritizes discussion,
benchmarking and connections.
Bring your
co-workers!
Team discounts save
you $50 to $100 per
registration. Visit
www.shsconference.org
and click Registration
to learn more.
Visit www.IIETrainingCenter.org to view all upcoming courses and full descriptions.
Register now! www.shsconference.org
December 2015 | Industrial Engineer
editors desk
Teaching the teachers

humanitarian logistics
To reach me,
email mhughes@iienet.org
or call (770) 349-1110.
Transmitting specialized operational knowledge to the front lines can

be a diffcult task in any organization.
The hurdles increase in sectors such as humanitarian logistics,
which often deals with personnel who dont have access to structured
learning or advanced education. Face it, in far-fung undeveloped
areas ravaged by repeated wars or natural disasters, most people wont
be earning their masters in supply chain operations anytime soon.
In such cases, lessons from the front lines or teaching the teachers, if you will can be invaluable. The Swiss Federal Institute of
Technology Zurich (ETH Zurich) and the Khne Foundation found
that out over the past several years as they adapted their humanitarian
logistics training program.
Bublu Thakur-Weigold, Jonas Stumpf and Stephan Wagner detail
the lessons in this months cover story, which starts on Page 24. Although the title is Helping the Helpers, its clear that the teachers
learned plenty. In many instances, they discovered that the problem
wasnt the gap in skills, it was poor information fow. Action learning,
using simulations of the participants own system and making sure
the class has personnel from multiple functions, not just the logistics
department, were also important.
The teachers have incorporated the lessons into the program, a
move that truly could beneft those in need. Far from being miserly
and sparse, humanitarian givings accumulated donations have grown
eightyfold since 2000.
The catch, of course, is that many billions of those dollars are wasted. Instead of corruption and incompetence, the wastes in humanitarian operations come from the same bad actors that bedevil commercial supply chains: process dysfunction, silo behavior, redundant
work and communication breakdown.
Humanitarian relief agencies worldwide are looking to improve
their operations so that more help gets to victims of disasters, both
natural and manmade. Examining Helping the Helpers is a good
place to start.
Promote Achievement through

Society and Division Awards
Nominate a colleague for a society or division award to help recognize
INDUSTRY AWARDS
Engineering Economy Wellington Award
Manufacturing & Design (M&D) Outstanding Service Award
Quality Control & Reliability Engineering (QCRE) Golomski Award
TEACHING AWARDS
Engineering Economy (EE) Teaching Award
Lean Teaching Award
Sustainable Development Teaching Award
STUDENT AWARDS
Engineering Economy (EE) Undergraduate Senior Design Award

Lean Student Paper Award
Manufacturing and Design (M&D) Division Student Paper Award
Process Industries (PID) Student Paper Award
Society for Engineering & Management Systems (SEMS) Student Paper Award
Society for Health Systems (SHS) Student Paper Award
Student Chapters IAB YouTube Contest
Sustainable Development Student Paper Award
Check website for deadlines. Visit www.iienet.org/SDAwards.
Industrial Engineer
Executive Editor
Art Director
Monica Elliott
melliott@iienet.org
(770) 449-0461, ext. 116
Tara Ott
(770) 449-0461
Managing Editor
Director of Multimedia
Advertising Sales
Michael Hughes
mhughes@iienet.org
(770) 349-1110
Hope Teague
hteague@iienet.org
(770) 349-1127
Web Managing Editor
Exhibit Sales
David Brandt
dbrandt@iienet.org
(770) 449-0461, ext. 120
Dolores Ridout
ridout3@airmail.net
(281) 762-9546
serves the diverse audience of professionals and students whose

common interest is industrial engineering. Our mission is to provide useful,
interesting, timely and thought-provoking content that addresses the broad
spectrum of industrial engineering practice in all industries. As the pre-eminent
voice of the profession,
strives to give readers information they
can use to enhance their professional capabilities, improve their organizations
performance and advance the development of their profession.
In furtherance of this mission, we adhere to the following objectives:
To present accurate reporting and analysis of the most prevalent industrial
engineering topics
To serve as a career development resource to students and professionals
To maintain high editorial standards, journalistic integrity and credibility
To support the mission of IIE in its service to members and the industrial
engineering profession
INSTITUTE OF INDUSTRIAL ENGINEERS
BOARD OF TRUSTEES
President
Senior VP-at-Large, Industry
Senior VP, International
VP of Student Development
James E. Moore II, Ph.D.,

University of Southern California
Joan Wagner, P.E., Spirit

AeroSystems Inc.
Bopaya Bidanda, Ph.D.,

University of Pittsburgh
Jennifer Cross, Ph.D.,

Texas Tech University
President-Elect and CFO
Senior VP, Regional Operations
Senior VP, Publications
Secretary
Michael D. Foss,
Cameron International
Christopher Geiger,
Universal Orlando Resort
Alice E. Smith, Ph.D., P.E.,

Auburn University
Ariela Sofer,
George Mason University
Immediate Past President
Senior VP,
Continuing Education
Senior VP,
Technical Operations
Chief Executive Offcer
Dennis Oates, Amazon
Senior VP-at-Large, Academic
Scott Mason, Ph.D.,

Clemson University
Toni L. Doolen, Ph.D.,

Oregon State University
Don Greene,
Institute of Industrial Engineers
Randa Shehab, Ph.D.,

University of Oklahoma
POINTS OF CONTACT
Institute of Industrial Engineers
Chapters, Societies
and Divisions
Corporate Partnerships
and Strategic Alliances
Nancy LaJoice, ext. 122

nlajoice@iienet.org
Douglas Long, ext. 109

dlong@iienet.org
Customer Service
Annual Conference
Continuing Education
and Corporate Training
Bill Gibbs, ext. 126

bgibbs@iienet.org
Larry Aft, P.E., ext. 130

laft@iienet.org
3577 Parkway Lane, Suite 200

Norcross, GA 30092
www.iienet.org
(770) 449-0461
Operations
Donna Calvert, ext. 108
dcalvert@iienet.org
Publications
Ext. 102
cs@iienet.org
Monica Elliott, ext. 116

melliott@iienet.org
Website
David Brandt, ext. 120
dbrandt@iienet.org
(ISSN 1542-894X) is published monthly. Copyright 2015 Institute of Industrial Engineers. Established 1969. Subscriptions for members included in annual dues, not deductible. Single copy $16.50. USA
subscriptions: per year $194; two years $342; three years $437. Outside U.S.: one year $241; two years $411; three years $558. Airmail $99 additional. Institutions and agencies, call for rates: (770) 449-0460. Editors Note:
We treat all communications as letters to the editor unless otherwise instructed. This publication is designed to provide accurate information in regard to the subject matter covered. It is provided and disseminated with the
understanding that the publisher is not engaged in rendering legal or other professional services. If legal advice or other expert assistance is required, the services of a competent professional should be sought. Content:
Publication does not constitute endorsement of any product or material, nor does IIE necessarily agree with the statements or opinions advanced at its meetings or printed in its publications. This magazine acts as a
moderator, without approving, disapproving, or guaranteeing the validity or accuracy of any data, claim, or opinion appearing under a byline or obtained or quoted from an acknowledged source. All issues of
Industrial Engineer are available on microfilm or photocopy from University Microfilms, 300 Zeeb Rd., Ann Arbor, MI 48106. Copies of articles in
for personal or internal use may be made, with the
consent of IIE, in accordance with copying permitted in Section 107 or 108 of the U.S. Copyright Law (P.L. 94-553). For copying beyond that permitted by law, the copying agency must pay $3 per article copy through the
Copyright Clearance Center Inc., 21 Congress St., Salem, MA 09170, (978) 750-8400, using the following code: 0019-8234/90/$03.00/0. Copyright consent does not extend to other kinds of copying, such as copying for general
distribution, or advertising, or promotional purposes, for creating new collective works or for resale. Periodicals postage paid at Norcross, Ga., and at additional mailing offices. Permission requests should be submitted to the
Copyright Clearance Center at www.copyright.com.
Postmaster: Send address changes to:
, 3577 Parkway Lane, Suite 200, Norcross, GA 30092. CPC Publication sales agreement #1459430. In Canada:
Canadian returns: Access Worldwide, 1415 Janette Ave.,Windsor, ON N8X 1Z1. Printed in USA.
publications agreement #40031822.
March 21-24, 2016 | Disney's Coronado Springs Resort | Orlando, Fla.
Gather with Leading Minds in Ergonomics Attend AEC 2016!

Discuss best practices, research and applications with professionals in the ergonomics, human
factors, safety, risk management and industrial hygiene felds!
Early-bird registration ends January 10. Register now!
Hands-on Learning
One full-day and eight half-day pre-conference workshops are your opportunity to explore
ergonomics topics in depth. Plan to arrive early and enhance your skills!
Ergonomic Principles and Trends for the Ofce Environment
How Do I Choose the Correct Ergo Assessment Tool(s)?
Applied Anthropometrywww.appliedergoconference.org
for Practitioners
Ergonomics Programs: Strive for the Possible; Implement the Practical; Continuously Improve
Cost Justifcation for Ergonomics
The Bad Ergonomics of Order Fulfllment in Distribution Centers
May the FORCE Be with You: Improving Your Methods to Quantifying Forces in the Workplace
Leadership in Ergonomics: Overcoming the Hurdles of an Ergonomics Program
Ergonomics Certifcation Review: Determining What You Do and Do Not Know (full day)
www.appliedergoconference.org
Trending @iie
Throughout the year, we have covered the upcoming vote on whether to add the word systems to
IIEs name, culminating in the article Name Recognition in the November issue. Below is a letter
from a 37-year member of the institute sharing his thoughts on the name change recommendation.
And be sure to review other election details and offcer candidates in The Institute on Page 52.
Mail
Name change is not the same as rebranding
I carefully read the Name Recognition comments of former
President Al Soyster and current President Jim Moore (November). With regard to the question of IIE or IISE, I am not
sure how much of a difference it will make. I am concerned,
however, that the membership not think that a name change
will address what in my opinion are much deeper issues for our
profession and our association.
In the introduction to the piece, the name change was described as an investment in rebranding, and in Mr. Moores
response he points out both that a more quantitative assessment would come at a nontrivial cost of its own and that
the institute is doing well fnancially.
We must be clear that a company or organizations brand is
not its name, its logo or its tag line. These are tools to reinforce
and communicate the brand. The brand itself is reputation amplifed by visibility.
Hours of discussions at conferences and meetings show that
as a profession we have a hard time reaching consensus on
the value proposition we provide to clients, and we have not
invested in fnding out what clients think the profession has
to offer them which may be very different from what we
assume they think.
The same holds true for the association in terms of the value
proposition to an enormous pool of potential members. Bot-
10
tom line, we do not have an effective strategy for enhancing

