This article was downloaded by: [14.139.108.

34] On: 12 July 2017, At: 23:50

Publisher: Institute for Operations Research and the Management Sciences (INFORMS)
INFORMS is located in Maryland, USA

INFORMS Transactions on Education

Publication details, including instructions for authors and subscription information:
http://pubsonline.informs.org

Case—The MotoTech Manufacturing Company: Process

Control and Improvement
Prakash Mirchandani,

To cite this article:

Prakash Mirchandani, (2010) Case—The MotoTech Manufacturing Company: Process Control and Improvement. INFORMS
Transactions on Education 10(2):79-84. https://doi.org/10.1287/ited.1090.0041cs-a

Full terms and conditions of use: http://pubsonline.informs.org/page/terms-and-conditions

This article may be used only for the purposes of research, teaching, and/or private study. Commercial use
or systematic downloading (by robots or other automatic processes) is prohibited without explicit Publisher
approval, unless otherwise noted. For more information, contact [email protected].

The Publisher does not warrant or guarantee the article’s accuracy, completeness, merchantability, fitness
for a particular purpose, or non-infringement. Descriptions of, or references to, products or publications, or
inclusion of an advertisement in this article, neither constitutes nor implies a guarantee, endorsement, or
support of claims made of that product, publication, or service.

Copyright © 2010, INFORMS

Please scroll down for article—it is on subsequent pages

INFORMS is the largest professional society in the world for professionals in the fields of operations research, management
science, and analytics.
For more information on INFORMS, its publications, membership, or meetings visit http://www.informs.org
 Vol. 10, No. 2, January 2010, pp. 79–84
issn 1532-0545
10
1002
0079 informs ®

doi 10.1287/ited.1090.0041cs-a
I N F O R M S © 2010 INFORMS
Additional information, including supplemental material and rights and permission policies, is available at http://ite.pubs.informs.org.

Transactions on Education

Case
Downloaded from informs.org by [14.139.108.34] on 12 July 2017, at 23:50 . For personal use only, all rights reserved.

The MotoTech Manufacturing Company:

Process Control and Improvement
Prakash Mirchandani
Katz Graduate School of Business, University of Pittsburgh, 358 Mervis Hall, Pittsburgh, Pennsylvania 15260,
[email protected]

Introduction The Situation

You and four of your ex-classmates run a small, but MotoTech Manufacturing (MM) is a medium-sized
rapidly growing, management consulting firm that firm that produces integrated circuits for the
provides advice to high-technology companies. What electronics—particularly, the consumer electronics—
had started off as a half-baked idea late one night industry. Integrated circuits are used in comput-
when you were busy working on an MBA project six ers, automobiles, airplanes, appliances, toys, watches,
years ago has resulted in a highly successful consult- cell phones, cameras, televisions, stereo components,
ing firm today. Indeed, who would have dreamed that medical diagnostic equipment, etc. Integrated circuits
all of the fights and the arguments your group had allow electronic products to be smaller in size, have
had when doing school projects could possibly result greater functionality, and faster speed, yet they are
in such a closely-knit team just a few years later? You lower in cost. The electronics and computer industry
still challenge each other, still debate the pros and has grown steadily over the past several years and
cons of each issue, and are often frustrated with each now employs more people in the United States than
other, but the end-product is always much better as a the steel, automobile, and pharmaceutical manufac-
consequence. turing industries put together.1
Today is special. You are visiting a company called The demand for integrated circuits has also grown
MotoTech Manufacturing (MM) that was founded by sharply during the past decade, with newer appli-
another former batch mate. You were a bit surprised cations being identified every day, and this trend is
to get a phone call from the old friend requesting that expected to continue into the future. This increase in
one of you go over for a plant visit. That is how it
demand for integrated circuits has, in turn, resulted
typically works in your business: Just one of the five
in several highly profitable years for MM. MM is
visits the client along with some junior members of
proud to be a part of this growth in the United
the consulting staff. In this case though, because all of
States: Roughly 45% of the worldwide integrated cir-
you know the founder, and all five of you had wanted
cuit demand of $256 billion is met from domestic
to go.
production.2
You discuss the potential assignment during your
car ride to MM and conclude that you probably More recently, though, John Tagole, MM’s founder
would be required to make recommendations about and CEO, has been troubled by MM’s lower yield
MM’s manufacturing process. Almost all domestic
manufacturing firms have been under pressure to 1
http://www.bls.gov/oco/cg/indchar.htm. Checked on June 6,
improve their processes, and you speculate that you 2009. Latest available data on this Bureau of Labor Statistics website
are for 2006—before the 2008–2009 reorganization of the domestic
would probably be investigating how to improve
automotive industry.
manufacturing both in the short term and the long 2
http://www.sia-online.org/cs/industry_resources/industry_fact_
term. In doing so, you anticipate having to first deter-
sheet. The United States continues to be a market leader in this
mine whether MM’s process is capable of producing industry. One reason is that the rate of innovation in the technology
products that meet the customer requirements and, if for this industry is high, with annual research and development
not, to recommend ways of doing so. expenses touching 16% of sales.

