Chapter 08

Wallace J. Hopp, Mark L. Spearman, 1996, 2000 http://www.factory-physics.
com
1
Variability Basics
God does not play dice with the universe.
Albert Einstein
Stop telling God what to do.
Niels Bohr
Wallace J. Hopp, Mark L. Spearman, 1996, 2000 http://www.factory-physics.com
2
Variability Makes a Difference!
Littles Law: TH = WIP/CT, so same throughput can be
obtained with large WIP, long CT or small WIP, short CT.
The difference?

Penny Fab One: achieves full TH (0.5 j/hr) at WIP=W
0
=4 jobs if
it behaves like Best Case, but requires WIP=27 jobs to achieve
95% of capacity if it behaves like the Practical Worst Case.
Why?

3
Tortise and Hare Example
Two machines:
subject to same workload: 69 jobs/day (2.875 jobs/hr)
subject to unpredictable outages (availability = 75%)

Hare X19:
long, but infrequent outages

Tortoise 2000:
short, but more frequent outages

Performance: Hare X19 is substantially worse on all measures than
Tortoise 2000. Why?
4
Variability Views
Variability:
Any departure from uniformity
Can be random or controllable

Randomness:
Essential reality?
Artifact of incomplete knowledge?
Management implications: robustness is key
5
Probabilistic Intuition
Uses of Intuition:
Driving a car
Throwing a ball
Making investments

First Moment Effects:
Throughput increases with machine speed
Throughput increases with availability
Inventory increases with lot size
Our intuition is good for first moments
g
6
Probabilistic Intuition (cont.)
Second Moment Effects:
Which are more variable processing times of parts or batches?
Which are more disruptive long, infrequent failures or short frequent
ones?
Which helps more increasing times to failure or reducing times to
repair?
Our intuition is less secure for second moments, so we make more
errors (e.g., regression to the mean).

7
Variability
Definition: Variability is anything that causes the system to
depart from regular, predictable behavior.

Sources of Variability:
machine failures workpace variation
setups differential skill levels
material shortages material handling
yield loss demand fluctuations
rework engineering change orders
operator unavailability product variety
May be consequence
of business strategy
May be consequence
of manufacturing
practices
8
Measuring Process Variability
CV , variation of t coefficien
time process of deviation standard
job a of time process mean
= =
=
=
e
e
e
e
e
t
c
t
o
Note: we often use the squared
coefficient of variation (SCV), c
e
2

9
Variability Classes in Factory Physics

Effective Process Times:
actual process times are generally LV
effective process times include setups, failure outages, etc.
HV, LV, and MV are all possible in effective process times

Relation to Performance Cases: For balanced systems
MV Practical Worst Case
LV between Best Case and Practical Worst Case
HV between Practical Worst Case and Worst Case
0.75
High variability
(HV)
Moderate variability
(MV)
Low variability
(LV)
0 1.33
c
e

10
Measuring Process Variability Example
Trial Machine 1 Machine 2 Machine 3
1 22 5 5
2 25 6 6
3 23 5 5
4 26 35 35
5 24 7 7
6 28 45 45
7 21 6 6
8 30 6 6
9 24 5 5
10 28 4 4
11 27 7 7
12 25 50 500
13 24 6 6
14 23 6 6
15 22 5 5
t
e
25.1 13.2 43.2
s
e
2.5 15.9 127.0
c
e
0.1 1.2 2.9
Class LV MV HV
11
Natural Variability
Definition: variability without explicitly analyzed cause

Sources:
operator pace
material fluctuations
product type (if not explicitly considered)
product quality

Observation: natural process variability is usually in the LV
category.
12
Down Time Mean Effects
Definitions:

) / ( es repair tim of ty variabili of coefficent
repair to mean time
failure to mean time
parts/hr) e.g., (rate, capacity base
1
ty variabili of t coefficien time process base
time process base
0
0
0
0
r r r
r
f
m c
m
m
t
r
c
t
o =
=
=
= =
=
=
13
Down Time Mean Effects (cont.)
Availability: Fraction of time machine is up