our reputation and increasing our visibility, and this will not
be solved simply by changing our name.
Looking at U.S. Department of Labor Statistics and association membership statistics reported on Wikipedia, our successes and challenges are clear. As shown in Figure 1, industrial
engineering is one of the four largest engineering professions
in the United States.
Although I do not show them, none of the other engineering professions have even one-third the number of practitioners we have.
Our profession is important, ubiquitous and largely invisible. We improve products, processes, systems and organizations that other professions claim (and the public sees) as the
result of their efforts, not ours. Figure 1 also shows that the
professional organizations of the other three large engineering disciplines have almost 10 times or more the number of
members relative to IIE. This may be because of the splintering of industrial engineering organizations into specialty
associations.
I would suggest that at a time when the institute is doing well fnancially, it would be a better investment to spend
some or all of the reported $100,000 on hiring branding professionals who can apply best practices to perform independent
research regarding what others value in our work, to better defne the value of the profession and IIE, to benchmark against
our sister professional associations, and to then develop a long-
FIGURE 1
By the numbers
According to the U.S. Department of Labor, industrial engineering is one of the countrys top four largest engineering professions.
U.S. Department
of Labor employment
by title
Employment numbers
(2012)
Professional
association
Association
global membership
Electrical and electronics

engineers
306,100
IEEE
430,000
Civil engineers
272,900
ASCE
140,000
Mechanical engineers
258,100
ASME
140,000
Industrial engineers,
including health and safety
247,400
IIE
14,000
term strategy to improve the reputation and visibility of our

profession and our association.
This strategy may include collaboration with (or perhaps
consolidation with) other industrial engineering-oriented
associations, and it may even include a recommendation,
grounded in research, to change our name, logo and tag line.
John M. Corliss Jr.
IIE fellow and founder of the IIE Sustainability Division
Andover, Massachusetts
Share and discuss

Wed love to hear from you. Send letters to the editor to Michael Hughes
at mhughes@iienet.org or be retro and mail them to his attention at
3577 Parkway Lane, Suite 200, Norcross, GA 30092. And join the
discussion on IIEs social media sites by sharing your professional
insights, questions, multimedia, kudos and more. Go to www.iienet.org/
networking to get into the conversation.
11
The front line
News from the feld
All I want for Christmas? A drone delivery

FAA allows HorseFly octocopter to test fights around Ohio airport
If Steve Burns gets his way, youll get your holiday order
via the frst FAA-approved drone delivery just in time
for Christmas.
Of course, your house must be in the 5-square-mile
area surrounding Wilmington Air Park in Ohio. Burns
company, Workhorse Group, has Federal Aviation Administration approval to test and do live deliveries with
his companys drone-equipped electric trucks. The
company and the University of Cincinnati have been
working on the HorseFly octocopters for two years.
The goal is to make the frst FAA-approved drone
package delivery.
The drones attach to the roof of the companys
trucks. Drivers load a package into the drones basket. A
fat-screen panel in the truck shows the satellite image
of the delivery address, and the driver touches the screen
to place the drone in the exact delivery spot.
HorseFly octocopters cost only 2 cents a mile to operate compared to
While Amazon famously wants drones to deliver
the roughly $1 a mile for a standard delivery truck.
directly from a central warehouse, adding the human
factor of a courier who knows the neighborhood is benefcial, Burns noted. It gives the FAA comfort to have
Watch that drone
a person nearby in case something goes wrong, and the
To view the HorseFly octocopter
driver can recall the drone by pressing a button.
delivery process in action, visit
Eventually, Burns said, the FAA will move beyond
http://bit.ly/1LpRqlx.
line of sight, and thats where the real economic benefts
of a driver plus a truck plus a drone take off. Delivery
companies have optimized current technology and logistics to besides dragging a 20,000-pound truck up to your doorstep.
where theres not much left to squeeze. But improving techImagine a driver at a stoplight with one delivery a mile to
nology has limits after 100 years of refnement, the typical her left and three to her right. She gives the delivery to the left
medium-duty delivery truck still gets only about 5.5 mpg and to the HorseFly and completes her three stops to the right. The
costs about $1 a mile to operate. Electric trucks cost about 30 drone delivers and returns to the truck.
cents a mile.
Drones wont make economic sense for every delivery. In a
But the HorseFlys fuel costs are only 2 cents a mile and residential area with three stops in a row, the drive could be a
thats a mile as the crow fies, literally.
quicker option. But for the delivery thats a mile in the oppoWith those kinds of dynamics, thats how we know this is site direction or a half mile up a farmers driveway, the Horsegoing to happen, Burns said.
Fly could signifcantly reduce operational costs, Burns said.
Its too compelling, and with e-commerce, well the
Saving him that 1 mile round trip to the left could have
Achilles heel with e-commerce is in the end a big truck has to been the one [the delivery company] lost money on, Burns
bring it to your house and take it back if you dont like it. And said. [Companies] dont make money on every delivery.
this is all for a fve-dollar book. So theres got to be something Some of them are just a pain.
12
Take fewer knocks to the head

App makes athletes think about concussions
Despite all the publicity about sportsrelated concussions, many student-athletes dont recognize the symptoms or
wont report them if they do.
University of Arizona systems engineering professor Ricardo Valerdi wants
to change that with an app. Having
student-athletes spend 10 minutes on a
virtual athletic feld that shows them the
immediate and delayed side effects of
concussions could change their behavior
and attitude toward head injuries.
Valerdi, also I dustrial E gi eers systems engineering columnist, and his
colleagues at the University of Arizona
College of Medicine-Phoenix have
made it through the second round of the
NCAA Mind Matters Challenge, secur-
ing $100,000 to build a prototype and

release it to athletes.
They will present their prototype this
winter to NCAA offcials, and the winning approach will be made available to
some 400,000 NCAA student-athletes.
The more student-athletes know
about concussion and the risks of hiding
symptoms, the more confdent theyll
be in making the right choices, Valerdi
said. And the right choice is simple:
Dont play through a suspected concussion.
The app is designed to be a tool for
athletic training programs and will
be free with any smartphone. It uses
Google Cardboard, a foldout cardboard
mount with lenses, magnet and fasten-
Ricardo Valerdi hopes his teams app

will teach student-athletes to avoid
playing through concussions.
ers that can become an instant virtualreality headset when assembled with a
smartphone slipped inside.
Sustainability, island style

The Gorona del Viento hydrowind plant eventually could provide all the energy necessary for the
isolated Canary Island El Hierro,
according to an industrial engineer
working on the project.
Juan Pedro Sanchez told the
BBC that the plant covered all of
the islands energy needs for two
hours on Aug. 9.
The next steps will be to try it
for 24 hours and then weeks at a
time, he said.
I think that in a year or so, the
plant could supply all the electricity the island needs for about 200,
250 days, Sanchez told the British
news organization.
The island, which relies on supplies of diesel fuels shipped
over unpredictable seas from Tenerife, 124 miles away, saved
300 tons of fossil fuels in July, a number that could increase to
500 tons a month before long. That equals 40,000 barrels of oil
and 19,000 tons of carbon dioxide emissions per year.
The system has fve wind turbines and two reservoirs.
Photo courtesy http://gleanproject.org
Hydro-wind combo plant could generate all the electricity El Hierro needs
When the wind is blowing, excess power pumps water from

the lower reservoir to one that is 2,300 feet above sea level.
When the wind dies down, that water falls through a set of
hydraulic turbines to generate electricity, ameliorating the key
disadvantage of wind power its unreliability.
Engineers say the island will have to expand reservoir capacity if it wants to eliminate the use of fossil fuels.
13
Technology allows self-folding of complex objects

In the future, the words some assembly required could mean heating up your new desks
components so they self-assemble.
Researchers at Georgia Tech and Singapore
University of Technology and Design have used
smart shape-memory materials with different responses to heat to demonstrate four-dimensional printing, which creates complex self-folding
structures. The material responds to external
stimuli such as temperature, moisture or light to
change to a programmed shape.
The demonstrations produced a mechanism
that can switch from a fat strip into a locked
confguration as one end bends and threads itself
through a keyhole, along with a fat sheet that
can fold itself into a box with interlocking faps.
Each smart-shape memory polymer responds at
a different rate to application of uniform heat.
Possible applications include space structures, deployable medical devices, robots, toys and other
structures.
Earlier work with such materials required applying different levels of heat, which made the
process complicated. The team turned that approach around and used a uniform temperature,
which is easier to apply, exploiting the ability of
different materials to control their rate of shape
change through their molecular design, said
Georgia Tech professor Jerry Qi.
The work, Sequential Self-Folding Structures by 3-D Printed Digital Shape Memory
Polymers, was published in the journal Scie tifc
Reports.
Photos courtesy Qi Laboratory, Georgia Tech
the
thefront
frontline
line
3-Ds nothing; lets go 4-D
Starting at the top, this series of photos shows the self-folding process of a
box from beginning to end. This folder box is intended to simulate a postal
mailer.
Prime Number
Meeting demand
Industrial engineer ranks 24th on Indianas list
of Hot 50 jobs, the dia apolis Busi ess Jour al
reported. The University of Indianapolis will begin
offering bachelors degrees in industrial engineering
in fall 2016 to meet rising demand. IEs earned a
mean salary of $72,170 in Indiana last year, $85,110
nationally, according to the journal.
14
24th
How to make crazy talk not
Book
Help arrives for those who have to explain the rational to the irrational
Industrial engineers try to live in a world of rationality where the numbers lead to conclusions. Unfortunately,
co-workers, associates and CEOs dont always inhabit that
world, which leads to process improvement types dealing
with all sorts of irrationality.
Perhaps Mark Goulston can help with Talki g to Crazy:
How to Deal with the Irratio al a d Impossible People i our
Life. The psychiatrist and crisis counselor offers a six-step
sanity cycle, useful for understanding when a person is
unable to think rationally, not take it personally, offer empathy and gradually guide that person back to a saner way
of thinking.
Chapters detail how to keep your own crazy at bay when
under attack, ways to handle such strife in your personal life
of the
Month
and 14 tactics for talking to crazy.