79
 Mirchandani: Case: The MotoTech Manufacturing Company: Process Control and Improvement
80 INFORMS Transactions on Education 10(2), pp. 79–84, © 2010 INFORMS

(higher rejection rate) values, compared to those of its material for virtually all integrated circuits is silicon,
Additional information, including supplemental material and rights and permission policies, is available at http://ite.pubs.informs.org.

competitors. Moreover, overseas competitors, particu- which is the primary constituent of common sand.
larly from Taiwan, Japan, Singapore, Korea, Thailand, The circuits are manufactured from many “die” that
and now China, have started to encroach upon MM’s are fabricated on silicon wafers (in a facility referred
traditional customer base. It has been six years since to as a “fab”). The die are also known as chips. The
John Tagole started MM after getting his MBA from silicon wafer can be from about four to 10 inches in
the Great Western Pennsylvania Graduate School of diameter and may contain from a hundred to several
Business (GWPGSB), but the competitive pressures thousand chips. During the wafer fabrication process,
Downloaded from informs.org by [14.139.108.34] on 12 July 2017, at 23:50 . For personal use only, all rights reserved.

attributable to global competition have never been each wafer goes through several steps to produce
higher. While John Tagole knows that this type of the final chip. These steps include processes such as
competition is good for the consumer in the long plasma etching, ion implantation, chemical deposi-
run because it forces companies to continuously make tion, and photolithography. Once a wafer has been
productivity improvements, he also realizes that his processed and tested, it is cut into individual chips.
dreams of growing MM into a global chip supplier for The chips are then packaged for sale and used in the
the consumer electronics industry might evaporate in electronic systems of cars, computers, and many other
thin air if he makes a bad decision now in what is products.”
clearly a critical juncture in MM’s short history.
On this cool November morning, John Tagole
has a premonition of receiving some adverse news.
The End of the First Visit
John Tagole is clearly proud of his success as he fin-
Sure enough, he gets a call from Danika Katz,
ishes giving The Famous Five a tour of his plant.
VP of Procurement, Semicon International, one of
They too must be proud of his achievement (as he
MM’s largest customers. Semicon Transcontinental is
is of theirs), because all five of them have shown
a consumer electronics manufacturer. Danika first dis-
up for this fact-finding trip. They ask the right ques-
cusses the long and successful vendor-buyer rela-
tions, and John Tagole is impressed by the speed with
tionship that MM and Semicon have had for six
which they zero in on the problem. The “diffusion”
years. Then she informs MM that Semicon engineers
stage, The Famous Five conclude, is the problem with
have drastically tightened the design specifications
the current process. “Let’s address the problems at
for the components going into the new line of super-
the diffusion stage first, and then we will look at the
high-definition televisions with three-dimensional-
remaining stages of the process.” John Tagole mar-
like imagery. Danika informs John Tagole that
vels: “These people must have a sixth sense. Danika
although MM’s products meet Semicon’s current
Katz’s new specs all deal with the diffusion stage.” In
specification limits, they do not meet the newly
any event, he goes ahead and describes the diffusion
designed specification limits, and unless MM can
stage to The Famous Five.
meet the new specification, Semicon might have to
“In many process steps, the wafers are processed in
take its business elsewhere.
a batch, with several wafers processed together at the
After the call John Tagole wonders, “Should I just
same step. One such process is the diffusion step. As
ignore this call? Are these new specification lim-
many as two hundred wafers may be batch processed
its temporary—imposed by a young, overenthusias-
together in a diffusion furnace. During this step, a
tic design engineer?” “Maybe,” he speculates, “if we
layer of material (e.g., silicon dioxide) is deposited
wait long enough, Semicon will eventually be forced
onto each wafer. The thickness of the deposit must be
to relax the specifications and buy MM’s current prod-
tightly controlled at a specified target thickness. The
uct, especially if none of the other chip manufacturers
target thickness for this diffusion process is 3,000 Å.3
can meet the new specifications. On the other hand,
To ensure the desired functionality and reliability of
what if Semicon does not relax the specifications and
the chip, the deposit thickness must lie in the range
drops MM as its supplier? Can I really afford to lose
of 2,900 to 3,100 Å.”
such an important customer?”
The Famous Five go into a “huddle.” John Tagole
He makes a quick decision and decides to call “The
is used to their loud arguments from his time back
Famous Five”—his batch mates who are now running
at the GWPGSB—but his staff is not. He chuckles to
a consulting company. (As he places the call, why,
himself as he overhears worried staff members mut-
he tries to remember, were they called “The Famous
tering about calling plant security.
Five” at GWPGSB?)
After about an hour, The Famous Five emerge and
discuss their plan of action with John Tagole. They
The Process
John Tagole describes the manufacturing process to 3
An angstrom (Å) is a unit for measuring length and equals 10−10
The Famous Five during their visit: “The critical raw meters.
 Mirchandani: Case: The MotoTech Manufacturing Company: Process Control and Improvement
INFORMS Transactions on Education 10(2), pp. 79–84, © 2010 INFORMS 81