Effective Processing Time and Rate:

r f
f
m m
m
A
+
=
A t t
Ar r
e
e
/
0
0
=
=
14
Totoise and Hare - Availability
Hare X19:
t
0
= 15 min
o
0
= 3.35 min
c
0
= o
0
/t
0
= 3.35/15 = 0.05
m
f
= 12.4 hrs (744 min)
m
r
= 4.133 hrs (248 min)
c
r
= 1.0

Availability:

Tortoise:
t
0
= 15 min
o
0
= 3.35 min
c
0
= o
0
/t
0
= 3.35/15 = 0.05
m
f
= 1.9 hrs (114 min)
m
r
= 0.633 hrs (38 min)
c
r
= 1.0

A = A =
15
Down Time Variability Effects
Effective Variability:

Conclusions:
Failures inflate mean, variance, and CV of effective process time
Mean (t
e
) increases proportionally with 1/A
SCV (c
e
2
) increases proportionally with m
r
SCV (c
e
2
) increases proportionally in c
r
2

For constant availability (A), long infrequent outages increase SCV
more than short frequent ones

0
2 2
0
2
2
2
0
2 2
2
0 2
0
) 1 ( ) 1 (
) 1 )( (
/
t
m
A A c c
t
c
A
t A m
A
A t t
r
r
e
e
e
r r
e
e
+ + = =
+
+ |
.
|
\
|
=
=
o
o o Variability
depends on
repair times
in addition to
availability
16
Tortoise and Hare - Variability
Hare X19:

t
e
=

c
e
2
=
Tortoise 2000

t
e
=

c
e
2
=

17
Impact of Variability
Hare X19

CT = 28 hours
WIP = 81 jobs
Tortoise 2000

CT = 8 hours
WIP = 23 jobs
Hare
Tor-
toise
2.875 jobs/hr 2.875 jobs/hr
Conclusion: Capacity and arrival variability
are the same. CT and WI P are greatly
inflated by process variability due to failures
18
Setups Mean and Variability Effects
Analysis:

2
2
2
2
2
2
2
0
2
0
1
time setup of dev. std.
duration setup average
setups between jobs no. average
e
e
e
s
s
s
s
s
e
s
s
e
s
s
s
s
s
s
t
c
t
N
N
N
N
t
t t
t
c
t
N
o
o
o
o
o
=
+ + =
+ =
=
=
=
=
Capacity Effect setups
inflate average process time
Variability Effect setups
also inflate process time CV
19
Setups Mean and Variability Effects (cont.)
Observations:
Setups increase mean and variance of processing times.
Variability reduction is one benefit of flexible machines.
However, the interaction is complex.

20
Setup Example
Data:
Fast, inflexible machine 2 hr setup every 10 jobs

Slower, flexible machine no setups

Traditional Analysis?
jobs/hr 8333 . 0 ) 10 / 2 1 /( 1 / 1
hrs 2 . 1 10 / 2 1 /
hrs 2
jobs/setup 10
hr 1
0
0
= + = =
= + = + =
=
=
=
e e
s s e
s
s
t r
N t t t
t
N
t
jobs/hr 833 . 0 2 . 1 / 1 / 1
hrs 1.2
0
0
= = =
=
t r
t
e
No difference!
21
Setup Example (cont.)
Factory Physics Approach: Compare mean and variance
Fast, inflexible machine 2 hr setup every 10 jobs
31 . 0
4475 . 0
1
jobs/hr 8333 . 0 ) 10 / 2 1 /( 1 / 1
hrs 2 . 1 10 / 2 1 /
0625 . 0
hrs 2
jobs/setup 10
0625 . 0
hr 1
2
2
2
2 2
0
2
0
2
2
0
0
=
=
|
|
.
|
\
|

+ + =
= + = =
= + = + =
=
=
=
=
=
e
s
s
s
s
s e
e e
s s e
s
s
s
c
N
N
N
c
t
t r
N t t t
c
t
N
c
t
o
22
Slower, flexible machine no setups

Conclusion:
0625 . 0
jobs/hr 833 . 0 2 . 1 / 1 / 1
0625 . 0
hrs 2 . 1
2
0
2
0
2
0
0
= =
= = =
=
=
c c
t r
c
t
e
e
23
New Machine: Consider a third machine same as previous machine
with setups, but with shorter, more frequent setups

Analysis:

Conclusion:
hr 1
jobs/setup 5
=
=
s
s
t
N
16 . 0
2350 . 0
1
jobs/hr 833 . 0 ) 5 / 1 1 /( 1 / 1
2
2
2
2 2
0
2
=
=
|
|
.
|
\
|

+ + =
= + = =
e
s
s
s
s
s e
e e
c
N
N
N
c
t
t r
o
24
Other Process Variability Inflators
Sources:
operator unavailability
recycle
batching
material unavailability
et cetera, et cetera, et cetera
Effects:
inflate t
e

inflate c
e

Consequences:
25
Flow Variability
Process Variability is bad enough
Inflates CT
Inflates WIP
Forces lower utilization of capacity

But, variability also propagates
Causes uneven arrivals downstream
Inflates CT and WIP at other stations
Forces lower utilization of capacity throughout the line
26
Illustrating Flow Variability
t
Low variability arrivals
t
High variability arrivals
27
Measuring Flow Variability
times al interarriv of variation of t coefficien
arrivals between time of deviation standard
rate arrival
1
arrivals between mean time
= =
=
= =
=
a
a
a
a
a
a
a
t
c
t
r
t
o
o
28
Propagation of Variability
High Utilization Station
HV
LV HV
HV
HV
HV
LV
LV
LV
LV
LV HV
Conclusion: flow variability out of a high utilization station is
determined primarily by process variability at that station.
29
Low Utilization Station
HV
HV
HV
LV
LV
LV
Conclusion: flow variability out of a low utilization station is
determined primarily by flow variability into that station.
HV
LV
LV
LV
HV HV
30

Single Machine Station:

where u is the station utilization given by u = r
a
t
e

Multi-Machine Station:

where m is the number of (identical) machines and
2 2 2 2 2
) 1 (
a e d
c u c u c + =
) 1 ( ) 1 )( 1 ( 1
2
2
2 2 2
+ + =
e a d
c
m
u
c u c
c
d
2
(i) = c
a
2
(i+1)
m
t r
u
e a
=
i i+1
departure var
depends on
arrival var
and process
var
c
e
2
(i)
c
a
2
(i)
31
Variability Interactions
Importance of Queueing:
manufacturing plants are queueing networks
queueing and waiting time comprise majority of cycle time

System Characteristics:
Arrival process
Service process
Number of servers
Maximum queue size (blocking)
Service discipline (FCFS, LCFS, EDD, SPT, etc.)
Balking
Routing
Many more
32
Kendall's Classification
A/B/C

A: arrival process
B: service process
C: number of machines

M: exponential (Markovian) distribution
G: completely general distribution
D: constant (deterministic) distribution.
A
B
C
Queue
Server
33
Queueing Parameters
r
a
= the rate of arrivals in customers (jobs) per unit
time (t
a
= 1/r
a
= the average time between
arrivals).
c
a
= the CV of inter-arrival times.
m = the number of machines.
r
e
= the rate of the station in jobs per unit time =
m/t
e
.
c
e
= the CV of effective process times.

u = utilization of station = r
a
/r
e
.
Note: a station
can be
described
with 5
parameters.
34
Queueing Measures
Measures:
CT
q
= the expected waiting time spent in queue.
CT = the expected time spent at the process center, i.e., queue time
plus process time.
WIP = the average WIP level (in jobs) at the station.
WIP
q
= the expected WIP (in jobs) in queue.

Relationships:
CT = CT
q
+ t
e

WIP = r
a
CT
WIP
q
= r
a
CT
q

Result: If we know CT
q
, we can compute WIP, WIP
q
, CT.
35
The G/G/1 Queue
Formula:

Observations:
Useful model of single machine workstations
Separate terms for variability, utilization, process time.
CT
q
(and other measures) increase with c
a
2
and c
e
2

Flow variability, process variability, or both can combine to inflate
queue time.
Variability causes congestion!
e
e a
q
t
u
u c c
t U V
|
.
|
\
|
|
|
.
|
\
|
+
~
~
1 2
CT
2 2
36
The G/G/m Queue
Formula:

Observations:
Useful model of multi-machine workstations
Extremely general.
Fast and accurate.
Easily implemented in a spreadsheet (or packages like MPX).
e
m
e a
q
t
u m
u c c
t U V
|
|
.
|
\
|
|
|
.
|
\
| +
~
~
+
) 1 ( 2
CT
1 ) 1 ( 2 2 2
37
b
a
s
i
c

d
a
t
a

f
a
i
l
u
r
e
s

s
e
t
u
p
s

y
i
e
l
d

m
e
a
s
u
r
e
s

VUT Spreadsheet
MEASURE: STATION: 1 2 3 4 5
Arrival Rate (parts/hr) r
a
10.000 9.800 9.310 8.845 7.960
Arrival CV c
a
2
1.000 0.181 0.031 0.061 0.035
Natural Process Time (hr) t
0
0.090 0.090 0.095 0.090 0.090
Natural Process SCV c0
2
0.500 0.500 0.500 0.500 0.500
Number of Machines m 1 1 1 1 1
MTTF (hr) m
f
200 200 200 200 200
MTTR (hr) m
r
2 2 8 4 4
Availability A 0.990 0.990 0.962 0.980 0.980
Effective Process Time (failures only) t
e
' 0.091 0.091 0.099 0.092 0.092
Eff Process SCV (failures only) c
e
2
' 0.936 0.936 6.729 2.209 2.209
Batch Size k 100 100 100 100 100
Setup Time (hr) t
s
0.000 0.500 0.500 0.000 0.000
Setup Time SCV c
s
2
1.000 1.000 1.000 1.000 1.000
Arrival Rate of Batches r
a
/k 0.100 0.098 0.093 0.088 0.080
Eff Batch Process Time (failures+setups) t
e
= kt
0
/A+t
s
9.090 9.590 10.380 9.180 9.180
Eff Batch Process Time Var (failures+setups) k*o0
2
/A
2
+ 2m
r
(1-A)kt
0
/A+os
2
0.773 1.023 6.818 1.861 1.861
Eff Process SCV (failures+setups) c
e
2
0.009 0.011 0.063 0.022 0.022
Utilization u 0.909 0.940 0.966 0.812 0.731
Departure SCV c
d
2
0.181 0.031 0.061 0.035 0.028
Yield y 0.980 0.950 0.950 0.900 0.950
Final Departure Rate r
a
*y 9.800 9.310 8.845 7.960 7.562
Final Departure SCV yc
d
2
+(1-y) 0.198 0.079 0.108 0.132 0.077
Utilization u 0.909 0.940 0.966 0.812 0.731
Throughput TH 9.800 9.310 8.845 7.960 7.562
Queue Time (hr) CT
q
45.825 14.421 14.065 1.649 0.716
Cycle Time (hr) CT
q
+t
e
54.915 24.011 24.445 10.829 9.896
Cumulative Cycle Time (hr) Ei
(CT
q
(i)+t
e
(i)) 54.915 78.925 103.371 114.200 124.096
WIP in Queue (jobs) r
a
CT
q
458.249 141.321 130.948 14.587 5.700
WIP (jobs) r
a
CT 549.149 235.303 227.586 95.780 78.773
Cumulative WIP (jobs) Ei
(r
a
(i)CT(i)) 549.149 784.452 1012.038 1107.818 1186.591
38
Effects of Blocking
VUT Equation:
characterizes stations with infinite space for queueing
useful for seeing what will happen to WIP, CT without restrictions

But real world systems often constrain WIP:
physical constraints (e.g., space or spoilage)
logical constraints (e.g., kanbans)