In particular, The Butter-up: Getting a Know-It-All to Behave could
be quite effective with that recalcitrant C-level executive.
Through it all, Goulston explains
why people act in unreasonable
ways, giving insight into the brains
natural defense mechanisms and
how to recognize an irrational persons modus operandi.
Talki g to Crazy: How to Deal with the
Irratio al a d Impossible People i our Life is available from
Amacom ($24.95).
Quote, unquote
Teach entrepreneurship early
People become entrepreneurs because they think they are good at it and are going to
be successful, but students dont always feel that way when they graduate. Our fndings
show the need for more goal-specifc programs that give students the confdence that
founding ones own frm can be a controllable and potentially successful career. Colleges and universities can play an important role in convincing students that the noncorporate path is a viable option.
Erik Monsen, shown here with the University of Vermonts Entrepreneurship Club, describing his upcoming Journal of Small Business
Management article Founder, Academic, or Employee? A Nuanced Study of Career Choice Intentions. Monsen is the Steven Grossman
Endowed Chair in Entrepreneurship at the University of Vermont.
Dilbert
2014 Scott Adams. Used by permission of UNIVERSAL UCLICK. All rights reserved.
15
performance
Nine magic numbers
This article fell into place as I was wrapping up another Baldrige Performance
Excellence Award site visit. This year,
my focus was on information analysis
and sharing processes, which suited me
just fne.
As I prepared questions for my visit,
new queries popped into my mind relative to ideas, best practices and innovations. Specifcally, what is the difference between each one, and does it
make sense to track the frequencies
at which we are generating, evaluating and implementing them?
Most people would struggle to
explain the difference between an
idea, a best practice and an innovation. I did at frst, but the more I
thought about it, clarity and distinctions
emerged. Ideas come frst. Some ideas
are good, some not. Some have been
tried before and worked, while others
have failed time after time. We are after
only the ideas that give us great results.
Best practices are ideas that are proven to work (or at least they should be).
Their results compare well against relevant benchmarks. If you want to be one
of the best hospitals nationally for a given measure, your results for that measure
should compare favorably with national
top percentiles. Best practices are needed to sustain such results. Be cautious,
though, as best practice claims are made
much more often than they are proven.
True innovations are even harder to
fnd. Innovations lead to breakthrough
results what some call step changes
in performance. As with best practices,
16
By Kevin McManus
claims of innovation tend to be overstated. People often think things are innovative simply because they have not seen
them before. In other cases, a change is
new to a given business sector, but commonplace in others.
As I thought about ideas, innovations
and best practices, the notion of the nine
magic numbers came into my mind.
The three improvement types would
Sadly, best practices and true

innovations are tough to fnd
for this process area.
make up one side of this matrix. Three
rate types would make up the other side
of the table the generation, evaluation
and implementation rates for each type
of improvement. High performance organizations should know the nine magic
numbers that are driving performance
improvement across their work systems.
Test yourself can you describe the
difference between an idea, a best practice and an innovation? If so, do you
know the rates at which each type of
improvement is being generated across
your work groups? Do you also know
your respective evaluation and implementation rates?
Most organizations fail to provide evidence for their three idea rates: the rates
at which ideas of any type good or
bad are being generated, evaluated and
implemented over time. Ask for similar
rates for higher order improvements like

best practices and innovations, and the
result tends to be the deer in the headlights look. How many great ideas are in
the bank for future use?
Can people claim to be, let alone
prove they are, high performers if they
cant produce the nine magic numbers?
Are we willing to focus on outcomes
alone to tell us if our work systems are
sustainable and agile enough for the
future? As with most measures, the
nine rates tell us different things
about our idea generation and stakeholder engagement processes both
the rate at which we are currently
improving and what our improvement potential might be.
The nine magic improvement numbers are needed to give us insight into
who we are involving and how we are
involving them in our formal process
improvement efforts. Ad hoc, poorly
aligned improvement efforts produce
inconsistent results coordinated idea
capture, evaluation and implementation efforts contribute signifcantly to
accelerated progress on the road to operational excellence. Sadly, best practices
and true innovations are tough to fnd
for this process area.
Kevi
cMa us is a performa ce improveme t coach based i ai ier, Orego , a d
a 33-year member of IIE. He has writte
workbooks about perso al a d team effectiveess. McMa us is a lum i exami er for the
Malcolm Baldrige Natio al Quality Award.
Reach him at kevi @greatsystems.com.
17
management
How are you doing?

By Paul Engle
Pla s ar othi g; pla i g is everythi g.
Dwight Eise hower, U.S. Army Chief of
Staff a d 34th preside t of the U ited States
Many readers spend a lot of their time

developing and implementing plans to
achieve strategic goals, measuring performance against these plans. Annual
budgets, projects and strategic business
plans represent our attempt to direct
company resources effectively and effciently.
Most performance goals and metrics
are fnancially driven. Sales revenue,
gross margins, expenses and profts rank
high on the list of company objectives. Bob Kaplan and David Norton revolutionized this process by
introducing the balanced scorecard,
described in their 1996 book The
Bala ced Scorecard: Tra slati g Strategy i to Actio . For the frst time, management monitored key performance
indicators as well as fnancial metrics
when tracking an organizations progress against strategic plans.
Key performance indicators typically come in two favors: lagging and
leading. Lagging indicators such as last
months revenue and expenses are compared to the previous period or year.
This information may indicate a trend
that, when extrapolated into the future,
can guide decisions. Lagging indicators
represent historical facts.
However, leading indicators may
be even more important because they
might allow management to take direct
action to improve future performance.
18
Sales revenue is an excellent example.

Most organizations understand their
revenue cycle and assign signifcant resources toward meeting annual goals.
Sales and marketing costs take a large
chunk. Many executives struggle to understand how effectively these resources
are deployed. Leading indicators may
provide the answer.
Advertising campaigns can sway consumer behaviors and grow sales. The
amount and effectiveness of advertising
may be a leading indicator of future revenue increases. Metrics associated with
brand awareness and addressing con-
Leading indicators may be

even more important.
sumers' needs give management a basis
for forecasting future demand.
Since companies control their advertising spend and determine the most effective media campaigns, management
decisions affect these companies future
revenue growth. Advertising spending
and effectiveness are important leading
indicators of revenue and should be considered when developing sales forecasts.
Other leading indicators include employee productivity metrics. Employee
training and incentive plans can improve
fnancial results. Resources allocated to
training and bonuses may provide management with an important leading indicator for employee productivity.
Executives monitor external lead-
ing indicators because they might affect

company results. An example might include foreign currency exchange rates.
Todays strong U.S. dollar enhances our
global competitors ability to reduce
prices and grow their market share. Domestic producers may experience lower
revenue, pricing, margins and profts
as a result. Because interest rates infuence exchange rates, tracking short- and
long-term rates may provide management with a powerful leading indicator
of future revenues.
Selecting lagging and leading indicators may require management to
take a fresh look at their processes
and the external factors that affect
their business. Analyzing historical data may provide clues. Use of
big data, defned as vast amounts of
unstructured data points related to
all aspects of a companys business, also
can supply management with important
decision-making tools.
Can a company accurately forecast
future performance using lagging and
leading indicators? In many cases, the
answer is yes, provided management
selects the most important metrics that
impact performance and takes timely
actions to maximize improvement.
Paul E gle is a ma ageme t co sulta t
with a
BA i
a ce. He has more
tha 0 years of experie ce i a ageme t,
operatio s, product developme t, sales a d
marketi g, strategic pl i g d busi ess
process improveme t. You may co tact him at
paulfe gle@outlook.com.
ARE YOU CONNECTED?

IIE makes it easy for members to connect with other IE professionals. Members have access to:
12 FREE DIVISIONS
JOIN ANY OR ALL
Applied Ergonomics (GOErgo)

Computer & Informaton Systems
Constructon
Engineering Economy
Lean
Logistcs & Supply Chain
Manufacturing & Design
Operatons Research
Process Industries
Quality Control & Reliability Engineering
Sustainable Development
Work Systems
IIE CHAPTERS
PROFESSIONAL & STUDENT

Enhance your professional development
Build your leadership skills
Connect with peers, experts, even IE legends
Advance the reputaton of the profession
Partcipate in regional conferences
TWO SOCIETIES
IIE MEMBERS CAN CHOOSE

TO JOIN FOR AN ADDITIONAL $35 EACH
The Society for Engineering &
Management Systems (SEMS)
Examine management operatons and business
processes, and discuss solutons that can aect
the whole organizaton.
Society for Health Systems (SHS)
Improve healthcare systems by connectng to an
actve community of peers who share ideas, case
studies and applicatons.
IIE YOUNG
PROFESSIONALS GROUP
UNITING AND MEETING THE NEEDS OF
MEMBERS AGED 35 AND YOUNGER
Strengthen relatonship with the industrial and

systems engineering body at large
Create personal and professional growth
opportunites
Develop leadership skills
Make connectons and network
Get career guidance and job opportunites
Support the transiton to the professional
community
IIE members also have exclusive access to the member directory.

Find and connect with professional IEs in your town, your state, your eld globally.
ONLY IIE MEMBERS HAVE ACCESS TO THESE BENEFITS

To learn more, visit www.iienet.org or contact the IIE customer
service team at cs@iienet.org or 800-494-0460.
19
health systems
The prediction predicament
Forecasting is an ungrateful profession.

Many times engineers working in
the healthcare sector are called upon to
provide models that forecast or predict
activity to optimize resources. As engineers we tend to want to use those techniques and methods that will provide
the best and most accurate predictions.
The problem is we also are treated
like the weyward sisters of William
Shakespeares tragedy Macbeth.
As you might recall, the frst time
the witches enter Macbeths life and
give answers, Macbeth, while apprehensive, tends to be happy because they give him good news. His
doubts about their abilities fade, and
he plans his ascension to the throne.
However, the second time, the
news is not so good, and he brushes
them aside and takes a more optimistic
view of things than the actualities the
witches provide.
This is the folly of forecasting things
that no one really wants to know.
In a similar vein, health systems engineers use methods that are a bit more
scientifc than the nonsense chanting of
Double, double, toil and trouble ...
and the whole eye of newt and toe of
frog business, but to most nontechnical
executives it sounds about the same.
This, again, does not really seem to
matter, at least until you are wrong or
more correctly, the method produces an
inaccurate forecast.
Was it three toes and one eye? Or
one toe and two eyes? Ah, yes, it was
the loaded overhead estimate that was
20
By William Ike Eisenhauer

fed incorrectly to the AMARA time
series prediction BIC model selection
criterion module that was the source of
the problem.
Upon uttering this during your visit
to the boardroom, of course, you receive
a wall of blank stares.
This is the folly of using methods that
are too complex for the end user to diagnose. If you are forecasting things that
no one really wants to know (folly No.
Knowing where and how