instruct John Tagole on what data to collect, and then room to 60 F. The Famous Five construct a histogram
Additional information, including supplemental material and rights and permission policies, is available at http://ite.pubs.informs.org.

say, “Play time, John. You owe us at least a lunch! of the “Post_Improvement” observations and con-
Let’s catch up on what’s been happening on the per- clude that the distribution is normal.4 Answer the fol-
sonal front.” lowing questions for the data used in Part C.
Part A. Should John Tagole have “waited out the 1. What is the cumulative probability for a single
storm,” hoping that Semicon will relax its new, more wafer to have a thickness of 3,000 Å?
stringent specifications? Would this strategy have 2. What is the percentage of defectives being pro-
worked in the short term? In the long term? duced under the current setup?
Downloaded from informs.org by [14.139.108.34] on 12 July 2017, at 23:50 . For personal use only, all rights reserved.

Part B. John Tagole provides your group, The 3. Compute the process capability index, Cp , for
Famous Five, with the data set in Table 1 and in the this situation. Is the process capable of meeting the
“Pre_Improvement” worksheet of the file “MotoTech customer’s requirements? Is Cp the appropriate met-
(PCI) Data.xls.” The data were taken from the diffu- ric for measuring process capability in this case? If so,
sion process. Samples of three wafers were randomly why? If not, why not? What other metric would you
selected from each batch of 200 wafers, and the thick- suggest?
ness of the silicon dioxide deposition was measured 4. Is this a six sigma process? If not, what is its
on the wafers. The data are from 72 batches. sigma level? Assuming that the process mean can
Answer the following questions for this data set. drift by, at most, 1.5 sigma in either direction of the
1. For this data set, construct the X-bar and range
target without being detected, compute the approx-
control charts. Attach a printout showing your X-bar
imate number of defectives out of a million. (Hint:
and R control charts.
In Motorola’s experience, process mean shifts of up
2. Use the range chart to determine whether the
to 1.5 sigma can go undetected, and so they recom-
process variation is in control.
mended the use of this assumption to calculate the
3. Use the X-bar chart to determine if the process
proportion of defectives for a given sigma level.)
mean is in control.
4. If the range chart is found to be out of control in Part F (Looking Ahead).
Part B2 above, would the control limits of the X-bar John Tagole is pondering over what to do next   
chart have been valid? Why or why not? the secret of his success has been making many small
Part C. After observing the control charts in Part B, process improvements continuously and making fun-
The Famous Five investigate the reasons for the iden- damental process reengineering changes as and when
tifiable causes of variation. They find that the pro- needed. The diffusion process has been brought under
cess went out of control during times when the plant control following the initial analysis done by The
air conditioning system was shut down for preven- Famous Five. However, the rejection rate is still high,
