Académique Documents
Professionnel Documents
Culture Documents
Better Together
W
hich is better: lean or Six Sigma? The question is absurd, yet it
Combining lean is often debated among continuous improvement professionals.
and Six Sigma Through a simple classroom or group demonstration, participants
can gain an understanding of the advantages of lean and the advantages of
leads to Greater Six Sigma. More importantly, however, they will see how beneficial it is when
lean and Six Sigma work together.
Success Lean and Six Sigma can exist separately, but the benefits of bringing
them together are tremendous, such as the alignment of an organiza-
tions resources and the creation of a single improvement strategy. If kept
separate, the disadvantages are just as great: Employees are sent confusing
messages for improvement, and competition is established for continuous
improvement resources.
Six Sigma can improve the effectiveness of a system by optimizing it and by
reducing process variability. Lean can improve the efficiency of that system by
stabilizing work in process (WIP), reducing inventory and eliminating waste.
A traditional process includes so much WIP that the people involved have
no idea where to look for problems. Consider a process in which a large
amount of WIP must account for a process with lots of variability. In this
process, some imaginary work (or value-added processing) is accomplished
By Lloyd C. at each of four work stations.
The stations are in series, so A must complete its task before B can
Patterson, Master perform its task. Each station draws from a bufferor inventoryof goods
located just before it in the flow. Variations at each station, such as defects
Black Belt, ITT and downtime, cause output levels to vary.
Communications The process flow (see Figure 1) works like this:
1. The buffer at each work station and the warehouse starts with 16
Systems items. Each area draws from its buffer of items. Items sent or pushed
to the next station are not available until the next day.
2. Receiving sends seven items to the buffer of the first work station each
day.
3. The number of items each station completes is based on the roll of
a pair of dice. The number rolled is the amount of items sent to the
buffer of the next station.
4. The last station moves a certain number of items to the warehouse,
again based on the roll of the dice.
5. Shipping sends seven items each day from the warehouse to the cus-
tomer.
The customer expects seven items per day. Because the roll of two dice
results in seven more than any other number, each station will move an aver-
age of seven units. Items can be represented by poker chips or other small
items that are easily stacked and counted. Round one requires 150 items.
ice
ice
ice
Available
Available
Available
Available
Available
day
day
d
d
wo
wo
wo
wo
per
per
of t
of t
of t
of t
en
en
Sev
Sev
Sum
Sum
Sum
Sum
Receiving Station A Station B Station C Station D Shipping
Lean applied
dice
dice
dice
Available
Available
Available
Available
Available
day
day
wo
wo
wo
wo
per
per
of t
of t
of t
of t
en
en
Sev
Sev
Sum
Sum
Sum
Sum
Receiving Station A Station B Station C Station D Shipping
15
the numbers one through six displayed. Together,
they can display two through 12, with seven being
10
most likely. That should keep deliveries at seven items
per day, but a low roll cant always be compensated
5 for if the buffers begin small and variability is high
(see Figure 3).
0
But what if we used a custom set of dice at each
2 4 6 7 8 10 12 station (see Figure 4)? The custom dice only work
Result of rolling two dice when properly paired. One needs to have four sides
Existing dot
Existing dot removed
New dot added
with four dots, and the other needs to have four sides duce a total of six, seven or eight. The number seven
with three dots. is most likelymatching customer expectationbut
This can be accomplished by placing the correct variability is greatly reduced (see Figure 6).
pair in a clear plastic case large enough to roll the dice
in or by using multi-colored dice so matching colors Round threereduce variability
create a correct pair.
To remove dots, use a small drill to clean out the For the final round, run the factory for 10 days, but
paint. Drill a small indentation, and use paint to add start with a WIP of eight items at each station and use
new dots. the custom dice, which reduce variability (see Figure
Together, the custom dice (see Figure 5) will pro- 7, p. 14).
25
Mean 7
Standard 0.6761
deviation
20
Number 36
Frequency
15
10
0
2 4 6 7 8 10 12
Result of rolling two dice
dice
dice
dice
Available
Available
Available
Available
Available
tom
tom
tom
tom
day
day
cus
cus
cus
cus
per
per
wo
wo
wo
wo
en
en
Sev
Sev
of t
of t
of t
of t
Sum
Sum
Sum
Sum
Receiving Station A Station B Station C Station D Shipping
How much WIP do we have at the beginning? Other combinations could be created and used
8 items x 5 locations = 40 items. to make different points. The distribution could be
How much WIP do we have at the end? further tightened, with one die having five sides with
Probably about the same. four dots and the other having five sides with three
dots. The mean could be shifted by altering the dice to
What is the PCT? have sides with four and five dots, effectively increas-
Estimated to be five to six days. ing capacity.
Was the customer satisfied? Regardless of the approach taken, this hands-on
Probably, but deliveries might fall short. exercise is an effective way to illustrate the power of
How many items were delivered? lean and Six Sigma working together.
Notes
Discussion points
This demonstration is an expansion of the one created by John McCon-
nell that illustrates one of the most vital quality principlesthe fundamen-
Round three doesnt always result in 100% delivery. tal importance of minimizing variation.
An animated and shorter version called The Dice Experiment can be
If it doesnt, should we allow more WIP or tighten found in PQ Systems Quality Game Box.
the variability even more? The answer is whichever is Special thanks to Atomic Hobbies & Games, the source of the standard
dice.
economically more feasible. In some cases, tighten-
ing the process variability is a small investment in
process changes or new equipment, and the items
being produced are expensive. Other times, the cost
of the items is small compared to the cost of variability
reduction, and larger buffers make more sense.
Should Six Sigma or lean be applied first? Lean can
lead Six Sigma in implementation, but they must be
applied together, with Six Sigma used to attack prob- What do you think of this article? Please share
lems and lean to combat waste. What will happen if
your comments and thoughts with the editor by e-mailing
just Six Sigma is applied? The financial advantages of
reduced WIP and improved PCT will not be realized. godfrey@asq.org.