the forecast missed its mark
is just as important as
heeding its news.
1) and the end user cant fgure things
out (folly No. 2), you will not succeed.
You typically can avoid the frst woe
by asking a hard question ahead of time.
Do you have an action plan in place that
depends on the results of this forecast?
If there is not one, then there is no need
for the forecast because no one wants to
hear it anyway.
In the case of the second, it is a reality
of forecasting in healthcare operations
(as opposed to the dealings of kings) that
one who is responsible for the actuals
not adhering to a forecast must be able
to explain why. If your audience cannot
comprehend the methodology, limitations and causal chains in a forecast in
order to match them up with actual performance parameters, they are left in the
precarious positions of stating either I

have no idea or I do not understand
how the forecast is made. They need to
know why it has to be an owlets wing
in the cauldron and not just any old bird
appendage.
This is a risky position to be placed in.
It is not a lack of awareness or availability of the latest techniques and predictive
methodologies; it is purely a risk mitigation strategy. Knowing where and how
the forecast missed its mark is just
as important as heeding its good
and bad news, with just enough
blind-worms sting to take action if the prediction is not what
you expected.
As engineers we need to be cognizant of these follies and realities
when we introduce our models,
techniques and forecasts to provide the
most valuable, useful information in an
understandable manner.
William Ike Eise hauer is a gi eeri g
professor at Portla d State U iversity a d
atio al director of the Vetera s E gi eeri g
Resource Ce ters for the Vetera s Health
Admi istratio . His i terests are i tegrati g
e gi eeri g a d healthcare professio als to
i crease the value of health systems a d adva ci g e gi eeri g scie ce to address healthcare delivery challe ges. He ca e reached at
wde@pdx.edu.
Inspiring Ideas, Connections

and Professional Development
All at the IIE Annual
Conference & Expo 2016
Get excited for the largest industrial and systems engineering
event of the year. Workshops, case studies, thought leadership,
research and connections are all reasons to attend.
Heres what makes this event so great:
Case
studies and applications by industry
Thought
leadership from keynote, oral and poster

presentations
Research
from academia and industry scholars
Workshops
that provide new techniques and

understanding of important topics
Connections
Tours
with your colleagues from around the world
that show how ISE principles are put into action
Exhibitors
showcasing the latest products and services
Recognition
of achievement at all career stages
Register now!
www.iienet.org/Annual
21
innovation
DMDII tackles digital industrial innovation
The federal government hopes investing in digital manufacturing can drive

prosperity and help rekindle American
industrys competitiveness.
Washington hopes to stimulate the
same kind of creative synergies found
in much of Europe and several Asian
countries, where industry and government often join forces to motivate
industrial growth. A case in point is
the Fraunhofer Society, an applied
research organization with 67 dedicated institutes in Germany investigating in felds as diverse as optics
and nanotechnology to industrial
mathematics and process engineering.
In 2012, the U.S. government founded the National Network of Manufacturing Innovation (NNMI) as a joint
federal initiative by the Department
of Defense, the Department of Energy, the National Science Foundation
and the National Institute of Standards
and Technology. Modeled after the
Fraunhofer Society, NNMI will build a
network of regional research and development institutes for manufacturing innovation (IMIs). Each IMI will focus on
a unique research area and serve as a hub
for manufacturing innovation. Research
and development by the hubs and their
partners will boost U.S. technologies
and products, hopefully spurring more
domestic high-tech manufacturing and
employment.
One of the frst and among the
most promising from the perspective of
sustainability is the Digital Manufac-
22
By Nabil Nasr
turing and Design Innovation Institute
(DMDII). It is a consortium of more
than 100 companies, universities, nonprofts and government agencies that
support digital design and manufacturing. DMDII is based in Chicago.
More than 40 companies have signed
on to DMDII. Some partners are small
Each IMI will focus on

a unique research area and
serve as a hub for regional
manufacturing innovation.
businesses noted for creativity and innovation, but others are notable high-tech
manufacturers like Autodesk, Boeing,
Caterpillar, Cray, Lockheed Martin,
Microsoft, Rockwell Collins, RollsRoyce and Siemens. Likewise, the list of
universities and labs includes many with
strong manufacturing research and education credentials. For full disclosure, I
am serving on DMDIIs executive committee.
DMDII intends to help harness new
technology to address declines in manufacturing employment, lack of digital
manufacturing capability, breakdowns
in the manufacturing life cycle, and
communication barriers that inhibit the
exchange of data. Therefore, it is focusing on three crucial opportunities to
empower change: research and development by participant project teams, outreach to manufacturers to help compa-
nies assess and benchmark their digital

competencies, and workforce development to better prepare and align workers
for success in the digital manufacturing
arena. The results will apply to nearly
every manufacturing industry sector
and could decrease costs by roughly 10
percent across the entire manufacturing
enterprise.
To make this happen, DMDII is
committed to advanced manufacturing enterprise, an area that has
real sustainability enhancing implications. With advanced manufacturing enterprise, DMDII will
develop and implement modeling
and simulation tools to allow faster
time to market and effcient production
of complex systems. It also includes a focus on tools and practices to minimize
multiple designs, prototypes and test iterations typically required for product or
process qualifcation, all connected via
the digital thread to enable designer,
analyst, manufacturer and maintainer
collaboration.
Visit http://dmdii.uilabs.org to see what
DMDII has accomplished and what it
plans to do next. I think you will be impressed.
Nabil Nasr is director of the Golisa o I stitute for Sustai ability (GIS) at the Rochester
I stitute of Tech ology (RIT) a d director
of the Ce ter for I tegrated Ma ufacturi g
Studies, a tech logy developm t d tr sfer arm of GIS. He fou ded the Natio al
Ce ter for Rema ufacturi g a d Resource
Recovery (NC3R) at RIT.
Prepare for the PE Exam with the ONLY

Industrial Engineering PE Review Class!
Feel confident sitting for the Principles of Engineering exam by taking the IIE PE Exam
Review for Industrial Engineers class this February. Review the key concepts from the
industrial engineering body of knowledge and practice sample exam questions.
Where: Norcross, GA
Cost: $1,595 for members | $1,945 for non-members
Register at www.iietrainingcenter.org/PEI
ING
AIN1948 CE
LE AR
ER
NT
When: February 23-26, 2016
IIE T
R
This five-part review includes: Systems Definition, Analysis and Design; Facilities
Engineering and Planning; Supply Chain and Logistics; Work Design; and Quality
Engineering. Plus, participants receive an exclusive set of review notes only available to
class attendees!
N. IMPROVE. PROFIT.
The PE Exam Review course is also available as an online course.

23
HELPING
THE HELPERS
Adaptive tr
comm
yc
By Bublu Thakur-Weigold, Jonas Stumpf and Stephan Wagner
24
Hardly a month goes by without us witnessing scenes of disaster somewhere on

our planet. The images of Haiti, Syria and Nepal linger in our minds, only to be
displaced by the next reports of humanitarian need. In fact, historical data confrm that disasters, be they natural (earthquakes or storms) or man-made (war
and its aftermath), are increasing in frequency.
With every report, the urgent appeals resume. The media coverage implies
that the biggest challenge of aid is the fnancing itself. However, as we enter the holiday
season for many parts of the world, a glimpse into what happens with the many players
who spring to action and to the funds once they have been collected suggest another story.
In the years that we have been working with humanitarian operations, we have become
familiar with a hidden need behind the well-publicized emergencies the need behind
the need, so to speak. Contrary to the impression created by urgent campaigns like Give
now for Nepal! the accumulated donations are prodigious, having grown eightyfold from
under $2 billion in 2000 to reach $156 billion in 2014. Evidently, both private and public donors are giving generously, leaving many organizations, especially those run by the
United Nations, cash-rich.
Rather, the challenge is how well these budgets are deployed to meet benefciary need.
Research indicates that 80 percent of all funds are spent by logistics teams, mostly because
they are procuring the goods. Unfortunately, a staggering 40 percent of it does not fulfll
its purpose and goes to waste.
Contrary to another common perception, the reason for this loss is not corruption, incompetence or disproportionately high overheads. The waste can be traced to the same
kinds of ineffciencies that bedevil commercial supply chains: process dysfunction, silo behavior, redundant work and, especially, communication breakdown.
But unlike the most successful for-proft frms, humanitarian organizations do not usually attract top managers and supply chain experts, which puts the best practices of industry
out of their reach. This is an avoidable tragedy since, with every penny put to proper use in
getting the right goods to the right place at the right time, human suffering will be reduced.
At a 2012 workshop, the logistics director of the International Federation of Red Cross
and Red Crescent Societies urged her peers to professionalize and to do more with less,
reminding them that the needy benefciary cannot ask you to become more effcient so
you can reach his uncle in the next village. This was our call to action.
Pragmatic solutions from the ivory tower

The Swiss Federal Institute of Technology (ETH) and the Khne Foundations HELP
Logistics AG have been working together for the past four years to increase the effciency
of humanitarian logistics.
The Khne Foundation does not fund relief operations directly. However, it does provide grants for training and consulting through its HELP Logistics AG program. On the
other hand, despite its supply chain training prowess, the ETH is not one of the many universities that offer degree programs in humanitarian logistics. Nevertheless, the combined
capabilities of the two organizations seemed to be a dream team: academic prowess and
best practices, together with an extensive worldwide network and real-life experience in
humanitarian work.
Early on, however, it became clear that ETHs commercial research fndings did not
quite ft the logic of nonproft work. In many cases, the teaching material required a thorough translation before it could be delivered in the feld.
To enable agile delivery and impact, team-based action learning combined with a version of MIT professor Jay Forresters beer game proved to be a powerful way to drive the
necessary improvements. Communication breakdown is arguably the No. 1 cause of ineffciency in humanitarian operations, and our own assessments confrmed that poor information fow, not knowledge gaps, compromised effectiveness.
25
Helping the helpers