tive maintenance. They recommend that a back-up which leads to higher unit costs and cuts into MM’s
air conditioner be installed and that the tempera- profits. He has been advised by The Famous Five to
ture in the diffusion room be maintained at 60 F. do a “design of experiments”5 study. This study will
This recommended temperature setting is based on help in determining whether 60 F is the right tem-
The Famous Five’s general experience, although local perature, or whether this setting needs to be fine-
atmospheric conditions and raw-material composi- tuned. He believes that such experimentation is in line
tion can also potentially affect the recommended with the philosophy of continuous improvement—
temperature. Until the new air conditioner can be advocated by many quality gurus. This approach has
installed, The Famous Five recommend that the pre- made Toyota an automotive powerhouse worldwide,
ventive maintenance be carried out on weekend resulting in its overtaking General Motors in 2008
nights, when the diffusion process is stopped. They global sales,6 even before the reorganization of GM.
ask John Tagole to collect data for an additional Through this investigation, he hopes to identify the
72 batches under these controlled conditions. These right temperature setting, possibly also identifying
data are enclosed in Table 2 and in the worksheet any underlying interaction between temperature and
“Post_Improvement.” other input factors. This study is ongoing, and its
1. For this data set, construct the X-bar and range findings will be implemented in the next few weeks.
charts. Attach a printout showing your control charts.
2. Use the range chart to determine whether the 4
You should check this by constructing your own histogram using
process variation is in control.
Excel or by doing a check for normality, using a statistical package.
3. Use the X-bar chart to determine whether the 5
The case “The MotoTech Manufacturing Company: Design of
process mean is in control. Experiments/ANOVA” shows how to use Design of Experiments
Part D. Based on your analysis in Parts B and C, for making further quality improvements.
what would you recommend? 6
“GM Fell Behind Toyota in 2008.” 2009. The Wall Street Journal
Part E. In Part C, we found that the process is in (January 22). (Toyota sold almost nine million vehicles in 2008
control when we set the temperature in the diffusion worldwide, while GM sold about 8.3 million.)
 Mirchandani: Case: The MotoTech Manufacturing Company: Process Control and Improvement
82 INFORMS Transactions on Education 10(2), pp. 79–84, © 2010 INFORMS

Table 1 Pre_Improvement Data

Additional information, including supplemental material and rights and permission policies, is available at http://ite.pubs.informs.org.

Data and Initial Computations

Thickness Thickness
(in thousands of Å) (in thousands of Å)
Sample Sample Sample Sample
Sample no. 1 2 3 mean range Sample no. 1 2 3 mean range

1 3
265 3
097 3
201 3
1877 0
1680 38 3
170 3
171 3
229 3
1900 0
0590
2 3
273 3
244 3
125 3
2140 0
1480 39 3
067 3
146 3
178 3
1303 0
1110
Downloaded from informs.org by [14.139.108.34] on 12 July 2017, at 23:50 . For personal use only, all rights reserved.