Blocking Models:
estimate WIP and TH for given set of rates, buffer sizes
much more complex than non-blocking (open) models, often
require simulation to evaluate realistic systems
39
The M/M/1/b Queue
) 1 ( / ) 2 ( where
) / 1 / / (
) / 1 / / (
) / 1 / / (
1
1
) / 1 / / (
1
) 1 (
1
) / 1 / / (
1
1
1
e e
a
b
b
b
b
t t u
b M M TH
b M M WIP
b M M CT
r
u
u
b M M TH
u
u b
u
u
b M M WIP
=
=
=
+
+
+
B buffer spaces
Infinite
raw
materials
Model of Station 2
1
2
Note: there is room
for b=B+2 jobs in
system, B in the buffer
and one at each station.
Goes to u/(1-u) as b
Always less than WI P(M/M/1)
Goes to r
a
as b
Always less than TH(M/M/1)
Littles law
Note: u>1 is possible; formulas valid for u=1
40
Blocking Example
B=2
t
e
(1)=21 t
e
(2)=20
jobs 8954 . 1
9524 . 0 1
) 9524 . 0 ( 5
20
1
) 1 (
1
) / 1 / / (
job/min 039 . 0
21
1
9524 . 0 1
9524 . 0 1
1
1
1
job/min 0476 . 0 21 / 1 ) 1 ( / 1 ) 1 / / (
jobs 20
1
) 1 / / (
9524 . 0 21 / 20 ) 1 ( / ) 2 (
5
5
1
1
5
4
1
=
=
= |
.
|
\
|
= =
= = = =
=
=
= = =
+
+
+
b
b
a
b
b
e a
e e
u
u b
u
u
b M M WIP
r
-u
-u
/b) TH(M/M/
t r M M TH
u
u
M M WIP
t t u
M/M/1/b system has
less WI P and less TH
than M/M/1 system
18% less TH
90% less WI P
41
Seeking Out Variability
General Strategies:
look for long queues (Little's law)
look for blocking
focus on high utilization resources
consider both flow and process variability
ask why five times

Specific Targets:
equipment failures
setups
rework
operator pacing
anything that prevents regular arrivals and process times
42
Variability Pooling
Basic Idea: independent sources of variability tend to cancel each
other out, reducing total amount of variability.

Example (Time to process a batch of parts):
n
c
batch c
n
c
nt t n
n
batch t
batch
batch c
n batch
nt batch t
t
c
t
0
0
2
0
2
0
2
0
2
0
2
2
0
2
0
2
0 2
0
2
0
2
0
0 0
0
0
0
0
0
) (
) (
) (
) (
) (
) (
part single process to time of CV
part single process to time of deviation standard
part single process to time
= = = = =
=
=
= =
=
=
o o o
o o
o
o
Batches are less variable
than parts, because high
and low process times
average out.
43
Safety Stock Pooling Example
PCs consist of 6 components (CPU, HD, CD ROM, RAM,
removable storage device, keyboard)
3 choices of each component: 3
6
= 729 different PCs
Each component costs $150 ($900 material cost per PC)
Demand for all models is normally distributed with mean 100 per
year, standard deviation 10 per year
Replenishment lead time is 3 months, so average demand during
LT is u =25 for computers and u =25(729/3) =6075 for
components
Use base stock policy with fill rate of 99%
44
Pooling Example - Stock PCs
Base Stock Level for Each PC:

R = u + z
s
o = 25 + 2.33(\ 25) = 37

On-Hand Inventory for Each PC:

I(R) = R - u + B(R) ~ R - u = z
s
o = 37 - 25 = 12 units

Total (Approximate) On-Hand Inventory :

12 729 $900 = $7,873,200

cycle stock
safety stock
45
Pooling Example - Stock Components
Necessary Service for Each Component:
S = (0.99)
1/6
= 0.9983 z
s
= 2.93

Base Stock Level for Each Component:

R = u + z
s
o = 6075 + 2.93(\ 6075) = 6303

On-Hand Inventory Level for Each Component:

I(R) = R - u + B(R) ~ R - u = z
s
o = 6303-6075 = 228 units

Total Safety Stock:

228 18 $150 = $615,600

cycle stock
safety stock
92% reduction!
46
Basic Variability Takeaways
Variability Measures:
CV of effective process times
CV of interarrival times
Components of Process Variability
failures
setups
many others - deflate capacity and inflate variability
long infrequent disruptions worse than short frequent ones
Consequences of Variability:
variability causes congestion (i.e., WIP/CT inflation)
variability propagates
variability and utilization interact
pooled variability less destructive than individual variability

Chapter 08

Enviado por

Dados do documento

Descrição original:

Direitos autorais

Formatos disponíveis

Compartilhar este documento

Compartilhar ou incorporar documento

Opções de compartilhamento

Você considera este documento útil?

Este conteúdo é inapropriado?

Direitos autorais:

Formatos disponíveis

Chapter 08

Enviado por

Direitos autorais:

Formatos disponíveis

Wallace J. Hopp, Mark L. Spearman, 1996, 2000 http://www.factory-physics.

Você também pode gostar