26
Comparable to the commercial sector, education does not

automatically trigger solution implementation. No single
boss can command a global network of decision-makers to
collaborate. There would have to be alignment and buy-in by
the individual functions. Our workshops would also have to
engage teams that were more heterogeneous than those who
attend universities and prepare these trainees to implement
change.
Although it is often misunderstood as a rehearsal of supply
chain management, the beer game is a simulation of a dysfunctional distribution system, and (if not passively witnessed on a
computer screen), one of the most compelling ways to experience how the distortion, delay and amplifcation of information causes systems to fail.
Humanitarian logistics have unique characteristics, like the
built-in handoffs between headquarters (in quiet places like
Switzerland) and the feld, which is by defnition in a disaster area with poor bandwidth and infrastructure. Firefghting, with its tendency to overreact and under-plan, is the very
nature of most aid work. It is not surprising that reporting
material count or benefciary estimates in the heat of battle is
perceived as irritating bureaucracy that distracts staff from the
real work at hand.
It was our challenge to convince learners from a daunting
range of professional and educational backgrounds that passing
on information was as critical to the success of their projects
as distributing the goods themselves. In this report from the
feld, we look back on what both our instructors and participants experienced over several years at several stations around
the globe.
Station 1, Geneva: The instructors were on our home
turf and felt confdent that we had all the answers. How diffcult could this be?
We played the original version of the beer game with a
small, faith-based nongovernmental organization (NGO) that
provides healthcare and sanitation to hot spots in the world.
The game worked just as it had with our MBA students and
commercial managers. The stock-outs and breakdown of trust
were appalling.
But in this case, the players were unconvinced by the incentive to minimize cost. Why, they asked, should a high score
(high costs) really lose? They were under pressure to spend all
their budgets or risk losing them in the next round of funding.
This time, the point of this organizations operations wasnt
about cost-cutting this was about saving lives.
Slightly chastened, the instructors went back to the drawing
board to redesign plans for the future.
Station 2, Rome: A visit to the U.N. offces does not
equate to a visit to your typical NGO. Instead, the instructors were dealing with a team of highly trained professionals
accustomed to all the infrastructure that international politics
can provide. This site is the nerve center of food distribution
Empty warehouses across the board mean hunger or worse for

the benefciaries of humanitarian assistance.
to the worlds needy. The staff members carried blue passports

and came, proverbially, from all corners of the earth.
By now, we had redesigned MITs original beer game to
become the high energy biscuit or HEB distribution game,
relabeling the nodes to refect humanitarian nomenclature and
reworking the debrief to address their practices. In this session,
there was no argument about the need to control cost.
However, a silence briefy descended on the players at the
one round in play when every single one of the four warehouses was empty. Stockpiles of biscuits were on the road between them. Without a word from the instructors, the consequences of the delays and demand distortion were all too clear
to the participants. The cost of a stock-out would be hunger or
worse. The lesson was learned.
Station 3, Juba, South Sudan: This time, training was
happening in a war zone. This capital city had only two perpendicular paved roads, and some parking lots were guarded
by small boys with large rifes. Our host NGO was to provide healthcare and water sanitation because epidemics spread
when food is distributed to camps without latrines.
The international staff members were mostly universityeducated, and one turned out to be a mathematician. In contrast, the national staff members in the classroom were those
who had survived the confict that had destroyed settlements
and prevented them from completing primary school. Dur-
ASEAN government offcials play the high energy biscuit (HEB) game, a version of MITs beer game modifed to ft the needs of
humanitarian operations.
Revising ways to help refugees

As stories of refugees from Syria and other war-torn and impoverished Middle Eastern countries dominate the news, humanitarian relief
agencies are examining their supply chain networks to improve their response.
CHEP, which provides pallets and container pooling, is performing an in-depth study of the United Nations Commission on Human
Rights global supply chain network, according to the 21st Century Supply Chain Blog run by Kinaxis, which provides cloud-based
supply chain software. In addition, according to Kinaxis, the UPS Foundation has worked with UNCHR and the World Food Programme
to deliver relief shipments to thousands of refugees in Greece and along the Macedonian border. On September 5, 163 metric tons of
relief supplies arrived via air shipments, including 86 tons of high energy biscuits, 30,000 blankets, 25,000 sleeping mats and 1,000
solar lanterns.
Mobile technology is having its
own effect on humanitarian supply
chains. Refugees Welcome, a German
website, has connected 222 refugees
with accommodations from people
who have opened their homes to
displaced people. And Refugeemaps.
org is an independent project that uses
geospatial and technical knowledge to
help support the humanitarian network
for the refugee crisis in Europe.
According to its website, the crowdsourced online application addresses
the need for the visual display of
grassroots activity.
27
Helping the helpers

28
ing the class, not all of them could compute

the simple sums to tally inventory, nor fll out
the graphs. Our teaching skills were tuned to
the needs of professional managers and had
reached their practical limits.
Astonishingly, we saw that innumeracy
doesnt have to be a barrier to ingenuity
and improvisation. The teams understood
the wild fuctuations of the graphs and welcomed the recommendation that they set up
their game to win rather than to lose. The
gaming behavior triggered by poor visibility
and communication was all too familiar to
them. It was a relief to learn that hostile and
distrustful reactions could be the outcome of
system structure (which could be fxed) and The graphs show that one well-performing calm manager (at node 1) cant prevent
not always malice (which could not be fxed). the interlocking system from going out of whack.
One of the local buyers made a neat diagnosis: They always suspect that I am corrupt, and they unStation 5, Brussels: This time, the training team was at a
fairly correct my cash budgets, which, in turn causes me to worldwide meeting of logistics offcers of the same small faithover-request so I can buy the food we need. And then every based NGO we began our program with. After playing the
number is wrong no matter what! Again, lesson learned.
HEB game so many times, they now understand the value
Station 4, Jakarta, Indonesia: Indonesia is in the heart of managing the system and not just the sum of its functions.
of the typhoon and earthquake region. The news of our HEB
One of our students, a former child soldier from South Susimulation has spread, and the Future Leaders of Disaster dan, offered to assist with the game because he had learned so
Management of the ASEAN governments wanted to try it.
much from it. He wanted to make sure the learning was passed
To increase preparedness, the government offcials are plan- on to his colleagues in Afghanistan, the Middle East and Asia.
ning centralized warehouses in a region where tropical storms This was the last person we had expected to train the trainer,
hit with near predictable regularity. The game electrifed the but he got the job done. And so, apparently, had we.
room, and indignant cries of corruption rose to a chorus.
The delays and handoffs were wreaking their usual havoc: Looking back, looking forward
Players were struggling with a system that took 12 weeks for The special challenges of humanitarian logistics clearly are
information to fow end-to-end, from benefciary to factory being taken seriously by educators and researchers alike. The
and back. Where is the material I ordered weeks ago, and growing array of training options for humanitarian workers is
what have you done with it?
creative and diverse, ranging from e-learning from the Fritz
Although there were only four nodes to manage, supplies Institute to applied academic programs like those at Georgia
were running out at one node and piling up at another.
Tech and the University of Lugano. The ETH/HELP LogisIn this group, however, we had decision-makers who could tics AG program completes its portfolio of training with aginfuence how the international network was to be confg- ile workshops designed to improve continuously to support
ured. We discussed how long their own supply chains were set team-based action.
up to be, how many handoffs were built in from the beginning
Our experiences often counter the assumption that the keys
and how demand information was being passed through the to a successful class lay with the quality of instructor or trainsystem.
ing content. Instead, we make several observations on instrucThey asked themselves how better collaboration between tional design. Action learning has proven useful, especially to
countries could be set up to avoid misunderstandings and mu- mixed audiences, because it makes the educational qualifcatual distrust. Dont communicate by passing only reports, lists tion of individual participants irrelevant. Furthermore, we
or orders back and forth without explanation and context. The found that subjecting the group to well-orchestrated simulamisinterpretation of upstream performance and the gaming tions of their own system (and not just rehearsals of individual
that ensued (again, over-ordering because they did not believe transactions or scenarios) builds both judgment and shared
that their supplier would deliver as requested) would have to incentives.
stop. They left determined to put their pooled resources to
To create a common understanding of a complicated global
better use.
system, the selection of class participants must be prioritized as
FIGURE 1
Complexity in action
A humanitarian supply chain is an interconnected system involving fows of goods, funds and data.
highly as training content. It was often a struggle to convince

fnance or programs staff to attend what is announced as a
logistics class.
Yet if the learning group consists of only one function, the
training might reinforce the silo thinking that is natural in
specialized organizations, while implying that a single function (like logistics, which has the last touch) is in control and
alone accountable for delivery performance.
As practitioners and academics, we have worked with the
U.N. and tiny NGOs, government representatives, wellmeaning and diligent people from all walks of life and privilege. What surprised us most was the fact that skill gaps are not
the most pressing problem they faced. Without well-managed
information fows, their organization, natural turnover and
process architecture can set up humanitarian logistics to fail.
Collaboration is not a natural impulse in a high stress situation, be it commercial or humanitarian. Communicating
with order slips or reports is, however, natural and routine
behavior. Changing this one habit has a disproportionate impact on system performance by reducing handoffs and delays.
After almost two decades of playing the beer, and now the
HEB, game, one of the most frequent comments we hear is
how amazing it is that the bullwhip effect always kicks in
with every single group of players, independent of individual talent, disposition, education, experience or managerial
expertise.
Underlying this wonder is one of the most persistent false
assumptions about systems performance: that a charismatic
leader will make it work, neutralizing uncertainty. The shocking experience of the game, in which professors and CEOs
alike fail, makes clear that no one decision-maker, however

gifted, well-educated or strong-willed, can be held responsible
for or maintain control of what is going on across the board.
Uncertainty, if allowed to propagate unchecked, will always
exist and wreak havoc. Humanitarian supply chains, like their
commercial cousins, are, as shown in Figure 1, a complex, interconnected system involving fows of goods, funds and data.
Conversely, even the smallest link in the supply chain can
help to remove bottlenecks and expedite critical information
so others can plan realistically and in time. This single insight
can motivate all kinds of relief workers to reach out to a proliferating array of stakeholders: military and government representatives, local authorities, national and international suppliers, donors, partner organizations, clusters, transport partners
and, of course, their own headquarters.
Improved collaboration can leverage the multibilliondollar budgets in what some are calling the humanitarian
economy. No reduction of overhead or even corruption can
compare. Its about unlocking the potential that was in their
systems all along.
Bublu Thakur-Weigold is associate director, programs at the Swiss
Federal I stitute of Tech ology Zurich (ETH Zurich) a d is a part er
at e3 Associates ter atio al.
Jo as Stumpf is regio al program director for the Kh e Fou datio s
HELP Logistics AG, based i i gapore.
Professor Stepha
ag er holds the Chair of Logistics Ma ageme t
at the Swiss Federal I stitute of Tech ology Zurich (ETH Zurich).
29
A primer for
simulation in
manufacturing
IE methods,
is a powerful tool for
improv
By Edward J. Williams
30
As the world becomes conceptually smaller and more

tightly integrated economically, the challenges of designing, staffng, equipping
and operating a manufacturing process or plant intensify. These
challenges include, but are not limited
to, process design and confguration,
selection of personnel (staffng levels
and skill levels), selection of machines,
sizing and placement of buffers, production scheduling, capacity planning,
implementation of material handling
and choices for ongoing process revision
and improvement.
During its 50-year history of application to manufacturing operations,
simulation has successfully addressed
all of these and more. Correctly used,
simulation is a powerful force for organizational learning. Typical motivations
for initiating a manufacturing-context
simulation project include:
0. A system design is already anointed, but upper management wont
OK spending the money until a
simulation provides good news about
that design.
1. A design (or several) is (are) sketched
on a cocktail napkin, and simulation might give insight about the
designs potential capability and
indicate points amenable to improvement.
2. The system is already operational,
but not satisfactorily; several improvements have been suggested
and even hotly debated. Their merits
need investigation.
3. The system is already operational,
and contingency plans are needed in
case of increased product demand,
increased economic pressures, wider
variety of product mix or other plausible changes.
Note how a zero prefaces the frst
motivation. Beginning a simulation
project with this motivation is setting
foot on the road to ruin as the results
31
A primer for simulation in manufacturing
will be irretrievably contaminated by bias. The second motivation (or No. 1 in our nomenclature) is the one with the
greatest potential return on investment (ROI). In these cases,
many examples exist of a 10-to-1 ROI, occasionally reaching
100-to-1 ROI.
In this situation, estimating all needed input data required
for the simulation will be a challenge. After all, the system
does not yet exist. The power of sensitivity analysis (explained
below) is then extremely valuable.
In the last two situations, the existing systems input data,
which will be modeled as a baseline, will be more readily
available, although the numbers might not be easy to fnd. Its
quite possible that suggested improvements A and B will be of
little value when implemented separately, but implementing
both could yield great value. Statistical analysis of the output
can expose such valuable insights.
When undertaking a simulation project in manufacturing,
remember that unsatisfactory operation may refer to any or
all of low throughput, low utilization of expensive resources,
excessive in-process inventory or long makespan (likely including long waits in queues).
Where are we going?