3 3
197 3
174 3
186 3
1857 0
0230 40 3
819 3
898 3
309 3
6753 0
5890
4 3
215 3
158 3
122 3
1650 0
0930 41 3
216 3
253 3
310 3
2597 0
0940
5 3
321 3
200 3
320 3
2803 0
1210 42 3
156 3
200 3
067 3
1410 0
1330
6 3
253 3
254 3
177 3
2280 0
0770 43 3
141 3
019 3
162 3
1073 0
1430
7 3
186 3
172 3
138 3
1653 0
0480 44 3
226 3
236 3
250 3
2373 0
0240
8 3
255 2
927 3
173 3
1183 0
3280 45 3
247 3
250 3
233 3
2433 0
0170
9 3
195 3
145 3
125 3
1550 0
0700 46 3
164 3
051 3
136 3
1170 0
1130
10 3
205 3
160 3
131 3
1653 0
0740 47 3
012 3
260 3
202 3
1580 0
2480
11 3
008 3
098 3
094 3
0667 0
0900 48 3
209 3
002 3
174 3
1283 0
2070
12 3
191 3
122 3
151 3
1547 0
0690 49 3
125 3
132 3
134 3
1303 0
0090
13 3
108 3
103 3
105 3
1053 0
0050 50 3
111 3
023 3
177 3
1037 0
1540
14 3
005 3
301 3
010 3
1053 0
2960 51 3
306 3
119 3
133 3
1860 0
1870
15 3
228 3
007 3
000 3
0783 0
2280 52 3
170 3
203 3
210 3
1943 0
0400
16 3
080 3
069 3
063 3
0707 0
0170 53 3
011 3
197 3
076 3
0947 0
1860
17 3
152 3
029 3
202 3
1277 0
1730 54 3
137 3
031 3
039 3
0690 0
1060
18 3
144 3
133 2
993 3
0900 0
1510 55 3
178 3
003 3
193 3
1247 0
1900
19 3
244 3
308 3
202 3
2513 0
1060 56 3
105 3
068 3
249 3
1407 0
1810
20 3
062 3
100 3
091 3
0843 0
0380 57 3
044 3
005 3
010 3
0197 0
0390
21 3
141 3
025 3
113 3
0930 0
1160 58 3
164 3
058 3
197 3
1397 0
1390
22 3
179 3
037 3
207 3
1410 0
1700 59 3
008 3
025 3
053 3
0287 0
0450
23 3
180 3
137 3
171 3
1627 0
0430 60 3
236 3
091 3
271 3
1993 0
1800
24 3
195 3
098 3
173 3
1553 0
0970 61 3
011 2
982 2
959 2
9840 0
0520
25 2
960 3
051 3
046 3
0190 0
0910 62 3
170 3
176 3
098 3
1480 0
0780
26 3
175 3
215 3
096 3
1620 0
1190 63 3
304 3
243 3
064 3
2037 0
2400
27 3
286 3
212 3
265 3
2543 0
0740 64 3
037 2
968 2
959 2
9880 0
0780
28 3
005 3
123 3
138 3
0887 0
1330 65 3
075 3
174 3
027 3
0920 0
1470
29 3
065 3
062 3
030 3
0523 0
0350 66 3
255 3
245 3
241 3
2470 0
0140
30 3
173 3
177 3
075 3
1417 0
1020 67 3
273 3
067 3
000 3
1133 0
2730
31 3
068 3
090 3
064 3
0740 0
0260 68 3
108 3
132 2
904 3
0480 0
2280
32 2
919 2
970 2
979 2
9560 0
0600 69 3
001 2
987 3
064 3
0173 0
0770
33 3
001 3
021 2
930 2
9840 0
0910 70 2
978 2
890 2
952 2
9400 0
0880
34 3
250 3
127 3
230 3
2023 0
1230 71 3
116 3
110 3
115 3
1137 0
0060
35 3
179 3
104 3
194 3
1590 0
0900 72 3
022 3
007 2
997 3
0087 0
0250
36 3
109 3
100 3
227 3
1453 0
1270
37 2
996 3
042 3
063 3
0337 0
0670 Mean 3
1326 0
1165

Table 2 Post_Improvement Data

Data and Initial Computations

Thickness Thickness
(in thousands of Å) (in thousands of Å)
Sample Sample Sample Sample
Sample no. 1 2 3 mean range Sample no. 1 2 3 mean range