First, when a simulation project is to be started, vital questions
must be asked and answered.
1. Exactly what is to be modeled? For the frst question, and especially for an initial or early foray into simulation
usage (which management may be approaching charily), the
preferred answer is a small one. Extensive experience suggests
that an answer such as the milling department or the XYZ
line augurs much better for eventual success than an answer
such as the whole factory, or, worse, the whole factory plus
inbound and outbound shipments.
2. What questions shall the model and output analysis answer, and what decisions will be guided by those
answers? For the second question, example answers (these
answers are themselves questions) might be:
Raising and documenting these questions accomplishes several vital tasks. First, these questions will provide an unequivocal basis for answering the fnal question, Has the simulation
project successfully met its objectives? Second, the questions
guide decisions concerning the scope and the level of detail to
be incorporated into the model, guide data collection efforts
and help guide the choice of simulation software. Note that
the scope and level of detail should be as low as possible consistent with answering the chosen questions.
3. When are results needed? There are two typical answers to this third question. The frst answer is that results
must be available by a specifed drop-dead date to infuence
a key decision. If late, the results will be useless and ignored.
The second answer is that the sooner results are available, the
sooner the company can start earning greater profts via an
improved system. The second case favors quality over speed.
In cases with a drop-dead date, the project plan surely will
require modifcation. Possible modifcations include canceling
the project, reducing its scope and adding headcount to the
project at its inception. Adding headcount is quite dangerous,
being akin to the notion that if we need the baby in three
months, not nine, we will assign three women to produce it.
And adding headcount after the project is underway is even
more dangerous. Such a move likely will crash into the fgurative iceberg so aptly described by Frederick Brooks famous
article The Mythical Man-Month. He wrote, Adding
headcount to a late project makes it later.
4. Who will do the work if it is to be done at all?
The manufacturing enterprise also will have to address the issue of who will do the work. Reasonable alternatives include
doing simulation modeling and analysis in-house or contracting with a service vendor for this and all future simulation
projects. Companies also can hire service vendors to complete
this project while also training your staff to do future projects.
Now, if the project is to proceed, its time for data collection.
Data collection and analysis

Of the three proposed alternatives for production line expansion, which one will produce the greatest throughput
per hour?
Will a specifc proposal for line design be able to produce
at least 55 jobs per hour?
What level of staffng of machine repairmen/repairwomen
will ensure that the total value of inline inventory will not
exceed $40,000 at any time during one month of scheduled production?
Will the utilization of a particular critical and expensive
piece of equipment be between 80 percent and 90 percent?
Which of several proposed designs, if any, will ensure that
no part waits more than eight minutes to enter the brazing
oven?
32
Data collection is notoriously the hardest and most tedious,

time-consuming and pitfall-prone phase of a simulation project. First, consider the wide variety of data typically needed for
a manufacturing simulation:
Cycle times of automatic or semiautomatic machines; process times on manually operated machines
Changeover times of machines, whether occasioned by
product change (Next one is green, not red.), cycle count
(After making 55th part, sharpen the drill bit.), working
time (After polishing for 210 minutes, replenish the abrasive.), or elapsed time (Its been three hours since we last
recharged the battery.)
Frequency and durations of downtimes; whether down-
times are predicted by operating time, total elapsed time

or number of operations undertaken; whether a downtime
ruins or damages a work item in process
Travel time, load time, unload time, routes and availability of material handling equipment (conveyors, tug trains,
automatic guided vehicles, forklifts); whether travel time
differs for loaded vs. unloaded vehicles; accelerations and
decelerations also may be signifcant and important
Frequency of defective product; whether the defective
product is scrapped or repaired
Operating schedule number of shifts run, their durations
Workers their schedule, number and type of workers
available (operators, repair staff, material handlers), duties,
travel time between duties, absenteeism statistics
Buffer locations and capacities
Availability and frequency of delivery of raw materials
But for the simulations purpose, the milling machine actually suffered one downtime of 40 minutes.
Forewarned by these examples (all from experience), the
reader and practitioner will be alert to others. Downtime data

are particularly diffcult to gather. Too often, production personnel are reluctant to report downtimes, perhaps fearing that
such reports would cast aspersions on the rigor with which
maintenance policies and procedures are followed.
For instance, a 30-minute downtime might need to be subdivided as (a) the machine was down fve minutes before the
malfunction was noticed and reported, (b) it took the repair
expert 10 minutes to gather tools and get to the machine, and

(c) it then took her 15 minutes to make the repair. Neglecting
(a) overestimates the demands on the repair staff.
Next, the input data must be analyzed for best inclusion in
the model. For ease of checking and updating the data, practitioners routinely urge that constant values be kept in spreadI have yet to undertake a manufacturing-simulation project sheets and imported into the model (all modern simulation
in which the client added nothing to the above generic list.
software enables this task), not hard-coded in the model.
Next, be careful of misunderstandings that can undermine
When data is imported into the model in this way, the
data collection. Perhaps the client spokesperson said, Cycle numbers can be updated without updating the model itself.
time of this machine is six minutes. Its quite possible that the Eliminating this task eliminates the errors introduced by the
operator is needed for 45 seconds to load the machine, which overconfdence of I dont know this simulation software
then runs automatically for 4 minutes. Then the operator very well, but it cant be that hard to open the model and just
is needed for 45 seconds to unload the machine. During the change a cycle time.
4-minute run time, the operaFurthermore, the modeler or
tor can travel to/from and peranalyst must decide whether to
form other tasks.
use the data directly (i.e., samFor more simulation
Indeed, in such cases, the term
ple from an empirical distribuThis
article
is
adapted
from
Pitfalls
in
Managing
cycle time has no one standard
tion formed by the data points
a Simulation Project, a paper by Edward J.
defnition.
collected) or ft a closed-form
Williams and Onur M. lgen that appeared in
Or the person collecting data
distribution (e.g., exponential,
reported that the employees
gamma, Erlang, Weibull and
the Proceedings of the 2012 Winter Simulation
work from 8 a.m. to 5 p.m., with
others) to the data using readConference. The Winter Simulation Conference
15-minute breaks starting at 10
ily available software and sample
2015, sponsored by IIE, is scheduled for Dec. 6-9
a.m. and 2:30 p.m. and a halffrom this distribution. The latter
in Huntington Beach, California.
hour lunch at noon. The correct
approach has two signifcant adprocedure might include the fact
vantages: It realistically permits
that workers spend 10 minutes from 8 a.m. to 8:10 a.m. don- sampling values in the simulation that are outside the range of
ning protective clothing and equipment, which they take off actual data points collected, and it eases the drawing of conclufrom 4:50 p.m. to 5 p.m.
sions concerning the model and its results, since formulas are
And if production reports that the drill press was down for readily available for common closed-form distributions.
a whole hour from 9:20 a.m. to 10:20 a.m., double check.
However, realizing these advantages is contingent upon
It could be that the drill press was in working order but idle fnding a closed-form distribution that fts the empirical data
because a problem upstream prevented any work from reach- well. And this may be impossible, particularly if the empirical
ing it.
distribution is conspicuously bimodal or multimodal.
Overlapping data from different shifts also can cause probIn that case, re-examine the data. For example, the data set
lems. Take, for example, when the person assigned to collect cycle times of the lathe might really be two data sets: cycle
data during the 4 p.m. to midnight shift reported the milling time of the lathe on x-type parts and cycle time of the lathe
machine was down for 20 minutes beginning at 11:40 p.m. on y-type parts. In such a case, subdivide the data set and
The next shifts data collector reported that the milling ma- reanalyze each subset. Extensive detail on distribution-ftting
chine suffered 20 minutes of downtime ending at 12:20 a.m.
analyses and techniques is available in the literature.
33
A primer for simulation in manufacturing
Furthermore, looking ahead to the next step, data should

be used in the model under construction as it is collected. The
sooner the data actually enters a model (even one in early stages of development), the sooner signifcant errors in the data or
misunderstandings involving its collection will come to light.
Building, verifying and validating the model

The task of building the simulation model now waxes large.
Indeed, in actual practice, the simulation team should collect
data and build the model concurrently. The choice of software
to build the model may be clear if previous simulation projects
have been done using that software; if not, then various considerations might direct the choice of software.
Some might want to use package x because its salesperson
gave the fashiest demonstration and made the rosiest promises. This, of course, is defnitely contraindicated.
Better reasons to use package x include the fact that one or
more employees have received instruction in its use, perhaps in
a university course. Or one of your analysts attended a conference and did a detailed comparative examination of competing packages relative to your modeling needs. Or a trusted
consultant, one who has no vested interest in recommending x, clearly articulated substantive reasons for choosing that
package. Or you could choose package x because of assurance
that support (including both software documentation and vendor support) will be timely and of high quality.
The analyst choosing the software must ensure that it accommodates any modeling needs specifc to the system to be
modeled. Examples include:
Shifts of work, perhaps including situations where parts of
the facility run one shift and other parts run two, very likely including situations involving coffee breaks and/or meal
breaks
Changeover times, perhaps including situations where
more than one cause of changeover (as discussed in data
collection) exist
Downtimes whose occurrence is based on any or all of
elapsed working time, elapsed total time or number of cycles executed
Repair operations whose undertaking may be contingent
on the availability of repair staff with specialized skills or
the availability of specifc repair equipment
Bridge cranes, perhaps including multiple cranes and
bump-away priorities in one bay
Conveyors, accumulating or nonaccumulating, perhaps including confgurations in which the conveyors have curves
that reduce travel speed
Material handling operations, including equipment such as
tug trains, forklifts, high-lows or automatic guided vehicles
Situations in which several parts are joined together permanently in a subassembly
34
Situations where the expected remaining cycle time suddenly changes because a piece of equipment suddenly becomes unavailable; for example, two polishers working together need a half-hour more to complete a job; one breaks
down, and the estimated remaining cycle time suddenly
becomes one hour
Situations where parts are inspected, with some being
judged good (ready for shipment or use in an assembly),
some being judged needing rework, after which they may
become good, and some being judged scrap to be rejected
Situations when several parts are joined temporarily; for example, to travel together on a pallet
Situations where several parts are joined permanently for
shipment or for further assembly work
Situations that disassemble raw material or parts
The task of verifcation should be concurrent with the task
of building the model. Verifcation, conceptually equivalent
with debugging in the context of computer software coding, seeks to fnd and extirpate all errors in the model by testing the model. As Glenford J. Myers clearly stated in The Art
of Software Testi g, a successful test is one that exposes an error.
The analyst should not build the entire model and then
begin verifcation errors in the model are then diffcult to
expose and isolate for correction. Instead, the analyst should
build the model piecemeal, pausing to verify each component
(e.g., another segment of the production line). Verifcation
methods include stepwise examination of the animation (Are
items in the model going where they should?) and code or
model walkthroughs (The model builder explains the construction and operation of the model to a willing listener, often
becoming aware of an error in doing so.).
Validation is fundamentally distinct from verifcation. Verifcation answers the question: Is the model built right? But
validation answers the question: Is the right model built?
The right model is one that accurately mirrors the real or proposed system in all ways important to the client and does so as
simply as possible.
Therefore, validation requires the participation of the client more than verifcation does. Powerful methods of validation include a Turing test between the model and the current
system, temporarily removing randomness from the model,
tracking one entity or item through the model, and checking
for face validity (e.g., a long average queue directly upstream
from a machine with low utilization probably indicates an error).
The ultimate goal of verifcation and validation is model
credibility. A credible model is one the client trusts to guide
managerial decision-making.
Model execution and output analysis

After verifying, validating and making sure the model is credible in the opinion of the client, it must be executed to evaluate
and assess the merits of the system designs under investigation. Next, the analyst must decide appropriate warmup time,
length of replications and number of replications.
Warmup refers to the simulated time during which the
model runs to achieve typical system conditions, as opposed
to the time-zero empty and idle default condition of the
model. To select the warmup time, the analyst frst must decide whether the simulation is terminating or steady-state.
A terminating simulation models a system that itself begins
empty and idle, such as a bank. A steady-state simulation models a system that does not periodically empty and shut down,
such as a hospital emergency room or a telephone exchange.
Most manufacturing systems are steady-state even if operations pause over the weekend, for example, work probably
resumes Monday morning where it left off Friday afternoon.
Whereas terminating systems need and should have zero
warmup time, a manufacturing system model must run for
suffcient warmup time to reach typical long-term conditions
before the simulation software is instructed to begin gathering
output statistics and performance metrics.
The length of a replication (i.e., the simulated time it represents) is likewise a delicate statistical question. The longer
each replication is, the more confdence both the analyst and
the client can have that the replication will capture representative reality in the system being modeled. One useful rule of
thumb is that even the rarest of events (for example, a conveyor
breakdown) should have a chance to happen half a dozen times
during the replication.
The analyst would do well to remember that the rarest
events may be interactions. For example, if each of two particular machines fails occasionally and independently, both machines may, on occasion, be down simultaneously. Despite
the rarity of this occurrence, its important to have information on system performance during such situations. In addi-
tion, the length of a replication should be an integer multiple

of a canonical work period (e.g., eight or 24 hours).
From a statistical viewpoint, each replication represents another experimental data point, or another throw of the dice
using different random numbers generated by the simulation
software. Therefore, successive replications are statistically
independent. This lets the analyst use standard statistical formulas (for example, those pertaining to the students t-distribution) to calculate confdence intervals for the performance
metrics of interest.
These formulas provide confdence intervals whose width
varies inversely as the square root of n (where n equals the
number of replications). If the width of these intervals needs to
be halved to give the client enough confdence to make decisions based on the simulation analysis, the number of replications must be quadrupled.
Furthermore, the analyst must avoid the mistake of making one extremely long run (for example, using the previous
numbers, 9,600 hours) and mentally dividing it into 20 replications of 480 hours each. Such misconstrued replications are
not statistically independent.
After all, conditions in the system at 955 hours (near the
end of one subdivision) and conditions at 965 hours (near the
beginning of the next) are very similar, the result of positive
correlation. With independence eliminated, the foundations
underpinning the computation of confdence intervals for the
performance metrics are compromised.
When just a few alternatives are to be compared (e.g., a, b
and c), the analyst can reasonably build confdence intervals
for all comparisons needed (here, a relative to b, a relative to c,
and b relative to c).
However, the typical simulation needs to do multiple conversions on multiple factors. For example, the analyst may
need to investigate whether to buy a faster conveyor and/or
add one more fork truck and/or add one more repairman or
repairwoman.
In this example, a faster conveyor by itself might yield almost no improvement, the same with adding an additional
fork truck. However, making both changes could prove to
yield signifcant improvement. For these situations, the many
designs (e.g., factorial) of design of experiments (DOE) offer
much greater statistical power.
Simulation studies following this path provide signifcant
economic advantages and speak well of the powers of industrial engineering analytical techniques.
Edward J. Williams works part-time at PMC, a dustrial e gi eeri g co sulta cy, a d teaches part-time at the U iversity of Michiga ,
Dearbor i cludi g a masters level simulatio ourse). He bega pplyi g simulatio
he late 1970s at Ford Motor Co.
35
Capacity analysis for better

batch processing
produce hard data for executive

By John Preston
36
Industry and business design many operations around

batch processes. Batch operations process one type of
product at a time and then usually change the operation
to make a different type of product, much like how a
small bakery would mix only one type of dough at a time
in their mixer. The baker uses the single type of dough to
make a series of one product, such as chocolate chip cookies,
and then cleans the mixer and makes a different type of dough
for a different type of cookie.
The bakers decision on the type of equipment and the size
of the batch of cookies can have a profound impact on the
business. The baker doesnt want to make too many of one
type of cookie, as a lack of customer demand could leave those
cookies unsold and growing stale. The baker also doesnt want
to make too little dough and spend extra time cleaning out the
mixer between small batches.
Instead, the baker wants to make enough of each type of
cookie to satisfy customer demand and to optimize the bakerys time spent and ingredients used. These decisions are no
different than the decisions made by large batch manufacturers. However, large batch manufacturers often face more
complicated decisions because they have a greater variety of
processes and products, along with the added complexity of
having many decision-makers with different motivations.
This complexity of machines and minds can lead to dissatisfed customers and unhappy owners if the batch production
planning process is not practiced consistently throughout the
enterprise.
The risks of nonstandard capacity analysis

Batch businesses that do not effectively use standardized capacity analysis methods run the risk of not meeting their customers primary needs. These businesses can create overburdened operations that simply do not have enough capacity to
meet the needs of the customer without creating stress within
the workforce and overloading its machinery.
Operations that are overburdened for long periods of time
naturally have increased operating costs and extra shipping
costs and produce more defective product.
Companies that do not effectively plan their batch processes
will not use their operations well, with leaders misallocating
capital for operations and machinery that are not necessarily
needed. Expensive machinery can be found idle, not making any product, even though decision-makers thought these
machines were to be well-utilized to meet customer demand.
Valued capital is not aimed at fxing an operations bottleneck
because machine utilization wasnt predicted accurately.
Without effective capacity analysis methods, batch businesses will give potential customers inaccurate quotes on new
products. This, in turn, can erode proft margins with unexpected operating costs.
Batch businesses that dont understand the capacity of their
operations often set inaccurate performance targets. These

critical targets are used to drive continuous improvement in
the operations. If the performance targets are too high, the incentive for the workforce is unreachable and has less meaning.
When the performance targets are too low, critical resources
are misappropriated away from the greatest improvement opportunities.
Understand critical processes

with work measurement
Before creating a capacity model for a batch operation, a vivid
understanding of the process must be established. To develop
this understanding, an organization likely will need to perform work measurement studies on the machines and people
affecting the batch process. These studies will generate the
critical assumptions used in the capacity model.
To develop a capacity model that accurately represents a
batch process, it is important to collect key data, including
cycle time, planned downtime, unplanned downtime and expected scrap.
Cycle time is defned as the time it takes to process one
unit of a product by an operation step without any downtime.
Planned downtime is composed of all of the predicted reasons
that the operation step would need to stop while making the
product and the frequency (time or usage based) of the stop
occurrence.
Unplanned downtime ratio is the assumed percentage of
time that the operation step would stop unexpectedly while
it is producing product. Scrap ratio is the predicted percentage
of product that fails to meet the customers guidelines and cannot be reworked to meet the customers criteria. Unplanned
downtime and scrap ratio assumptions can be developed over
an average period of time that represents historical or expected
ineffciency rates.
Two main types of planned downtime are collected when
developing a batch process capacity model: time based and usage (piece) based.
Time-based downtime is planned on intervals of time
passed. For example, an operation is designed to be idle for a
30-minute lunch over an eight-hour shift. Other examples include quality checks, shift change delays, daily housekeeping,
preventive maintenance and shift meetings.
To collect time-based planned downtime, frst determine
the work pattern duration of a shift, including breaks and
lunch. Next, determine the length of each downtime event.
Finally, determine the frequency (over time) of each downtime event.
Usage (piece) based downtime is planned on intervals of
units produced. Common examples include raw material restocking time, changeover (setup) time, container/packaging
replacement time and tooling rework or change time.
To collect usage-based planned downtime, frst collect in-
37
Capacity analysis for better batch processing