1 3
039 3
011 3
095 3
0483 0
0840 38 3
032 3
105 3
111 3
0827 0
0790
2 3
041 3
071 3
049 3
0537 0
0300 39 3
157 3
079 3
014 3
0833 0
1430
3 3
047 3
089 3
102 3
0793 0
0550 40 3
073 3
049 3
056 3
0593 0
0240
4 3
074 3
047 3
019 3
0467 0
0550 41 3
068 3
041 3
042 3
0503 0
0270
5 3
081 3
109 3
010 3
0667 0
0990 42 3
059 3
024 2
970 3
0177 0
0890
6 3
018 3
148 3
030 3
0653 0
1300 43 3
057 3
109 3
068 3
0780 0
0520
7 3
061 3
093 3
094 3
0827 0
0330 44 3
003 3
032 3
046 3
0270 0
0430
8 3
075 3
080 3
074 3
0763 0
0060 45 3
025 3
051 3
066 3
0473 0
0410
9 3
042 3
067 3
019 3
0427 0
0480 46 3
046 3
105 3
074 3
0750 0
0590
10 3
083 3
133 3
057 3
0910 0
0760 47 3
060 3
133 3
121 3
1047 0
0730
 Mirchandani: Case: The MotoTech Manufacturing Company: Process Control and Improvement
INFORMS Transactions on Education 10(2), pp. 79–84, © 2010 INFORMS 83

Table 2 Post_Improvement Data (Continued)

Additional information, including supplemental material and rights and permission policies, is available at http://ite.pubs.informs.org.

Data and Initial Computations

Thickness Thickness
(in thousands of Å) (in thousands of Å)
Sample Sample Sample Sample
Sample no. 1 2 3 mean range Sample no. 1 2 3 mean range

11 3
057 3
119 3
105 3
0937 0
0620 48 3
108 3
007 3
108 3
0743 0
1010
Downloaded from informs.org by [14.139.108.34] on 12 July 2017, at 23:50 . For personal use only, all rights reserved.

12 3
086 3
040 3
021 3
0490 0
0650 49 3
068 3
105 3
007 3
0600 0
0980
13 3
084 3
040 3
087 3
0703 0
0470 50 3
037 3
060 3
074 3
0570 0
0370
14 3
122 3
056 3
033 3
0703 0
0890 51 3
047 3
056 3
067 3
0567 0
0200
15 3
132 2
987 3
048 3
0557 0
1450 52 3
008 3
052 3
004 3
0213 0
0480
16 3
096 3
044 3
126 3
0887 0
0820 53 3
058 3
059 3
095 3
0707 0
0370
17 3
048 3
103 3
057 3
0693 0
0550 54 3
075 3
017 3
043 3
0450 0
0580
18 3
067 3
040 3
108 3
0717 0
0680 55 3
042 3
133 3
111 3
0953 0
0910
19 3
014 3
004 3
062 3
0267 0
0580 56 3
042 3
048 3
041 3
0437 0
0070
20 3
089 3
006 3
038 3
0443 0
0830 57 3
072 3
020 3
034 3
0420 0
0520
21 3
094 3
066 3
005 3
0550 0
0890 58 3
031 2
981 3
053 3
0217 0
0720
22 3
031 3
139 3
106 3
0920 0
1080 59 3
023 3
105 3
075 3
0677 0
0820
23 3
051 3
086 3
058 3
0650 0
0350 60 3
043 3
077 3
085 3
0683 0
0420
24 3
086 3
018 3
104 3
0693 0
0860 61 3
032 3
073 3
060 3
0550 0
0410
25 3
045 3
026 3
036 3
0357 0
0190 62 3
094 2
991 3
080 3
0550 0
1030
26 3
039 2
980 3
034 3
0177 0
0590 63 3
017 3
046 3
044 3
0357 0
0290
27 3
047 3
030 3
099 3
0587 0
0690 64 3
070 3
099 2
970 3
0463 0
1290
28 2
961 3
076 3
015 3
0173 0
1150 65 3
063 2
970 3
057 3
0300 0
0930
29 3
044 3
053 3
027 3
0413 0
0260 66 3
073 3
120 3
086 3
0930 0
0470
30 3
051 2
982 3
137 3
0567 0
1550 67 3
078 3
075 3
109 3
0873 0
0340
31 3
071 3
046 3
074 3
0637 0
0280 68 3
062 3
129 3
002 3
0643 0
1270
32 3
101 3
093 3
007 3
0670 0
0940 69 3
078 3
113 3
051 3
0807 0
0620
33 3
073 2
998 3
103 3
0580 0
1050 70 3
046 3
069 3
079 3
0647 0
0330
34 3
096 3
071 3
040 3
0690 0
0560 71 3
152 3
073 3
053 3
0927 0
0990
35 3
075 3
085 3
049 3
0697 0
0360 72 3
035 3
094 3
127 3
0853 0
0920
36 3
046 3
098 3
046 3
0633 0
0520
37 3
079 3
030 3
135 3
0813 0
1050 Mean 3
0613 0
0677