38
formation on raw material stock quantities. Next, determine

the length and frequency of each downtime event. It also is
important to determine the size of the batch and the forecasted
customer demand over time. These key assumptions will be
used to populate the capacity model that is used to represent
your processes utilization.
Create a model that will

be embraced by the organization
A business that uses batch processes can manage its operations
better by developing simple mathematical models that accurately predict an operations throughput (the product produced
per unit of time, e.g., pieces per hour).
These models must be designed in a way that is easily understood. Complex models with unclear assumptions are less
effective than transparent models. The simpler they are, the
greater the chance that the models will be adopted across the
business unit. The models will create the standard for planning
the business capacity.
For standards to be effective, they must be understood and
followed. Microsoft Excel is a great platform for creating
models. It can be understood easily without needing to embed
complex logic. Managers, salespeople and engineers often are
familiar with the software, so they are comfortable with its
basic functionality. This makes it a great choice for a capacity
models platform.
Within the model, create a worksheet that, in a preface, describes how the model functions. In the preface worksheet, list
all of the assumptions and defnitions of items/terms found in
the model.
Create a simplifed area of the worksheet where data can be
entered. This will reduce entry errors and help users understand all of the necessary inputs required to make the model
accurate. Allow only raw data to be entered into the spreadsheet, such as the time units, pieces in a batch and customer
demand volumes. This will reduce the chance of calculation
error and help users clearly understand the important factors
that affect the model.
Try to simplify the calculations as much as possible to allow future users to understand the functionality of the model.
Limit the use of complex formulas and stick to basic elementary math as much as possible. Use colors, wide margins and
bold fonts to organize data into types and highlight important
items. Doing this will allow users to understand the model
more quickly.
Finally, try to consolidate the model so that it can be printed
on a single sheet of paper and still be legible. This may require
formatting to condense the most important information. Having only the most important information on just one piece of
paper will allow users and decision-makers to understand the
models key inputs and outputs, making it easier to review in
meetings.
Establishing standardized capacity analysis

Once data has been collected and populated into the capacity
model that accurately represents the way batch production machines and people function together, the next step is proving
the models worth.
Test the model by modifying it to represent your organizations largest constraint process, the most expensive asset or
the process with the greatest variable operating costs. Try to
choose a process that receives executive-level attention. Compare the models forecasted effciency to the historically performed or predicted future performance. The new capacity
model could highlight these differences in performance expectations.
The model may show that the current planning assumptions
are too aggressive and that the operation doesnt have enough
capacity to meet the customers needs. The model may show
that enough capacity does exist if the process changes operations to reduce process cycle time, downtime or scrap. Alleviating a bottleneck operation that carries heavy operating costs
or helping to avoid a large capital investment most likely will
get the attention of key decision-makers.
If you want the new capacity model to succeed, it is important to document its positive impact on the business. Make
sure that you share the results with the key stakeholders. Leaders will favor using the model if it improves the accuracy of
their judgments when they have to make important resource
and capital decisions. A good capacity model can reduce the
stress executives incur when they have to make qualitative decisions without hard data.
After documenting the models effectiveness, train infuential managers on the value of using the model. Inform the people who are responsible for quoting and planning new business
on how to use the model. Teach those who schedule and manage the operations about how the model could beneft them.
Share the model with fnancial controllers and explain the
importance of the causal factors that have the most signifcant
outcomes on the effectiveness of the process. Show them how
changing these factors, such as batch size and changeover time,
can have a dramatic impact on capacity. Educating these stakeholders on how a batch operation performs will have a lasting
positive impact on the business and will greatly improve the
chances that leaders use the model when making capital investment decisions.
Once the model has been accepted by leadership, share it
with the entire organization by scheduling a meeting to disseminate the models worth quickly. In the meeting, introduce
participants to the basics of capacity planning. Describe how
they could use the model to improve their business. Teach
how to collect the data that will populate the model. Guide
them on how to interpret the models results. Walk through
specifc circumstances in which they could use the model.
Show how the model can be used to justify a capital invest-
Smaller batches for producing

biopharmaceuticals
GE Life Sciences hopes to become a leader in single-use production
systems that could help transition the biopharma industry away from big
batch production to more fexible systems, The Boston Globe reported.
The division of General Electric Co. has a roster of products that include
prefabricated factories that would cost between $60 million and $80 million
each, far less than the $300 million to $800 million it costs to build a
traditional bioprocessing plant, according to the newspaper.
Such modular production sites could work well in developing countries,
and one already is planned for China. The factories make drugs in smaller
batches and have single-use production systems, support equipment and
analytics software.
The division also is examining cell and gene therapy, felds so new that
production processes are still being developed. GE Life Sciences President
Kieran Murphy said the company is taking a total systems approach to such
development.
39
Capacity analysis for better batch processing

40
ment that would reduce downtime on

a machine.
Present an example of using the
model to reduce downtime due to
waiting on material resupply. Material feeding equipment can often create usage-based planned downtime
ineffciencies. This equipment is flled
with the raw material used to make
the product. As the raw material runs
out, sometimes the machines need to
be turned off to add new material,
causing a downtime event. Processes
or equipment often can be redesigned
to reduce or eliminate this planned
downtime.
Often, a capital investment is required to increase the size of the material storage or to purchase feeding
equipment that allows a queue of material to form. This queue
creates buffer time in the material feed system that allows the
operation to refll the material feeding equipment without
turning off the value-adding machine or process. The capacity
model can be used to calculate the operation hours saved when
making such an investment. This savings can be translated into
capital avoidance or an operating cost savings.
Show the team an example of how both reducing changeover time and batch sizes can reduce inventory costs without
adding extra operating time. Use the model to justify investing in quick changeover equipment and effcient processes that
reduce the amount of downtime associated with changing the
operation from running one product to another.
Businesses often invest time or money to improve changeovers in an attempt to reduce batch sizes. Smaller batch sizes
lead to less overproduction and inventory costs. A well-understood capacity model can help executives comprehend how an
operation can reduce batch sizes without increasing operating
costs. These efforts to reduce batch sizes can free cash held as
inventory that can be used to invest in the business to better
serve the customer.
Operations that have not focused their efforts on reducing
changeover time often can reduce their downtime signifcantly without spending any capital on new equipment. These
operations can achieve signifcant reductions in changeover
downtime by just focusing on prestaging needed items before
the changeover begins, performing basic preventive maintenance and coordinating resources better.
Simplifed capacity models on batch processes will quantify
how an operation that reduces its changeover time in half can
dramatically cut its inventory levels without increasing operating costs.
Once decision-makers realize the value of the model, the
organization will need to standardize how the model will be

used throughout the business. In most batch process centered
operations, the model could be used for quoting new business,
guiding capital investment decisions, scheduling changeovers,
determining batch sizes and setting operations performance
targets. Leadership must decide where to embed the model
into its decision-making processes and to develop process confrmation steps to validate its use.
Many processes in businesses and industries make their products in batches. Organizations that develop and use standardized methods of predicting a batch process overall equipment
or effectiveness will have better targeted strategic planning.
Critical resources and capital will be used more effectively to
reduce operating costs and to improve delivery performance.
The organizations continuous improvement process can
be improved with more precise quantitative methods of justifying projects that reduce wastes such as inventory, waiting,
overproduction, overprocessing and defects. Standardizing the
capacity planning process will align the sales, fnance, purchasing, scheduling and production management departments
to each other and the goals of the greater organization.
An organization simply sharing the common knowledge
and control of a process capabilities and goals can have a dramatic impact on the corporations proftability and customer
satisfaction.
Joh resto s a lea
a ufacturi g ma ager at MacLea -Fogg
Compo e t Solutio s a d a se ior member of IIEs Greater Detroit
Chapter. Outside of the automotive sector, he has served i the semico ductor, co sumer goods, weapo ystems a d biotech i dustries. He
graduated from the U iversity of Michiga ith a B.S.E. i dustrial
a d operatio s e gi eeri g, a d he received a BA from Northwood
iversity.
Effcient knowledge transfer

Compiled by Michael Hughes
Northrop Frye, one of the most infuential literary critics and theorists
of the 20th century, posited that the most technologically effcient machine that man has ever invented is the book.
In the age of the Kindle, the Nook and tablet computers, some may
doubt that quotes current validity. But those devices are simply a new
format for the venerable and honored method of transferring knowledge the book. (And unlike more modern contraptions, a printed book
doesnt need an AC outlet or a battery.)
So in honor of how much continuous process improvement types love
technology and effciency, weve compiled a list of some of the best books
of 2015. Enjoy and learn.
41

603999-DEC 2015 - Selected-Pages PDF

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

603999-DEC 2015 - Selected-Pages PDF

Uploaded by

Copyright:

Available Formats

Industrial Engineer

Engineering and management systems at work

How simulation leads

VOLUME 47 : NUMBER 12: $16.50

Make better decisions

Expand Your Abilities

Process Improvement Courses

Management and Specialty Courses

Six Sigma Green Belt

Introduction to Engineering Analytics

March 21-24, 2016 | Orlando, FL

Engineering Decision Making and Risk

April 26-28, 2016 | Norcross, GA

March 14-18, 2016 (Session 3)

February 1-5, 2016 (Session 1)

April 4-8, 2016 (Session 3)

Healthcare and Ergonomics Courses

April 4-8, 2016 (Session 3)

PE Exam Review for Industrial Engineers

March 14-18, 2016 (Session 2)

Facilities and Workplace Design

March 14-18, 2016 (Session 2)

December 2015 | Volume 47 | Number 12 | www.iienet.org/IEmagazine

24 | Improving humanitarian logistics

36 | Baking in better batch processing

41 | These pages transfer knowledge

Compiled by Michael Hughes

the Front line

DMDII tackles digital industrial innovation

12 | All I want for Christmas? A drone delivery

Industrial Engineer | www.iienet.org/IEmagazine

Lead Efforts to Improve Healthcare Quality Discuss

Visit www.IIETrainingCenter.org to view all upcoming courses and full descriptions.

Register now! www.shsconference.org

December 2015 | Industrial Engineer

Teaching the teachers

Industrial Engineer | www.iienet.org/IEmagazine

Transmitting specialized operational knowledge to the front lines can

Promote Achievement through

Manufacturing & Design (M&D) Outstanding Service Award

Quality Control & Reliability Engineering (QCRE) Golomski Award

Sustainable Development Teaching Award

Engineering Economy (EE) Undergraduate Senior Design Award

Manufacturing and Design (M&D) Division Student Paper Award

Process Industries (PID) Student Paper Award

December 2015 | Industrial Engineer

Web Managing Editor

serves the diverse audience of professionals and students whose

INSTITUTE OF INDUSTRIAL ENGINEERS

Senior VP-at-Large, Industry

Senior VP, International

James E. Moore II, Ph.D.,

Joan Wagner, P.E., Spirit

Bopaya Bidanda, Ph.D.,

Jennifer Cross, Ph.D.,

President-Elect and CFO

Senior VP, Regional Operations

Senior VP, Publications

Alice E. Smith, Ph.D., P.E.,

Immediate Past President

Chief Executive Offcer

Dennis Oates, Amazon