Planning for the more distant future, John Tagole the current equipment) leasing and operating cost of
wants to evaluate if modernizing the diffusion equip- $15,000,000 per year. This new machine will have a
ment is economically worthwhile. He does not want process standard deviation of 15 Å, but as with the
to purchase new equipment (which can cost several current equipment, it will be difficult to detect process
tens of millions of dollars), because Intel has recently mean shifts of up to 1.5 sigma. The capacity of the
claimed7 that it has found a replacement material for new machine will be the same as that of the current
silicon dioxide; the claim, if substantiated, will make machine. John Tagole calls The Famous Five again
the current chip manufacturing technology obsolete in to evaluate whether the MM should lease the new
the next five to seven years. Even if Intel’s technology equipment.
is unsuccessful, IBM’s new technology, which allows John Tagole informs The Famous Five that MM pro-
chips to consume less power, might become available duces a million wafers per year. MM has an enviable
in this time frame.8 (Power consumption by chips has brand equity that allows it to sell its entire produc-
become a major issue with computer manufacturers.) tion and has led to a sales growth rate of 20% per
To summarize: John Tagole’s current planning hori- year. Now that the quality has been improved, John
zon is five to seven years. MM’s equipment sup- Tagole expects this growth rate to continue in the near
plier has informed John Tagole that MM will be future. You therefore expect the production and sales
able to lease, on an annual basis, very precise diffu- to increase by 20% per year, starting from the current
sion equipment at an incremental (over the cost of value of a million wafers per year.
The wafers have a unit contribution of $150 unless
7
“Intel claims a breakthrough in chip design,” 2003. The Wall Street any rework is needed. Because Semicon is a major
Journal (November 5). MM customer, The Famous Five decide to do the
8
“Start-Up’s power-saving chip to be based on IBM technology.” analysis using Semicon’s specifications which allow
2005. The Wall Street Journal (October 24). thicknesses between 2,900 and 3,100 Å. John Tagole
 Mirchandani: Case: The MotoTech Manufacturing Company: Process Control and Improvement
84 INFORMS Transactions on Education 10(2), pp. 79–84, © 2010 INFORMS

informs The Famous Five that the rework costs for 1. Should The Famous Five recommend that MM
Additional information, including supplemental material and rights and permission policies, is available at http://ite.pubs.informs.org.

wafers with thicknesses in different ranges is as lease the new equipment?9 Why or why not? (In the
follows: computation of the expected costs, you can ignore the
time value of money because some of you have not
Thickness between Rework cost yet taken the financial management course and may
not know how to incorporate it. You can also ignore
(in Å) per unit ($)
inflation.)
2,860 to 2,880 125 2. What else would you recommend for the future?
Downloaded from informs.org by [14.139.108.34] on 12 July 2017, at 23:50 . For personal use only, all rights reserved.

2,880 to 2,900 50
3,100 to 3,120 50 Supplementary Material
3,120 to 3,140 125 Files that accompany this paper can be found and
downloaded from http://ite.pubs.informs.org.
Thus, the unit contribution for a wafer that orig-
inally had a thickness of 3,110 Å gets reduced to Acknowledgments
$150 − $50 = $100 because of rework. Any wafers with I am grateful to Dr. Skip Weed, formerly of Motorola Uni-
a thickness exceeding 3,140 Å or less than 2,860 Å versity, for suggesting the semiconductor environment as a
cannot be reworked; for these wafers, MM must backdrop of the case and providing a short description of
expense the variable cost of $500, which it has already the production process. The actual problem situation has
incurred. been modified for pedagogical reasons.

9
To find the probabilities, you can use the normal tables, or you
can use Excel’s = NORMDIST function.